LAND-USE/LAND COVER CHANGE AND ITS RELATIONSHIP TO 
CHANGING CLIMATE AND CROP PRODUCTION IN THE LOWER 
VOLTA BASIN, GHANA 
 
 
THIS THESIS IS SUBMITTED TO THE UNIVERSITY OF GHANA, LEGON IN 
PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE AWARD OF 
MPHIL ENVIRONMENTAL SCIENCE DEGREE 
 
 
 
BY 
BOSOMPEM, OHENE ASA 
(10278106) 
 
 
 
INSTITUTE FOR ENVIRONMENT AND SANITATION STUDIES (IESS),  
UNIVERSITY OF GHANA, LEGON 
 
 
 
 
 
JULY, 2015 
University of Ghana http://ugspace.ug.edu.gh
i 
 
DECLARATION 
I hereby declare that I have under supervision, undertaken this study herein submitted 
for the award of a Master of Philosophy in Environmental Science. This study is my 
own production and has not been presented anywhere for another degree whatsoever. 
The thesis report has been prepared and presented in accordance with academic rules 
and ethical conduct, I have fully cited, referenced and acknowledged all materials and 
results that are not original to this work. 
 
SIGN…………………………………  DATE………………….. 
BOSOMPEM OHENE ASA 
(STUDENT) 
 
SIGN…………………………………  DATE……………………. 
DR. TED NII YEMOH ANNANG 
(PRINCIPAL SUPERVISOR) 
 
SIGN…………………………………  DATE………………….. 
DR. OPOKU PABI 
(CO-SUPERVISOR) 
 
  
University of Ghana http://ugspace.ug.edu.gh
ii 
 
DEDICATION 
To My Family and Brenda Dzidzor Eleblu. 
  
University of Ghana http://ugspace.ug.edu.gh
iii 
 
ACKNOWLEDGEMENT 
First and foremost, I sincerely thank the Almighty God for his Grace, Protection and 
Mercies throughout this thesis work. Undoubtedly he has been my source of inspiration.  
I would like to express the deepest appreciation to my cherished supervisors, Dr. Ted 
Anang and Dr. Opoku Pabi for reviewing this work in spite of their very busy schedules. 
Their insights, energy and expertise were just indispensable to the successful 
completion of this research. The made several important contributions that enhanced 
this work and for that I am so grateful.  
My sincere thanks are also extended to Dr. Daniel Nukpezah (IESS) and Prof. Chris 
Gordon (IESS, Director) who showed me how to go about things right from the 
beginning, they have been so resourceful. 
I am grateful to Mr. Nii Amartey (GIS Technical Assistant) and Mr. Asubonteng (Geo-
Information Analyst at the UNU, Natural Resource Management Department). They 
passed unto me their in-depth knowledge of Remote Sensing and GIS.  
In addition, I would like to thank all those who provided me with the needed data for 
the study, most especially, Mr. Edem Akubia, Mrs. Paulina (GMET), Mr. Seyidu, Mr. 
Ayisi (Ada East Agriculture Department), Mr. Sefas Laryea (Ada East Planing 
Director), Mr. Larsey Sowah and Dr. Lawson (Lower Manya Krobo, Agricultural 
District Director). Without their support with data I wouldn’t have finished this work. 
The support of Mr. Manase, Mr. Isaac Adjaotor, Ms. Brenda Eleblu and all my friends 
who helped me in various ways is appreciated. God Bless you all. 
The author wishes to acknowledge the financial and resource support by the Water 
Resource Commission (WRC) and for the opportunity to be a part of the 
“Reoptimization of Operations of Akosombo & Kpong Dams on the Volta River in 
Ghana to Restore Downstream Livelihoods and Ecosystems”. I finally wish to 
acknowledge Hijmans and others for providing the EcoCrop model. 
  
University of Ghana http://ugspace.ug.edu.gh
iv 
 
ABSTRACT 
The construction of the Akosombo and Kpong dams have led to a change in the 
watershed ecology which has in turn contributed to the change in land use and land 
cover of the Lower Volta Basin (LVB). Inevitable changing global climate has also 
played a role in the alteration of land cover and crop production which have affected 
livelihoods. In this study, a post-classification comparison change detection algorithm 
was used to determine changes in land use and land cover that have taken place from 
2003-2015. A statistical analysis of rainfall and temperature variability for a period of 
five decades was studied as well as the perceptions and knowledge of farmers on 
climate change and land use/cover modification. The Diva GIS program was used to 
model and generate maps of crop suitability for current and projected future climatic 
conditions. The LUCC maps showed that between 2003 and 2015, the amount of mixed 
savannah vegetation, forest vegetation, grass lands and water decreased from 14.8% to 
9.9, 35.1 to 32.3%, 3.6 to 3.2% and 6.1% to 4.3%, respectively of the total area, while 
farmlands (agriculture) and built up areas increased from 14.8% to 29.8% and 4.2% to 
5.4%, respectively. The grasslands vegetation cover remains the largest and dominant 
land cover type and agricultural land use is second and increasing at a very fast pace. 
The small-scale shifting cultivation agriculture practiced in the LVB is the main factor 
responsible for the conversion of savannah and grassland vegetation into cropland. The 
changes in land cover is however a continuum which is influenced by changing climate. 
An increase trend in temperature and a decrease in rainfall have been identified with 
the Lower Volta Basin over the past five decades. These actual observed trends of 
rainfall and temperature affect crop yield in the area. However an increase in cropped 
area, application of fertilizer and the use of drought resistant seeds increases the general 
production of crops in the area.  
Developed cropped suitability maps of the Lower Volta Basin, show that projected 
future climatic conditions will affect crop production and thus adaptation measures 
required. 
 
 
University of Ghana http://ugspace.ug.edu.gh
v 
 
TABLE OF CONTENTS 
DECLARATION ........................................................................................................................ i 
DEDICATION ........................................................................................................................... ii 
ACKNOWLEDGEMENT ....................................................................................................... iii 
ABSTRACT .............................................................................................................................. iv 
LIST OF TABLES ............................................................................................................. viii 
LIST OF FIGURES ............................................................................................................... ix 
LIST OF PLATES .................................................................................................................. x 
LIST OF ABBREVIATIONS AND ACRONYMS .............................................................. xi 
CHAPTER ONE ........................................................................................................................ 1 
INTRODUCTION .................................................................................................................. 1 
1.1  Introduction ............................................................................................................. 1 
1.2  Problem Statement ................................................................................................... 8 
1.3  Justification ............................................................................................................ 10 
1.4  Research Questions................................................................................................ 11 
1.5  OBJECTIVES ........................................................................................................ 12 
1.6  Conceptual Framework .......................................................................................... 13 
CHAPTER TWO ..................................................................................................................... 15 
LITERATURE REVIEW ..................................................................................................... 15 
2.1  Volta Basin and Lower Volta Basin ...................................................................... 15 
2.2   Challenges associated to the Lower Volta Basin ................................................... 16 
2.3  Definition of Terms ............................................................................................... 19 
2.4  Understanding Land-use Change (Theories of Land-Use/Cover Change) ............ 20 
2.5  Land-use and cover change in Ghana and the Volta river Basin ........................... 23 
2.6  Drivers of Land-Use/Cover Changes .................................................................... 24 
2.7  Image classification ............................................................................................... 26 
2.8  Change detection ................................................................................................... 27 
2.9  Agriculture in the Ghana and the Volta River Basin ............................................. 29 
2.10  Climate change and rainfall variability ................................................................. 31 
2.11  Exposure and sensitivity of crops to changing climate ......................................... 32 
2.12  Crop suitability mapping ....................................................................................... 33 
CHAPTER THREE ................................................................................................................. 36 
MATERIALS AND METHODS ......................................................................................... 36 
University of Ghana http://ugspace.ug.edu.gh
vi 
 
3.1  Study Area ............................................................................................................. 36 
3.1.1 Population....................................................................................................... 38 
3.1.3  Climatic Conditions........................................................................................ 39 
3.1.4  Geology, Relief, and Soils.............................................................................. 40 
3.1.5  Land-use and Land Cover .............................................................................. 41 
3.1.6  Vegetation ...................................................................................................... 42 
3.2  Methods ................................................................................................................. 43 
3.2.1  Land-use/Cover Study .................................................................................... 43 
3.2.2  Climatic Study ................................................................................................ 51 
3.2.3  Crop Suitability Study .................................................................................... 53 
3.2.4  Data ................................................................................................................ 55 
3.3  Socio Economic Study........................................................................................... 56 
3.3.1  Statistical Analysis ......................................................................................... 56 
CHAPTER FOUR .................................................................................................................... 57 
RESULTS............................................................................................................................. 57 
4.1  Land-Use/Cover Classification and Change Analysis........................................... 57 
4.2  Land-Use/Cover Change (Change Detection) ....................................................... 61 
4.3  Incidence and Trend of Climate Factors ................................................................ 76 
4.3.1  Rainfall ........................................................................................................... 76 
4.3.2  Temperature ................................................................................................... 78 
4.4  Crop yield in the Ada East & Lower Manya districts ........................................... 80 
4.5  Crop suitability (EcoCrop Model) ......................................................................... 84 
4.6  Socio Economic Study........................................................................................... 89 
4.6.1 Background information from respondents ......................................................... 89 
4.6.2   Population of the LVB ................................................................................... 93 
4.6.3  Perception, Knowledge and Observed LUCC................................................ 94 
4.6.3.1  Statistical Analysis ...................................................................................... 96 
4.6.4  Knowledge and Perceptions on Climate Change ........................................... 97 
CHAPTER FIVE ............................................................................................... 103 
DISCUSSION ............................................................................................................................... 103 
5.1  Land –Use/Cover Change of Lower Volta Basin ................................................ 103 
5.2  Incidence and Trend of Climate .......................................................................... 111 
5.3  Perception and Knowledge on Land Use ............................................................ 112 
University of Ghana http://ugspace.ug.edu.gh
vii 
 
5.4  Perception and Knowledge on Climate Change .................................................. 112 
5.5  Crop Suitability.................................................................................................... 114 
5.6  Socio Economic Study......................................................................................... 115 
5.7  Agriculture In the LVB ........................................................................................ 117 
CHAPTER SIX ................................................................................................. 124 
CONCLUSION AND RECOMMENDATIONS ..................................................................... 124 
6.1  Conclusion ........................................................................................................... 124 
6.2  Recommendations ............................................................................................... 127 
REFERENCES .................................................................................................. 129 
APPENDICIES ................................................................................................. 148 
 
University of Ghana http://ugspace.ug.edu.gh
viii 
 
LIST OF TABLES 
Table 3. 1 Distribution of the Basin among the six riparian countries. ....................... 36 
Table 3. 2: Population of districts in the Lower Volta Basin....................................... 38 
Table 3. 3: Typology of vegetation of the Lower Volta River Basin .......................... 42 
Table 3. 4: Land-Use/Cover Classification Scheme and Description .......................... 44 
Table 4. 1: Accuracy results for maximum likelihood classification 2003 ................. 59 
Table 4. 2: Accuracy results for maximum likelihood classification 2015 ................. 60 
Table 4. 3: Land-Use/Cover Change Detection Statistics (%) .................................... 66 
Table 4. 4: Land-Use/Cover Change Detection Statistics (Area/ha) ........................... 67 
Table 4. 5: Area coverage of each land-use/cover type in the LVB ............................ 73 
Table 4. 6: Household size of respondents. ................................................................. 91 
Table 4. 7: Land tenure type of respondents ................................................................ 91 
Table 4. 8: Respondents farm sizes.............................................................................. 91 
Table 4. 9: Reasons given for fallowing farm land. ..................................................... 92 
Table 4. 10: Month for planting crops ......................................................................... 98 
Table 4. 11: Month in which rains begin (respondents view) ...................................... 99 
Table 4. 12: Advice given to respondents by agricultural extension officers ............ 101 
 
  
University of Ghana http://ugspace.ug.edu.gh
ix 
 
LIST OF FIGURES 
Figure 3. 1: Location Map of the Lower Volta Basin Showing Some Important Towns
...................................................................................................................................... 37 
Figure 3. 2 NDVI map of the two Landsat images ...................................................... 51 
Figure 4. 1: Selected ROI’s used for maximum likelihood supervised classification.
 57 
Figure 4. 2: Selected ROI’s used for maximum likelihood supervised classification . 58 
Figure 4. 3 Land-Use/Cover map of the LVB (2003 & 2015). Source: (Author’s 
Construct 2015). ........................................................................................................... 68 
Figure 4. 4: Land-use/cover change over the period of study (2003-2015) ................. 69 
Figure 4. 5: Land-Use/Cover change in Kpong area ................................................... 70 
Figure 4. 6: Land –Use/Cover Change detection (2003 and 2015) ............................. 71 
Figure 4. 7: Commercial irrigation farming area at Asutuare. ..................................... 72 
Figure 4. 8: Average annual rainfall totals, and trend in the Lower Volta Basin. ....... 77 
Figure 4. 9: Rainfall anomaly in the Lower Volta Basin ............................................. 77 
Figure 4. 10: Average temperature in the Lower Volta Basin ..................................... 79 
Figure 4. 11: Temperature anomaly in the Lower Volta Basin ................................... 79 
Figure 4. 12: Cassava Production in Ada East ............................................................. 80 
Figure 4. 13: Maize Production In Ada East................................................................ 81 
Figure 4. 14: Cassava Production in Lower Manya Krobo ......................................... 82 
Figure 4. 15: Maize Production in Lower Manya Krobo............................................. 83 
Figure 4. 16: Crop suitability map for maize crop, current climatic conditions (1950-
2010). ........................................................................................................................... 84 
Figure 4. 17: Crop suitability map for maize crop, future climatic conditions (~2050).
...................................................................................................................................... 85 
Figure 4. 18: Crop suitability map for pepper crop, current climatic conditions (1950-
2010). ........................................................................................................................... 86 
Figure 4. 19: Crop suitability map for pepper crop, future climatic conditions (~2050).
...................................................................................................................................... 86 
Figure 4. 20: Crop suitability map for cassava crop, current climatic conditions (1950-
2010). ........................................................................................................................... 87 
Figure 4. 21: Crop suitability map for cassava crop, future climatic conditions 
(~2050). ........................................................................................................................ 88 
Figure 4. 22: Distribution of respondents in the study area ......................................... 89 
Figure 4. 23: Age groups of respondents ..................................................................... 90 
Figure 4. 24: Educational level of respondents in the study area. ............................... 90 
Figure 4. 25: Number of years respondents have been farming. ................................. 92 
Figure 4. 26: Type of tree used as fire wood by respondents ...................................... 93 
Figure 4. 27: Land-use change as observed by respondents. ....................................... 95 
Figure 4. 28: Drivers of land-use change ..................................................................... 96 
Figure 4. 29: Knowledge level of respondents on climate change .............................. 97 
Figure 4. 30: Respondents view on effects of rainfall and temperature on their yield 
and soil condition ....................................................................................................... 100 
Figure 4. 31: Factors that influence farmer crop yield ............................................... 100 
Figure 4. 32: Farmer adaptation methods to changing rainfall and temperature ....... 101 
 
  
University of Ghana http://ugspace.ug.edu.gh
x 
 
LIST OF PLATES 
Plate 3.1: Feature verification using “kml” overlay in google earth……..…….……47 
Plate 3.2: Feature verification using spear tool in ENVI ………………….………..47 
Plate 3.3: False colour rendition image. Showing vegetation as bright red …….…..47 
Plate 4.1: Rice farms at Asutuare ……………………………………………….…..72 
Plate 4.2: Banana plantation (Golden exotics Co. Ltd) ………………….………….72 
Plate 4.3: Astruim imagery showing a section of irrigation scheme …………….….73 
Plate 4.4: Image showing some farms, reservoir and pump station ……………..….73 
Plate 4.5: Astrium image showing a section of Dordoekorfe irrigation scheme ……74 
Plate 4.6: Reservoir at Dordoekorfe ………………………….……………………...74 
Plate 4.7: Pump station at Dordoekorfe  ...……… ..…………………….…………..74 
Plate 4.8: Satelite image of Ada Foah area ………………………………………….75 
Plate 4.9: Irrigation farms at Ada Foah ……………………………………………...75 
Plate 4.10: Irrigation farm at Ada Foah ……………………………………………..75 
  
University of Ghana http://ugspace.ug.edu.gh
xi 
 
LIST OF ABBREVIATIONS AND ACRONYMS 
ABM LUCC – Agent-Based Model of Land-Use and Land Cover Changes 
AWF – Africa Water Facility 
CCSR – Canadian Centre for Remote Sensing 
CSIR – Council for Scientific and Industrial Research 
DN – Digital Number 
ENVI – Environment for Visualizing Images 
ETM – Enhanced Thematic Mapper 
FAO – Food and Agricultural Organization 
GCM – Global Climate Model 
GDP – Gross Domestic Product 
GIS – Geographic Information Systems 
GLSS – Ghana Living Standards Survey 
GMET – Ghana Meteorological Agency 
GPS – Global Positioning System 
HADCM3 – Hadley Centre Coupled Model version 3 
IGBP – International Geosphere biosphere Program 
IHDP – International Human Dimensions Programme 
IPCC – Intergovernmental Panel on Climate Change 
ISODATA – Iterative Self Organizing Data Analysis 
ITCZ – Inter Tropical Convergence Zone 
IWMI – International Water Management Institute 
LUC – Land-use change 
University of Ghana http://ugspace.ug.edu.gh
xii 
 
LULC – Land-use/land cover 
LULCC – Land-use and land cover changes 
LVB – Lower Volta Basin 
MoFA – Ministry of Food and Agriculture 
NAUTILUS - North East, Applications of Usable Technology in Land Planning for 
Urban Sprawl 
NDVI – Normalized Difference Vegetation Index 
NIR – Near Infra Red 
OLI – Operational Land Imager 
PCA – Principal Component Analysis 
RMS – Root Mean Square 
ROI – Region of Interest 
SHS – Senior High School 
SST – Sea Surface Temperature 
TIRS – Thermal Infrared Sensor 
UN – United Nations 
UNFCCC – United Nations Framework Convention on Climate Change 
UNEP GEF – United Nations Environmental Program, Global Environment Facility 
UTM – Universal Transverse Mercator 
VRA – Volta River Authority 
VRB – Volta River Basin 
WWAP – World Water Assessment Program 
 
University of Ghana http://ugspace.ug.edu.gh
1 
 
CHAPTER ONE 
INTRODUCTION 
1.1  Introduction 
Although dams have been built in the world since time immemorial, large dam 
construction was earlier not possible though needed, because of lack of adequate design 
knowledge, construction equipment, new materials like cement and concrete and 
technology of construction (Shultz, 2002). Moreover dams have solved many problems 
of communities served and have provided basis for economic development that has 
sustained itself. According to Gleick (2009), dams and reservoirs have long been used 
for various purposes which include, hydro power generation, irrigation, flood control 
and recreation.  
 
Although the societal benefits gained from dams are immense, there exists a risk, 
particularly downstream, that needs to be addressed for public safety and infrastructure 
resilience (Woldemichael et al., 2012). While some might argue that dam construction 
has reached the stage where the risk of structural failure is now almost non-existent, 
actual occurrences show that failures related to extreme hydrologic events (e.g., 
overtopping or unscheduled opening of spillways) still continue to occur (Saxena, 
2005). 
 
Research shows that there is also a potential impact of the reservoir on the local climate 
triggered by atmospheric feedback mechanisms that may physically modify the extreme 
hydro-climatology of the region (Woldemichael et al., 2012). In some studies, (Degu 
et al., 2011; Yusuf and Salami, 2009) it is indicated that the hydro meteorological 
variables like evaporation, precipitation and humidity are the first-order atmospheric 
University of Ghana http://ugspace.ug.edu.gh
2 
 
descriptors to show an increase in the post dam period while temperature and wind 
speed may also show a gradual decrease. Other changes occur in the post-dam era and 
are constituted as anthropogenic land-use and land cover changes (LULCC) around the 
dam. Factors responsible for the changes in the post dam era manifest themselves over 
a long period of time since anthropogenic (human-induced) alterations around dams, 
particularly of the land surface, take place continuously after the commissioning of the 
dam (Woldemichael et al., 2012). 
 
In Ghana the construction of the Akosombo dam in 1964 and Kpong dam 25km 
downstream in 1981 brought about hydro-electric power which has served as the 
backbone of Ghana’s industrialization and social advancement. Both dams are of 
significant importance as they provide hydro-electric power to most parts of Ghana as 
well as neighbouring countries such as Togo. Lake transport, tourism, irrigation and 
flood control are other benefits that came with the construction of the dams, but the 
problems associated to the dam construction both upstream and downstream include 
water borne diseases such as bilharzia and guinea worm, disrupted farming and fishing 
activities and infestation of water weeds which hinders lake transport and fishing 
(Geker, 1999). 
 
Apart from the construction of the Akosmbo and Kpong dams affecting land-use and 
land cover changes (LULCC), these benefits and challenges associated to the post dam 
era have also contributed further to land-use/cover change (LUCC) in the Lower Volta 
Basin (LVB). The drivers of LUCC are both human induced and naturally caused, 
especially by change in climate. Land-use and land cover change is inevitable with 
changing climate and population as well as man’s activities. According to Vitousek 
(1992), land-use/land cover change (LULCC) is a key driver of global change and has 
University of Ghana http://ugspace.ug.edu.gh
3 
 
significant implications for many international policy issues (Nunes and Auge, 1999). 
In recent years, LULCC has emerged as an important research question, because it has 
been identified as a key factor responsible for environmental modification worldwide 
(Xiao et al., 2006). To understand how LULCC affects and interacts with global earth 
systems, information is needed on what changes occur, where and when they occur, the 
rates at which they occur and the social and physical forces that drive the changes 
(Lambin, 1997).  
 
Amatekpor (1999), declared in his study that all categories of land-use occur in the 
Volta Basin, including rain fed and irrigated cropping, grazing, forestry, settlements, 
wildlife preservations, protected lands and many more. "But the current spatial 
distribution of these land-use types have not been mapped and published in a single 
document for easy reference and use by planners" (Amatekpor, 1999). However over 
the years some research studies including Koku (2002), Braimoh (2004), Codjoe 
(2004), Duadze (2004), Nyamekye (2010) and Akubia (2014), have been conducted on 
land cover and land-use types as well as its impact on the environment in the Volta 
Basin.  
 
This study forms an integral part of an ongoing Lower Volta project, “Reoptimization 
of Operations of Akosombo & Kpong Dams on the Volta River in Ghana to Restore 
Downstream Livelihoods and Ecosystems”. This project aims at restoring the natural 
flow of the Volta River, ecosystems and livelihoods which were changed after the 
construction of the Akosombo and Kpong dams and has also contributed to changes in 
land-use and land cover. This research study focuses on land-use/cover changes which 
influences and interacts with livelihoods and ecosystems. 
 
University of Ghana http://ugspace.ug.edu.gh
4 
 
Amatekpor (1999), points out that soil is the basis of all land-use planning. Moreover 
selecting a location for a specific land-use in an available space depend on the 
interpretation of the soil conditions as they relate to climatic and socio-economic 
conditions in the area. 
 
Stretching from latitude 5⁰ 30'N (Ghana) to 14⁰ 30'N (Mali) and covering six West 
African countries in total, is the Volta Basin (Barry et al., 2005). Two climatic zones 
are associated to the Volta River due to its length. The northern part having a single 
peak rainfall and the southern part having a bimodal rainfall pattern (Gordon, 1999). 
Over the past several years, the Volta River Authority (VRA) had focused more on 
tackling the problems upstream of the dam. It is in recent times that attention has been 
shifted to the Lower Volta Basin (LVB) which has experienced major ecological and 
land-use changes. 
 
The Lower Volta Basin which is defined as the area between the Kpong dam at Akuse 
and the estuary at Ada covering an area of about 68,600km2 (Barry et al., 2005), has 
undergone some changes because of the construction of the two dams (Akosombo and 
Kpong). There is a change in the normal flow regime of the river which has halted the 
periodic flooding of the Basin downstream. This has resulted in the change of the 
ecology and ecosystem services, thus having an impact on the livelihoods of people 
living here. Fishing and floodplain agriculture which were the main economic activities 
of the people living in communities downstream has due to the cessation of annual 
floods which deposited alluvial soils highly suitable for the cropping of pepper, cassava, 
okro and other crops collapsed. Also oyster, shrimps and tilapia which were in 
abundance before the construction of the dams are no longer available (Agbenyo, 
University of Ghana http://ugspace.ug.edu.gh
5 
 
2009). The characteristic in the flow regime created favourable conditions for plant 
growth. The plants hindered water transportation and created suitable habitat for other 
organisms which led to outbreak of diseases such as malaria and bilharzia.  
 
With growing concerns on ecosystems and livelihoods, learning projects and research 
studies have brought about the need to explore possible re optimization of dams. 
Moreover the negative effects downstream of the two dams which include the loss of 
the dynamic interactions between the river and its floodplains, wetlands, deltas, 
estuaries and mangroves could possibly be improved with the possibility of a re-
optimization project. Furthermore a probable re-optimization project looks at an 
enhancement of the benefits of dams which are hydropower generation, irrigation water 
supply, flood control and a restoration of ecosystems and livelihoods (Gregory, 2008). 
The Akosombo and Kpong dams when re optimized would create suitable conditions 
and minimize degree of the problems that were created downstream. Through a change 
detection analysis this study revealed how changed livelihood options as a result of the 
effects of the constructed dams is affecting land-use change. The impact of changing 
rainfall patterns on crop productivity was also evaluated.  
 
Pertinent literature indicates that the agricultural sector is the largest contributor to the 
country's Gross Domestic Product (GDP). The Ghana Statistical Service (GSS) in 2010 
reported that crops make up 66.2% of the agricultural subsector by GDP (GSS, 2010). 
According to Codjoe et al. (2005), rain-fed and some irrigated agriculture is the main 
economic activity of the majority of the population living in the Volta Basin region. 
This situation has however been reversed through rapid population growth and low 
economic standards of living with a lot of consequences for agricultural land resources 
in this region. An extensive engagement in crop farming coupled with variability in 
University of Ghana http://ugspace.ug.edu.gh
6 
 
rainfall and unreliable rainfall patterns in the Volta River Basin where poverty is 
predominant, has far reaching implications on the environment and food availability 
(Barry et al., 2005).  
 
Temperature and rainfall are known to be important climatic variables with 
considerable effect on crop yield. Titriku (1999) and MOFA (2011), reported that 
rainfall variability and unreliability is one important climatic variable that leads to 
moisture deficit that affects crop production.  
 
According to the FAO (2000), crops have different optimal growing conditions with 
regard to temperature and rainfall. Codjoe (2004), reported that precipitation in the 
Volta Basin is characterized by large variability as expressed in periodic droughts. 
These suggest without doubt that rain fed agriculture and crop yield will be affected, 
because of the large variability in precipitation patterns, the development and optimum 
use of surface water resources is key to agricultural crop production in the area.  
 
Moreover the interplay between land/atmosphere, energy, water, land-use and 
significant shifts in land-use patterns will cause spatio-temporal changes in weather 
patterns and rainfall characteristics. Concerns about land-use/cover change emerged in 
the research agenda on global environmental change several decades ago with the 
realization that land surface processes influence climate (Lambin et al., 2003). It came 
to light in the 1980s that the land-use/cover change affected the terrestrial ecosystems 
which served as sources and at the same time sinks of carbon, and up until today 
decreasing vegetation (forests) remains a serious challenge (Woodwell et al., 1983). 
Man and his activities are perceived to be the major drivers behind land-use/cover 
change and therefore have become essential in any land-use /cover change study 
University of Ghana http://ugspace.ug.edu.gh
7 
 
(Codjoe, 2004). Some of these drivers of land-use and cover change include increased 
population growth due to reduced infant mortality and migration (rural to urban or 
urban to rural), conversion of rural areas to urban areas, lack of value placed on 
ecological services, socio economic (jobs and poverty), and climate change/variability 
(Mallupatu and Reddy, 2013). Over the last few decades, numerous researches have 
improved measurements of land-cover change, the understanding of the causes of land-
use change, and predictive models of land-use/cover change. These researches have 
been largely promoted through the auspices of the Land-Use and Land-Cover Change 
(LUCC) project of the International Geosphere- Biosphere Programme (IGBP) and 
International Human Dimensions Programme on Global Environmental Change 
(IHDP). Remote sensing is essential in documenting and analysing actual changes in 
land-use/land cover (LULC) on regional and global scales (Lambin et al., 2003). 
Furthermore Geographic Information Systems (GIS) and Remote Sensing have proved 
to be useful tools for assessing the spatiotemporal dynamics of LULC (Hathout, 2002), 
and also assist decision makers to ensure sustainable development as well as to 
understand the dynamics of our changing environment (Iqbal and Khan, 2013). 
 
It is therefore pertinent to note that, even in the absence of population dynamics, and in 
an attempt to improve upon livelihoods in the Lower Volta Basin which have been 
affected over the years due to the dam construction, land-use/cover change and 
variability in rainfall will continue to have a dynamic interaction which will in turn 
influence livelihoods. This study assessed the relationship that exists between 
livelihoods and land-use/cover changes as well as rainfall variability effect on rain fed 
agriculture which influences livelihoods. LUCC mapping using remote sensing and 
GIS techniques was carried out by conducting change detection analysis. Furthermore 
University of Ghana http://ugspace.ug.edu.gh
8 
 
the socio economic surveys, analysed climate data and crop yield data and crop 
suitability modelling was performed. 
 
1.2  Problem Statement 
By eliminating the annual floods in the Lower Volta River floodplain and estuary, the 
dams have devastated the livelihoods of the downstream communities and the physical 
ecosystem processes on which they depend (AWF, 2010). It is estimated that about 
37% of freshwater-dependent species are on the verge of extinction as aquatic 
ecosystems worldwide and floodplain livelihoods in developing countries continue to 
be in a state of crisis. Furthermore recessional agriculture, estuarine fisheries and other 
land-uses have reduced and changed in the past several decades in Africa (Natural 
Heritage Institute, 2007). 
 
According to Gyasi (1999) "…..although the Akosombo and Kpong dams and their 
attendant artificial lakes have created significant economic opportunities, they also 
have created monumental national and international environmental, health and socio-
economic problems that warrant greater attention". 
 
The ecological functions and livelihoods of people of the Lower Volta Basin have 
changed since the operation of the Volta and Kpong dams. According to Barry et al., 
(2005), the establishment of the Akosombo dam has caused some of the soils in the 
Lower Volta Basin to become more acidic, and with a negative effect on yields from 
farms. Also human induced changes necessitated by the need to adjust to the change in 
ecosystem services especially in the provision of food resources have contributed to 
land cover change. As these land cover challenges have persisted over years there is no 
University of Ghana http://ugspace.ug.edu.gh
9 
 
doubt that changing rainfall patterns will also affect the currently existing agricultural 
land-use patterns and other land cover types downstream.  
 
According to Duadze (2004), land-use and land cover information is not available for 
many areas of the Lower Volta Basin which is undergoing rapid and wide-ranging 
changes being contributed largely by slash and burn farming. Pre-dam land-use/cover 
in the Volta Basin have changed considerably and there is thus the need to map detailed 
land-use of the Lower Volta Basin periodically to enhance and support planning and 
decision making. Rain fed agriculture is one of the common types of land-use in the 
Lower Volta Basin and is based on the land rotation system known also as bush fallow 
system (Titriku, 1999). Titriku further explains that rainfall variability and unreliability 
which leads to moisture deficits are a major hindrance to agricultural production. 
Further, Barry et al. (2005), reports that rainfall is seen to becoming more variable and 
unreliable. “Extensive crop farming coupled with variable and unreliable rainfall 
patterns in a region where poverty levels are high has far-reaching implications for the 
environment and food security” (UNEP GEF, 2010). 
 
Duadze (2004), further reported that many areas of the Volta Basin which are 
undergoing rapid and wide-range changes in vegetation and land-use due to the practice 
of shifting cultivation have no land-use and land cover information, suggesting the need 
to map land-use and perform change analysis more frequently due to the rapid changes 
that take place in the Lower Volta Basin. 
 
University of Ghana http://ugspace.ug.edu.gh
10 
 
1.3  Justification 
The damming of the Volta River has caused a number of negative effects in the Lower 
Volta Basin, leading to massive changes in land-use/cover downstream. A re-
optimization project could enhance the benefits of the dam which include increased 
hydro power output, irrigation water supply and flood control and a restoration of 
ecosystems and livelihoods. Mapping the land-use/cover changes that have occurred 
over the years just before the start of the re-optimization project is necessary because 
this is expected to inform policy makers and various stakeholders of how the 
implementation of the project has influenced land-use/ cover change and to what extent 
it can restore the terrestrial ecosystem. This study could help improve upon land-use 
planning and management which will also influence adaptation strategies to climate 
change (rainfall variability mainly). 
 
The inundation of the floodplain that could cause catastrophic damage or impinge upon 
settlements or specific land-use types to an unacceptable degree can and should be 
prevented. Moreover the pending change in the release of water from the Akosombo 
dam which will reintroduce annual controlled flow events into the floodplain is 
anticipated, as the re-optimization project seeks to identify scenarios that will increase 
water service benefit for all interests, including irrigation, fishery, flood control and 
hydro-electric power generation in the Lower Volta. Regional land-use patterns show 
the interaction between man and environment and the influence to mankind’s basic 
economic activities. Hence land-use and land cover mapping is of great significance in 
scientific research, planning and management.  
 
Furthermore changing climate might have possibly affected crop production in the 
Lower Volta Basin and thus affected livelihood. It is therefore necessary to investigate 
University of Ghana http://ugspace.ug.edu.gh
11 
 
to what extent rainfall pattern has affected crop production in the study area and what 
innovative measures could be put in place to improve crop production since the 
inundation of the flood plains have ceased as well. According to Gyasi et al. (2006), 
data for land-use planning seem inadequate and the increasing population as well as 
concomitant growing demands for land requires efficient use and management of land. 
This study therefore contributes to land-use and land cover data of the Lower Volta 
Basin needed for land-use/cover monitoring and effective land-use planning as well as 
management. 
 
1.4  Research Questions 
 What is the current land-use /cover pattern in the Lower Volta Basin? 
 What are the types, patterns and aerial extent of land-use change? 
 Which of the various land-use types in the Lower Volta Basin is predominant 
and why? 
 What have been the changes in climate (rainfall and temperature) of the Lower 
Volta Basin over the past three decades? 
 What is the trend of crop yield in the Lower Volta Basin? 
 Which areas of the Lower Volta Basin are suitable for crop production based on 
rainfall and temperature parameters?  
University of Ghana http://ugspace.ug.edu.gh
12 
 
1.5  OBJECTIVES 
1.5.1  Main Objective 
To study the relationship between changing land-use/cover using remote sensing and 
GIS techniques, changing rainfall patterns and agricultural crop production in the 
Lower Volta Basin. 
 
1.5.2  Specific Objectives 
1. To create a land-use/cover map of the Lower Volta Basin for the years 2003 and 
2015. 
2. To determine the types, pattern and magnitude of land-use/cover change. 
3. Analyse rainfall and temperature variability and changing crop yield in the 
Lower Volta Basin. 
4. Generate a crop suitability map for specific crops in the study area using 
different rainfall and temperature scenarios. 
 
  
University of Ghana http://ugspace.ug.edu.gh
13 
 
1.6  Conceptual Framework 
 
 
Dam creation, which is meant for hydro-electric power production mainly but comes 
with some other benefits and challenges upstream and downstream of the River Basin 
(Agbenyo, 2009). 
The dam creation has therefore affected the watershed ecology by altering the natural 
flow of the river. The natural annual flooding which occurs in the floodplain of the 
River Basin is affected and thus the creeks and wetlands which receive water annually 
become dry. Floodplain agriculture was also halted as the annual floods that replenished 
the floodplain with nutrients has ceased (Amatekpor, 1999).  
 
Dam Construction 
 
Changed watershed ecology 
 
Water for agricultural land use  
 
Climate change 
 
Land use/cover change 
 
Socio economic status 
University of Ghana http://ugspace.ug.edu.gh
14 
 
Water for agricultural land use has been affected because creeks and natural reservoirs 
have dried up and climate change has and continues to influence rainfall. 
 
The lack of access to adequate water for agriculture as well as changing climate has 
contributed to LUCC. Climate change affects water usage in five water demanding 
sectors namely; agriculture, industry, energy, municipal and reservoir (Yates and 
Strzepek, 1998). 
 
With the lack of adequate water for agricultural land use, the socio economic status of 
the people in the LVB especially farmers who constitute majority of the economically 
active population there are affected.  
 
  
University of Ghana http://ugspace.ug.edu.gh
15 
 
CHAPTER TWO 
LITERATURE REVIEW 
2.1  Volta Basin  
The Volta river Basin, located in West Africa covers an estimated area of 400,000 km2 
and stretches over six West African countries namely, Burkina Faso, Ghana, Togo, 
Benin, Cote d’Ivoire and Mali (Barry et al, 2005). In Ghana, the Volta River Basin 
covers nearly two-thirds (70%) of the land area and is divided into six sub-Basins 
namely, Black Volta, White Volta, Daka, Oti, Upper and Lower Volta. However studies 
conducted under the VBRP in 1997 suggested a revision of the geographic scope of the 
VRB to include the Tordzie/Aka Basin located south east of the Lower Volta in Ghana 
and Togo (UNEP-GEF, 2010). The basin therefore covers 8,500km2 in size with a 
volume of 150 billion m3 and an average depth of about 18.8m with the deepest part 
going as far as 90m and a shoreline of 5,500km (Barry et al., 2005). 
 
The Lower Volta Basin is the 80km stretch of river between the Kpong head pond and 
the estuary at Ada Foah (Gordon, 1999). The Lower Volta has a range of vegetation 
from closed forest to humid savannah, mangrove and coastal strand vegetation and 
thicket. Short grasses occur more within the Lower Volta that enhances the rearing of 
herbivores such as cattle, goat sheep among others. Short Grass Vegetation, refers to 
the type of vegetation in which short grasses abound or are relatively more with 
occasional trees. This type of vegetation occurs at the Lower Volta where the soil is 
generally suitable for rangeland agriculture and water resources abound. The soil types 
include the "Chichiwere" (Oxyaquic Quartzipsamment), "Amo" (Vertic Dystropept), 
"Tefle" (Alic Dystraquert) and "Avyime" (Rhodic Paleustalf), Amatekpor (1999). The 
land is normally low lying and in a few cases slightly hilly. These unique features are 
University of Ghana http://ugspace.ug.edu.gh
16 
 
conducive for grazing and watering of livestock, crop production, industrial 
development and community development. (UNEP-GEF, 2010). The Lower Volta 
Basin falls within the coastal savannah and transitional zone. 
 
2.2   Challenges associated to the Lower Volta Basin 
The construction of the Akosombo dam brought about changes which can be 
categorized or presented as opportunities and challenges to the environment and people 
of the Lower Volta Basin. The impact of the dam on the environment has been found 
to be varied. These include enhanced fishing upstream and diminished fishing 
downstream, opportunities for irrigated farming downstream, proliferation of aquatic 
weeds upstream and downstream, increase in some water borne diseases such as 
bilharzia and malaria and reduction and elimination of other diseases such as 
Onchocerciasis in some areas. Also included are seismicity, sediment load changes, 
sorphological changes and microclimatic changes (Amatekpor et al., 1999).  
Flooding 
According to Yeboah (1999), the Volta River had its annual cycle before the dams were 
constructed. Naturally the dry season was characterized by low flow rates (1,000 
cusecs) whereas the rainy season was characterized by high flow rates (125,000 cusecs), 
and this caused destruction of crops as well as property. He further reported that the 
annual floods did not only enrich the lands for cultivation but also the creeks, ponds 
and lagoons in these areas were filled with water for breeding fish which was exploited 
after floods had receded. Even though the construction of the dams completely cut out 
the annual flooding that occurred in the LVB, the opening of the floodgates of the 
Akosombo and Kpong dams occasionally causes flooding in certain areas of the Lower 
Volta Basin.  
University of Ghana http://ugspace.ug.edu.gh
17 
 
Flooding as observed in the Volta River Basin is blamed on extremely high rainfall 
rates and the creation of uncoordinated dams without appropriate management 
practices. Land-use conversions also exacerbate the problem as soils with significantly 
reduced vegetation cover that are exposed to atmospheric elements have little 
infiltration capacities to reduce storm water run-off (GEF-UNEP, 2002).  
Sand bars formation 
Before the dam construction, seasonal floods flushed out sandbars, which might have 
started during the dry season, thus the estuary was largely kept free of sandbars. 
However, after the construction of the dam, due to absence of annual floods, sandbars 
gradually formed at the estuary and in time virtually blocked it. The effect of this was 
that saline water, which during high tide flowed upstream into the river channel, 
completely ceased. Moreover, the absence of salt water in the river channel, made the 
areas even closer to the estuary prone to the presence of the schistosome snail and 
Bilharzia became prevalent (Yeboah, 1999). The VRA restored salinity to the river 
channel by placing a dredger permanently at the estuary to cut a channel through the 
sandbar. As a result, Bilharzia has been wiped out of these areas (Boubacar, 2005). 
 
Diseases 
All of the water-related diseases occurring in Ghana can be found in the Volta Basin. 
These include Bilharzia, which is very widespread in all four sub-Basins, except the 
Daka, onchocerciasis (except in the Lower Volta), Guinea worm, malaria, lymphatic 
filariasis, which is particularly common in the White Volta Basin, and diarrhoea (GEF-
UNEP, 2002). 
 
University of Ghana http://ugspace.ug.edu.gh
18 
 
Poverty 
In Lemoalle and Condappa (2010), it was reported that people are poor because of low 
agricultural productivity, limited access to markets, unstable prices, and insecure land 
tenure. Asante (2007) said, the three main causes of water related poverty in Ghana are 
low productivity of fishing, and agriculture, water insecurity (rainfall variability, poor 
access, health impacts, loss of labour) and water-related diseases (malaria, guinea worm 
and others). With regards to agriculture (Lachaud, 1998) the incidence of poverty was 
lower among cash crop farmers than among subsistence farmers. 
Land degradation and Deforestation 
The environmental problems arising from livestock production are becoming sources 
of great concern. The maintenance of large herds of livestock has tended to exceed the 
carrying capacity of the ecosystem (Barry et al., 2005). Barry explains further, that as 
a result of trampling, soils are severely exposed to erosion. The watersheds are also 
exposed to rapid evaporation. The prolonged exposure of the soil renders it susceptible 
to erosion and reduces its regenerative capacity. In the northern parts of the Basin, large 
tracts of arable land have become infertile and crop yields have declined tremendously. 
Livestock is also taking a significant toll on soil productivity in the LVB region. 
Although there is little data on the specific effects of livestock, it is clear that livestock 
negatively influences the area (Barry et al., 2005). 
 
One widely held view is that dryland degradation or desertification is due to population 
growth and poverty which contribute to increased pressure on natural resources through 
overgrazing, over-cultivation, and over-harvesting of woodlands. These activities, in 
turn, lead to deforestation, soil erosion and poor land management which result in 
further environmental degradation and desertification (Mortimore, 1998).  
University of Ghana http://ugspace.ug.edu.gh
19 
 
An estimated 2 million people in Ghana depend on forests for subsistence uses and 
traditional and customary lifestyles. Forest‐adjacent communities undertake a wide 
range of forest‐related activities, including fuelwood and charcoal production, wood‐
carving, canoe carving, rattan production and chew stick‐gathering (Tufouor, 2012). 
 
2.3  Definition of Terms 
Land cover is defined by the attributes of the earth’s land surface and immediate sub 
surface, including biota, soil topography, surface and ground water and human 
structures as in the case for agricultural expansion or deforestation (Lambin, et al., 
2003). 
 
Turner et al. (1995), defined land cover as “the biophysical state of Earth’s surface and 
immediate subsurface.” He also defined land-use/cover change as “the sum of short and 
long term effects of global change and human transformation of the landscape”. 
Moreover according to the FAO (1995), land use can be described as the purpose for 
which land is used by a population. 
 
Land-cover conversions (i.e., the complete replacement of one cover type by another) 
are measured by a shift from one land-cover category to another, as is the case in 
agricultural expansion, deforestation, or change in urban extent (Gonzales, 2009). 
Land-cover modifications are more subtle changes that affect the character of the land 
cover without changing its overall classification (Lambin et al., 2003). 
 
GIS technology is a computer-based data collection, storage, and analysis tool that 
combines previously unrelated information into easily understood maps. GIS software 
University of Ghana http://ugspace.ug.edu.gh
20 
 
can perform complicated analytical functions and then present the results visually as 
maps, tables or graphs. This allows decision-makers to virtually see the issues before 
them and then select the best course of action (Parthasarathy et al., 2008). 
 
According to the Canadian Centre for Remote Sensing (CCSR, 1997), remote sensing 
is the science of acquiring information about the earth’s surface without coming into 
contact with it. This is however done by, sensing and recording reflected or emitted 
energy and processing, analysing and applying the information obtained (Akubia, 
2014). 
 
2.4  Understanding Land-use Change (Theories of Land-Use/Cover Change)  
Land-use change is a complex process that encounters sophisticated parameters 
Tayyebi et al. (2008) and Lambin et al. (2003), claims LULCC can be classified as 
either a progressive or episodic change. It is revealed through time series remote 
sensing data that land-cover changes may show periods of rapid abrupt change followed 
by a quick recovery of ecosystems rather than the known gradual change. Furthermore 
such changes caused by the interaction of climatic and land-use factors have an 
important impact on ecosystem processes (Lambin, et al., 2003).  
 
Lambin et al., (1997) stated that, to understand how LULCC affects and interacts with 
global earth systems, information is needed on what changes occur, where and when 
they occur, the rates at which they occur and the social and physical forces that drive 
the changes. Research on land-use/cover change requires a combination of perspectives 
of understanding. These perspectives include the agent based, systems, and the 
narrative approaches (Lambin et al., 1999).  
University of Ghana http://ugspace.ug.edu.gh
21 
 
The agent based perspective is centred on the general nature and rules of land-use 
decision making by individuals (Lambin et al., 2003). These individuals who are agents 
of land-use/cover change have intentional and unintentional motives which are related 
to economic, traditional, emotional or biophysical factors (Leemans et al., 2003). 
Parker et al. (2003), explains how economic models of land-use/cover change assume 
that land managers attempt to fulfil their needs and meet their expectations.  
 
The systems perspective explains land-use/cover change through organizations and 
institutions of society (Ostrom, 1990). This perspective represents the dynamics of 
economy and environmental linkages which are operating at regional to global scales, 
and although some institutions are direct drivers of change, others, such as markets, are 
intricately linked to individual decisions (Lambin et al., 2003) 
 
Finally the narrative perspective provides an in-depth of understanding land-use/cover 
change through historical detail and interpretation (Crumley, 1994). It is important as 
well and might even be the most important as the historical analysis of landscape grasps 
all the complexity of events that affect LUCC (Klepeis and Turner, 2001). The narrative 
perspective is therefore able to avoid the simplifications and inaccurate interpretations 
that could result from a study focused on the present and immediate past (Batterbury, 
1999). An integration of these perspectives as stated by Lambin (2003), will not only 
help to understand the dynamics of LUCC but will yield unequivocal results in LUCC 
studies. 
 
According to Lambin et al. (2003), a major trend of global LUCC is the expansion of 
agricultural land. Currently, agricultural land covers about a third of the global land 
surface, and has expanded into forests to meet the growing demand for food. It is 
University of Ghana http://ugspace.ug.edu.gh
22 
 
important to know that climate-driven land-cover modifications interact with land-use 
changes. Land-use change is driven by synergetic factor combinations of resource 
scarcity leading to an increase in the pressure of production on resources, changing 
opportunities created by markets, outside policy intervention, loss of adaptive capacity, 
and changes in social organization and attitudes. Most importantly the changes in 
ecosystem goods and services that result from land-use change feedback on the drivers 
of land-use change (Lambin et al., 2003). Furthermore understanding of the causes of 
land-use change has moved from simplistic representations of two or three driving 
forces to a much more profound understanding that involves situation-specific 
interactions among a large number of factors at different spatial and temporal scales. 
 
Other forms of rapid land-cover change that are thought to be widespread are still 
poorly documented at the global scale. Local- to national-scale studies, however, 
demonstrate their importance and ecological significance (Lambin et al., 2003).  
In a study by Amanor and Pabi (2006), It was concluded that a collaboration between 
anthropological research within a regional economies perspective and remote sensing 
can be useful in providing framework for analysing the dynamics of localized patterns 
of change. 
 
Remote sensing data highlight high temporal frequency land-cover modifications of 
great importance for earth system processes. In particular, data from wide-field of-view 
satellite sensors reveal patterns of seasonal and inter-annual variations in land surface 
attributes that are driven not by human induced change but rather by climatic variability 
(Eastman, 1993). 
 
University of Ghana http://ugspace.ug.edu.gh
23 
 
As reported in d’Aquino et al. (2002), researches and experts in LUCC are developing 
a model “Agent-Based Model of Land-use and land cover changes (ABM LUCC) 
which will strongly compliment the already existing techniques of LUCC modelling. 
In this report, Parker and Berger discuss how ABM/LUCC can be used both as a means 
of implementing existing knowledge, thereby developing tools for scenario analysis, 
and for knowledge discovery, as integrated models are used to developed the process 
theory needed by the LUCC community. Finally, the editors discuss the potential for 
ABM/LUCC to integrate not only models, but researchers, across disciplines Berger et 
al. (2002). 
 
2.5  Land-use and cover change in Ghana and the Volta river Basin 
According to Agyarko (2001), Ghana’s landscape has been categorized into land-use 
types including, small and large scale farming, forestry, grazing fields, urban areas, tree 
plantations of indigenous and exotic species and game/park reserves. To a very large 
extent the vegetative cover of an area is determined by climatic and edaphic factors. It 
is documented and also observable that the land cover of the Volta River Basin consists 
predominantly of sub-humid savannah; mainly tall grasses interspersed with fire 
resistant tress in the Black and White Volta Sub-Basins with some forest reserves found 
in the white Volta area. (GEF-UNEP, 2002). 
 
With the latest developments in remote sensing and GIS technology, Ghana should now 
embark on mapping a detailed land-use map of the Volta Basin (Amatekpor, 1999). 
Amatekpor claims the use of Landsat TM or preferably SPOT-XS image is ideal for the 
production of the detailed land-use/cover map. Since the launch of the first Earth 
Resource Technology Satellite on July 23, 1972, the analysis of data has advanced from 
University of Ghana http://ugspace.ug.edu.gh
24 
 
simple visual observation to sophisticated interpretations based on first principles of 
spectroscopy and electromagnetic radiation (Ustin, 2004). Moreover, in present times 
the development of high resolution sensors such as Quick bird and Spot 5, provides 
high spatial resolution imagery which gives more accurate information about the earth’s 
surface (Gumma et al., 2010). Land-use/cover information is important because, it 
constitutes key environmental information required for many scientific, resource 
management and policy purposes, as well as for a range of human activities. This 
information is however currently not available for the Volta Basin even though it is 
undergoing rapid and wide-ranging changes in land-use and vegetation due to 
agricultural practices (Duadze, 2004).  
 
It is very important to have continual, historical and precise information on LULC 
changes of the earth’s surface for any kind of sustainable development program in 
which LULC serves as one of the major input criteria (Mei and Qing, 1999; El-Kawy 
et al., 2010). 
 
2.6  Drivers of Land-Use/Cover Changes 
Land-use change is always caused by a multitude of interacting factors originating from 
different levels of organization of the coupled human-environment system (Lambin et 
al., 2003). Land cover may change under the influence of biophysical conditions only 
but, most frequently; it results from human-induced land-use change (Briassoulis, 
2000) At the local level, causes of land-use/cover changes involve a physical action on 
land cover such as agriculture, forestry and infrastructure construction (Lambin and 
Geist, 2006). Driving forces of land-use and land-cover change are not only highly 
interrelated, but also can vary both in time and space, whereby the strength of their 
University of Ghana http://ugspace.ug.edu.gh
25 
 
interrelations is also temporally and spatially variable (Schindler 2009). However, most 
importantly land-use dynamics, which involve decisions of land-users, are major 
determinants of land cover changes. Thus, the critical element in land-use is the human 
agent, who takes specific actions to his own calculus or decision rules that drive land-
cover change (Lambin et al., 1999).  
 
There are proximate/direct causes of land-use change which constitute human activities 
or immediate actions that originate from intended land-use, which directly affect land 
cover (Ojima et al., 1994). There are also the underlying (or indirect or root) causes 
which are fundamental forces that underpin the more proximate causes of land-cover 
change (Leemans et al., 2003). The underlying causes are formed by a complex of 
social, political, economic, demographic, technological, cultural, and biophysical 
variables that constitute initial conditions in the human environment (Geist and Lambin, 
2002). 
 
Interestingly, Lambin et al. (2003), established that the five fundamental high level 
causes of land-use change are;  
i. resource scarcity leading to an increase in the pressure of production on 
resources, 
ii. changing opportunities created by markets, 
iii. outside policy intervention, 
iv. loss of adaptive capacity and increased vulnerability, and  
v. changes in social organization, in resource access and in attitudes. 
 
However all known drivers of LUCC which are (multiple causes, economic and 
technological factors, demographic, institutional, cultural and globalization), are all 
University of Ghana http://ugspace.ug.edu.gh
26 
 
strongly linked within and between levels of organization of human-environment 
systems. This therefore places humans as the main drivers of land-use change. It is the 
agent (an individual, household, or institution) that takes specific actions according to 
its own decision rules which drive land-cover change (Berger et al., 2002). Lambin et 
al. (2003), further conclude that, the various sectoral drivers of land-use change interact 
directly, are linked via feedback, and thus often have synergetic effects. 
 
2.7  Image classification 
An approach often used to derive main land-cover types is the analysis of satellite 
images via remote sensing using automatic classification methods. Such automatic 
classification methods extract land-cover types based on spectral information of the 
satellite image. But since some land-cover types display similar spectral properties, the 
accuracy of such automatic classification algorithms is often limited. These automated 
algorithms are therefore usually used in association with other information sources so 
as to interpret the automatically derived land-cover classes (Schindler, 2009). These 
other sources include, aerial photographs, a high-resolution satellite image, or ground-
truth data.  
 
Presently there are different image classification procedures used for different purposes 
by various researchers (Butera, 1983; Lo and Watson, 1998; Ozesmi and Bauer, 2002; 
Liu et al., 2002; Dean and Smith, 2003; and Pal and Mather, 2003). The latest high 
resolution sensors and others which are yet to come, together with new existing data 
process environments have led to important changes in classification methodologies 
(Manakos et al., 2000). According to Gumma et al. (2010), these techniques are 
distinguished in two main ways as unsupervised and supervised classifications. 
University of Ghana http://ugspace.ug.edu.gh
27 
 
Moreover, the supervised classification has different classification methods which are 
named as parallelpiped, maximum likelihood, minimum distance and fisher classifier 
methods. These are termed hard classifier.  
 
The unsupervised classification also has two techniques namely K-Means and 
ISODATA classifier (Guide EUS, 2008). However both classification approaches have 
pros and cons, however supervised classification is likely to provide more accurate 
classification result than unsupervised classification (Peter and Michael 2003; Butt et 
al., 2012). A Kappa coefficient is usually calculated after a classification process has 
been run on an image. Duadze (2004), recommends that an unmodified kappa index is 
utilized to report classification accuracy, primarily because it is documented and well 
known. Furthermore Duadze explains that errors of omission calculate the probability 
that a reference sample has been classified correctly whilst commission calculate the 
probability that a sample from the classified data actually represents that category on 
the ground. With the use of NDVI also aiding in image classification, (Tucker, 1979; 
Dawelbait and Francesco, 2011), acknowledged that NDVI has been mostly commonly 
used to map spatial and temporal variation in vegetation. However, it is limited in 
providing accurate estimates of shrub land cover in arid areas and limited utility in an 
arid ecosystem (Huette, 1992 and Asner, 2004). 
 
2.8  Change detection 
There are many techniques developed to detect changes in the environment (Wong et 
al., 1997). But there is still no standard method for the quantitative estimation of land 
cover. Recent efforts have also been made and addressed improved methods of LULC 
change detection (Hurd et al., 2001).  
University of Ghana http://ugspace.ug.edu.gh
28 
 
 
All the existing methods are based on different approaches which are suitable for 
specific locations or areas and satellite sensors (Purevdor et al., 1998). Several 
techniques including, multi date classification, post classification change detection 
(Hurd et al., 1992), cross correlational analysis (Hurd et al. 2001), multi date principal 
component analysis and RGB-NDVI colour composite change detection (Hoffhine, 
2000) have been evaluated by North East, Applications of Usable Technology in Land 
Planning for Urban Sprawl (NAUTILUS) investigators (Civco et al., 2002) .According 
to Sunnar (1998), the most appropriate change detection procedure for a given situation 
has significant impact on the results. Moreover this depends on the type of application, 
environment, targets of interest and the amount of detail required. Several changed 
detection methods have been developed (El-Raey et al. 1999), however the most 
commonly used change detection techniques include, image overlay, classification 
comparison, image differencing, principal component analysis (PCA), image rationing, 
differencing of NDVI images, and change vector analysis (Sunnar, 1998). 
 
With land cover change analysis, the most used approach is image differencing, which 
involves the subtraction of image brightness on a pixel-by-pixel basis (Duadze, 2004). 
The image differencing procedure involves the co-registration of two images of the 
same area taken at different times to assess the degree of change that has taken place 
between the dates of imaging and preparing a temporal difference image by subtracting 
the digital numbers (DN) for one date from those of the other date(s) (Lillesand and 
Kiefer, 1994). In this study, the classification comparison technique was used. This 
method makes use of two or more independent classifications of each satellite scene. In 
this method all classified satellite are classified into a single dataset and it identifies 
classes of pixels that have changed between dates (Richards, 1995). 
University of Ghana http://ugspace.ug.edu.gh
29 
 
In recent years, the developed satellite remote sensing techniques prove to be very 
useful for preparing accurate land-use/cover maps and monitoring changes at regular 
intervals of time and space (Zubair, 2006). 
 
2.9  Agriculture in the Ghana and the Volta River Basin 
Agriculture, including forestry, is the backbone of the Ghanaian economy. It provides 
43% of the Gross Domestic Product, 50% of export earnings and 70% of total 
employment (Agyarko, 2010). Agriculture in Ghana is also predominantly on a 
smallholder basis, with about 90% of farm holdings being less than 2 hectares although 
there are some large farms and plantations, particularly for rubber, oil palm and 
coconut, and to a lesser extent, rice, maize, and pineapples. With the main system of 
farming being traditional, the hoe and cutlass are usually the main farming tools used 
and production varies with the amount and distribution of rainfall (MOFA-SRID, 
2011). Published statistics of the MoFA reveal a vast amount of land that is potentially 
available for agricultural use. MoFA reports that of the total land area of 23.9 million 
hectares, 13.6 million hectares (57.1%) can be classified as agricultural land area but 
only 5.3 million (22.2% of all agricultural land area) were under cultivation at the last 
count (1995). With an economically active population between the ages 15 to 49, there 
are “424,458”, “531,635” and 145,034 people engaged in farming in the Volta, Eastern 
and Greater Accra regions respectively. The estimated annual population growth rate 
of Volta, Eastern and Greater Accra regions are 2.4, 2.0 and 2.8 respectively (MOFA-
SRID, 2011). The soils of the Lower Volta flood plain which covers an area of 294.0sq 
km (29,280ha) is good for agriculture (Gordon and Amatekpor, 1999). 
 
University of Ghana http://ugspace.ug.edu.gh
30 
 
According to Fisher et al. (2002), suitable land and potential production for staple cereal 
crops will decrease markedly in sub-Saharan Africa. The paradox is that areas that are 
currently most food insecure will be most affected by climate change (Lane and Jarvis, 
2007). However it should be asked if the area which falls within the Volta Basin of 
Ghana should be considered prone to low productivity due to changing climate. There 
is no doubt that with the available water provided by the lake rainfall should not hinder 
production even if temperature will increase and humidity will decrease. Fobil and 
Attaquayefio (2003), reported that “following construction of the Akosombo dam, there 
has been a steady decline in agricultural productivity along the lake and its tributaries, 
where inadequate soils have been exhausted.” The IWMI (2013), reported climate 
change will have an intense impact on water resources that could have serious effects 
for agriculture. A key message from the (WWDR3; WWAP, 2009a) states that, 
Adapting to climate change is a critical challenge, particularly for developing countries, 
whose capacity to adapt is low. For some, the incremental costs of climate change 
adaptation will soon approach the current value of aid inflows. Cassava and cocoa are 
the two most important crops in the coastal zone though their value are not as much as 
compared to those of the forest and savannah zones. Maize and yam account for more 
than half of the total crop harvest value in the savannah zone however maize is the least 
crop harvested in the coastal zone as compared to the savannah and forest ecological 
zones (GLSS, 2010). Cassava is reported a very significant crop in the coastal zone and 
has a higher capacity to withstand drought conditions. 
 
The Ghana Statistical Service of Ghana reports livestock rearing is concentrated in the 
rural savannah areas, where 86 percent of drought animals, 63 percent of cattle and 80 
percent of guinea fowls are being reared (GLSS, 2010).  
 
University of Ghana http://ugspace.ug.edu.gh
31 
 
2.10  Climate change and rainfall variability 
According to the IPCC (2007), climate change is defined as a change in state of climate 
that can be identified by changes in the mean (using statistical tests) of its properties 
and persists for an extended period, decades or longer. This therefore refers to changes 
in climatic elements over time, be it due to natural variability or human induced. The 
IPCC further reports, average global temperature has increased by 0.74 Celsius in the 
last 100 years; rainfall has trended downward during 1960–2000; and sea levels have 
risen between 1 to 3 millimetres per year. 
 
Rainfall variability is an inherent part of African climate and is deeply entrenched in 
West African society (FAO 2008). Moreover the impact of rising temperatures on 
rainfall distribution patterns in Africa remains far less certain (IPCC, 2007). 
Over the past 50 years, declining rainfalls have been reported throughout West Africa 
as part of a general southward shift in seasonal migration of the Inter-tropical 
Convergence Zone (ITCZ) (Weldeab et al., 2007)  
 
Sea surface temperatures (SST) really influence climate change in Ghana. Studies 
conducted by scholars including Gu and Adler, (2003) show that, a high south Atlantic 
SST is associated with high rainfall in the Guinea coast (south of 8○N) and vice versa. 
In a national study conducted by Opoku-Ankomah and Cordery (1994) suggests that 
variability in the 
 
Northern zone is distinct from the remainder of the country, due to the movement of 
the ITCZ and influence of Atlantic SSTs, as reported by Boateng (1967), as cited in 
Owusu (2009). 
University of Ghana http://ugspace.ug.edu.gh
32 
 
Climate models show that the probability of extremes of rainfall is likely to increase, 
resulting in more floods and droughts in regions already affected by these kind of events 
– often these are regions with low income levels per capita, widespread poverty, high 
population growth and rapid urbanization (WWAP, 2011). 
 
Water-related climate change impacts are already being experienced in some river 
basins in the form of more severe and more frequent droughts and floods. Higher 
average temperatures and changes in precipitation and other climatic variables are 
projected to affect the availability of water resources through changes in rainfall 
distributions (UN Water, 2009) 
 
2.11  Exposure and sensitivity of crops to changing climate 
Many studies have confirmed that temperatures will rise and rainfall will increase in 
some places, while in others, rainfall will decrease (Rosenzweig et al., 1993). However, 
there is general agreement among scientists that food crops are sensitive to the changing 
climate (Yumbya et al., 2014). 
 
According to Rosenzweig et al. (2001), altered and unpredictable weather patterns can 
increase crop vulnerability to pests, diseases and the effects of extreme climate events 
such as high temperatures, droughts and torrential rains. Sequential extremes, such as 
droughts followed by intense flooding rains, are catastrophic in themselves and are 
compounded by ecological effects such as the expansion of the ranges of pathogens, 
diseases, and pests that affect human populations and agricultural production. However 
the impact of climate change on production of various crops varies markedly depending 
mainly on the region, growing season, the crops and their temperature thresholds. Reilly 
et al. (1996), found that as temperatures move away from the favourable growing 
University of Ghana http://ugspace.ug.edu.gh
33 
 
temperature of crops, the growth of the crop is adversely affected. Similarly, if 
variability in temperatures is high, crop yields are lower. The authors therefore 
concluded that places that are too hot or are too close to the optimum temperature are 
likely to suffer the most in terms of crop production. 
 
Porter et al. (2007) predict that if global temperatures do not increase more than 4◦C 
over the next century, arable agricultural production can probably adapt to changes in 
mean global temperature using breeding, selection and management Anecdotal 
evidence from farmers suggests that the onset of rainy season has been shifting forward 
in time over the past two generations. Recently, detailed atmospheric modelling over 
the region shows that in the near future too, the onset of rainy season will shift to later 
periods in the year, roughly from April towards May. The end of rainy season as well 
as the total amount of rainfall will remain more or less fixed (Van de Giesen et al., 
2010). 
 
Climate change and its impact on agriculture may contribute to preventing strategic 
economic sectors and may further hinder the achievement of the Millennium 
Development Goal (MDG) 1 on poverty alleviation (Osman-Elasha, 2010)  
 
2.12  Crop suitability mapping 
The value of Geographic Information Systems (GIS) for assessing climate change 
impacts on crop productivity cannot be over-emphasised (Yumbya et al., 2014). 
 
Agriculture is a climate-dependent activity and hence is highly sensitive to climatic 
changes and climate variability (Ramirez-Villegas et al., 2013). Several authors report 
University of Ghana http://ugspace.ug.edu.gh
34 
 
that agricultural production could suffer progressive yield loses 17 in the next hundred 
years (Challinor et al., 2009; Challinor et al., 2010; IPCC, 2007; Lobell et al., 2008; 
Thornton et al., 2011). Acknowledging the gap of non-availability of production of 
minor crops data and models, Ramirez-Villegas et al. (2013), integrated the current 
expert knowledge reported in the FAO-EcoCrop database, with the basic mechanistic 
model (also named EcoCrop), originally developed by Hijmans et al. (2001).  
 
The Diva GIS program houses this model and is used to generate crop suitability maps 
based on the evaluation of the likely impacts of climate change on agricultural 
production. Ramirez-Villegas et al., (2013) furthermore highlighted the considerable 
opportunity of using EcoCrop to assess global food security issues, broad climatic 
constraints and regional crop-suitability shifts in the context of climate change. The 
model was named EcoCrop since it was based on the FAO-EcoCrop database (FAO, 
2000). The modeling of the crop’s suitability is a process that involves the evaluation 
of the model and the usage of the selected parameter set(s) to run the model using a 
certain (set of) climate scenario(s) (Ramirez-Villegas et al., 2013). Furthermore, in the 
model, there are two ecological ranges for a given crop. Each defined by a pair of 
parameters for each variable (i.e. temperature and rainfall). There is the absolute range 
(minimum and maximum absolute temperature and rainfall), optimum range (minimum 
and maximum optimum temperatures and rainfall respectively) and an additional 
temperature parameter (TKILL) used to illustrate the effect of a month’s minimum 
temperature (Ramirez-Villegas et al., 2013). 
 
In many studies (Yumbya et al., 2014; Lane and Jarvis, 2007; Geerts et al., 2006), 
scientist have used the EcoCrop model to predict suitable areas of crop production 
under current and future climatic conditions. Other methods using other GIS platforms 
University of Ghana http://ugspace.ug.edu.gh
35 
 
such as ArcView and ArcGIS are also used to model suitable geographical areas for 
crop production (Geerts et al., 2006). 
 
In Parthasarathy et al., (2008) the thirty years area and production curves of ginger 
growing states were compared with the Eco-crop suitability model which indicated that 
suitability has direct impact on production.  
 
 
  
University of Ghana http://ugspace.ug.edu.gh
36 
 
CHAPTER THREE 
MATERIALS AND METHODS 
A background and general information about the subject and study area which guided 
the methodology of this research is provided. Pertinent literature has identified some 
major occurrences as drivers of land cover and land-use changes in the Volta Basin. 
Thus the need to study these drivers and their effects in relation to the environment and 
peoples livelihoods. Furthermore the need to look at the relationship between changing 
climate (rainfall and temperature) and farming, a major livelihood option of the people 
in the Lower Volta Basin is relevant to this study. 
3.1  Study Area 
The Lower Volta Basin is a sub Basin of the Volta River Basin which covers six West 
African countries with an estimated area of about 400,000km2 and considered the 9th 
largest in sub-Saharan Africa (UNEP GEF, 2002). The widest stretch of the Volta Basin 
is longitude 5⁰W to 2⁰E along latitude 11⁰ N, it becomes narrower towards the coast of 
the Gulf of Guinea. In Ghana the Volta Basin is about 167,298 sq km representing about 
70% of the total land area of Ghana and 40% of the entire Volta Basin as seen in table 
3.1. 
Table 3. 1 Distribution of the Basin among the six riparian countries. 
Country Area of Volta River Basin 
(km2) 
% Basin % Country In 
Basin 
Benin 17,098 4.10 15.2 
Burkina Faso 178,000 42.65 63.0 
Côte d’Ivoire 12,500 2.99 3.9 
Mali 15,392 3.69 1.2 
Togo 26,700 6.40 47.3 
Ghana 167,692 40.18 70.0 
Total 417,382 100%  
Source: UNEP GEF, 2002 
University of Ghana http://ugspace.ug.edu.gh
37 
 
 
Figure 3. 1: Location Map of the Lower Volta Basin Showing Some Important Towns 
University of Ghana http://ugspace.ug.edu.gh
38 
 
3.1.1 Population 
The Lower Volta Basin traverses six districts of Ghana, including, Lower Manya 
Krobo, Shai Osudoku, North Tongu, Central Tongu South Tongu and Ada East. These 
districts fall within three regions of the country, namely; Greater Accra, Eastern and 
Volta region. 
 
The Volta Basin has an estimated population of about 8.85 million (2010) people and 
supplies close to eighty percent of the staple food requirement of the country. The 
population growth rate for these regions, Greater Accra, Eastern and Volta for the 
period, 2000 to 2010 are 3.1%, 2.1% and 2.5% respectively (UNEP GEF, 2010). 
Using the population of the various districts that fall within the Lower Volta Basin area, 
the total population for the study area is computed in the Table 3.2.  
Table 3. 2: Population of districts in the Lower Volta Basin 
Region District Population 
Greater Accra Ada East  130,795 
Eastern  Shai Osudoku 122,836 
Volta  North Tongu 149,188 
Volta Central 59,411 
Volta South Tongu 87,950 
Eastern Lower Manya Krobo 89, 246 
Total 639,426 
Source: GSS, 2010, population and Housing Census.  
3.1.2  Location and Administrative Area of LVB 
The six districts that are captured in the extent of the Lower Volta Basin as defined by 
the coordinates (0○45”E to 1○00”E and 6○15”N to 5○45”N) of the Volta Basin Research 
Project (VBRP). 
The Ada East district is located in the Eastern part of the Greater Accra region and 
forms part of the Accra plains.  
University of Ghana http://ugspace.ug.edu.gh
39 
 
North Tongu district lies within the Tropical Savannah Grassland, with a large tract of 
grassland which is suitable for large scale livestock rearing. The Volta River runs from 
the north to the south of the district roughly dividing it into two equal parts (MOFA, 
2015) 
South Tongu district, was carved out of the Tongu district, the district is drained 
mainly by the Volta River, which runs along the district’s western boundary with 
Dangme East District. 
The Central Tongu district forms part of the (25) MMDAs in the Volta Region of 
Ghana. The district is drained by the Alabo, Kolo, Aklakpa, Gblor, and Nyifla streams 
and their numerous tributaries into the Volta River, which runs North – South through 
the District. 
The Shai Osudoku District formerly Dangme West. The general pattern of drainage in 
the district is dendritic with most of the streams taking their source from the Akwapim 
range which is also serves as a watershed, and flows from northwest to southwest into 
lagoons on the coast.  
The Lower Manya Krobo district is relatively flat to the southeast with isolated hills to 
the northwest. Landscape is generally undulating with several streams, most of which 
drain into the Volta River.  
 
3.1.3  Climatic Conditions 
The climate of the Volta Basin is controlled by two air masses namely, the North-East 
Trade Winds and the South-West Trade Winds, which meet in a zone referred to as the 
Inter-Tropical Convergence Zone (ITCZ). The ITCZ moves northwards and 
southwards across the Basin from about March to October when rainfall is received in 
University of Ghana http://ugspace.ug.edu.gh
40 
 
the region (Barry, et.al, 2005). In the southern and middle parts of the Basin two rainy 
seasons are experienced with a double maxima regime in May to August (major peak) 
and September to October (minor peak) (Codjoe, 2004). 
 
Temperature 
Mean daily temperature in the LVB ranges from 32–44°C, whilst night temperatures 
can fall to as low as 15⁰C. Temperatures in the South could fall as low as 18⁰C under 
the influence of sea breeze, particularly during the months of July and August when 
marine coastal upwelling occurs (UNEP-GEF, 2010). CSIR-WRI (2000), projected a 
temperature increase by 2.5-3.2oC by 2100 across the country. 
 
Rainfall 
Furthermore three types of climatic zones can be identified in the Volta Basin: the 
humid south with two distinct rainy seasons; the tropical transition zone with two 
seasons of rainfall very close to each other; and, the tropical climate, north of lat 9° N, 
with one rainfall season that peaks in August. The average annual rainfall varies across 
the Basin from approximately 1500 mm in the south eastern section of the Basin in 
Ghana, to about 360 mm in the northern part of Burkina Faso. The Lower Volta Basin 
has two wet seasons, one from March to July, and a shorter wet season from September 
to November, corresponding to the northern and southern passages of the ITCZ across 
the region (Rodgers et al., 2007). 
 
3.1.4  Geology, Relief, and Soils 
The (UNEP-GEF, 2010) report, indicates that the geology of the main Volta is 
dominated by the Voltaian system, which consist of consists of Precambrian to 
University of Ghana http://ugspace.ug.edu.gh
41 
 
Paleozoic sandstones, shales and conglomerates. The other geological formations 
include the Buem formation, Togo series, Dahomeyan formation, and Tertiary-to-
Recent formations. However the Dahomeyan formation which consists of mainly 
metamorphic rocks, including hornblende and biotite, gneisses, migmatites, granulites, 
and schist occurs mainly in Southern portion of the Lower Volta. The Basin has on its 
western side a chain of mountains, the Akwapim Togo Ranges, Fazao mountain and 
Atakora ranges in Benin.  
 
The relief of the Lower Volta is generally low with an average elevation of (0-150m) 
above sea level, with the average mean altitude of the main Volta Basin approximately 
257m (Barry et al., 2005). The soils in the Volta Basin and Lower Volta Basin for that 
matter are related to the agro-ecological zones within the Basin (UNEP-GEF, 2010). 
The agro ecological zones of the Volta Basin includes Coastal savannah, Sudan 
savannah, Guinea savannah, Forest savannah transitional zone, Semi-Deciduous forest 
and High rainforest. The geological formations have given rise to different parent 
materials from which most of the soils in the Lower Volta Basin and Ghana are derived 
(Ministry of Lands and Forestry, 2002). Amatekpor (1999), reported that the soils 
developed over the Lower Volta Alluvium, Coastal Savanna Zone include; Chichiwere, 
Amo, Tefle and Aveyime. The soils of the Lower Volta flood plain which covers an 
area of 294.0sq km (29,280ha) is good for agriculture (Gordon and Amatekpor, 1999).  
 
3.1.5  Land-use and Land Cover 
All categories of land-use occur in the Volta Basin, including; rain-fed and irrigated 
agriculture, grazing, forestry, settlements, public services, wildlife preservation, 
University of Ghana http://ugspace.ug.edu.gh
42 
 
reservoirs (manmade and natural), protected lands and many more (Amatekpor, et al., 
1999). 
 
Rain-fed agriculture which is based on the land rotation farming system, also known as 
bush fallow system, is the most common type of land-use in the area. (Titriku, 1999). 
The land cover of the Volta River Basin consists predominantly of sub-humid 
savannah; mainly tall grasses interspersed with fire resistant tress in the Black and 
White Volta Sub-Basins whereas the Lower Volta has a range of vegetation from closed 
forest to humid savannah, mangrove and coastal strand vegetation and thicket. Short 
grasses also occur more within the Lower Volta that enhances the rearing of herbivores 
such as cattle, goat sheep among others (UNEP-GEF, 2010). 
3.1.6  Vegetation 
Climatic and edaphic factors are known determinants of vegetation cover of an area. 
The dominant tree species in the area is the "Neem" tree. It can be found everywhere in 
the Lower Volta Basin. Generally, the Lower Volta has a composite vegetative cover 
of moist semi deciduous forest, short grasslands with some mangrove zones (UNEP-
GEF, 2010) 
 
Table 3. 3: Typology of vegetation of the Lower Volta River Basin 
Northern section Mixed savannah woodland, tall grassland 
with fire resistant trees. 
Central section Derived savannah interspersed with semi 
deciduous rain forest 
Southern Semi-deciduous rain forest with patches 
of derived savannah with short grasses 
and mangroves 
Source: GEF-UNEP (2010) 
 
The area is predominantly a coastal savannah grassland and scrub, and the predominant 
vegetation type is grass with trees such as neem, sheanut, baobab and dawadawa in the 
north with acacia species and mangroves in the south (Ayivor, 1999). 
University of Ghana http://ugspace.ug.edu.gh
43 
 
3.2  Methods 
3.2.1  Land-use/Cover Study 
For this study, satellite imagery captured by LandSat 8 (OLI & TIRS) and LandSat 7 
ETM+ of different dates was obtained from the United States Geological Survey 
(USGS) website, (http://earthexplorer.usgs.gov/) and "global visualization viewer".  
The path and row of the LandSat 8 2015 images from the USGS website are wrs_path 
= 193 and wrs_row = 056 respectively. The acquisition date of this image was 4th 
January 2015. The 2003 LandSat 7 ETM+ images was also downloaded from the USGS 
website from wrs_path = 193 and wrs_row = 056. The acquisition date of this image 
was 12th February 2003. 
 
These two images were selected based on the quality of the satellite image and the 
amount of cloud cover in the image. Therefore a final decision was made after carefully 
considering these factors and how they could interfere with the accuracy of mapping 
land-use/cover for this study. The selected datasets were cloud free images acquired in 
between January and March, i.e, for the dry season.  
 
 
 
 
 
 
 
 
 
 
University of Ghana http://ugspace.ug.edu.gh
44 
 
Table 3. 4: Land-Use/Cover Classification Scheme and Description 
Land-Use/Cover Classes General Description 
Waterbody Rivers, Inland waters and reservoirs, 
creeks, permanent lakes and other 
intermittent ponds 
Farmlands (Rain fed/Irrigation) Land prepared for growing agricultural 
crops, covered with crops and land 
under preparation. This category 
includes rain fed and irrigation 
agriculture. 
Grassland/Shrub Areas covered with natural grass and 
small shrubs.  
Mixed savannah/Coastal veg Areas covered with patches of trees, 
bushes and woodlots. Riverine 
vegetation as well as mangroves and 
some species of water weeds. 
Semi Deciduous Forest/Protected Area Areas covered by trees with minimum 
size of 0.5ha. Dominated by natural 
high forests which are deciduous. 
Bare/Burnt Area Areas burnt by fire (bush burning) and 
exposed soils due to erosion and or 
misuse, especially overgrazing. 
Built Area Built-ups (buildings) in both rural and 
urban areas.  
Source: Adapted and modified from UNEP-GEF, 2010 
 
Remote Sensing and GIS Data Processing and Analysis 
Pre-processing 
The time series satellite images obtained between (2003 and 2015) and were processed 
using Environment for Visualizing Images (ENVI) and ArcGIS softwares. 
Georeferencing and co-registration 
 
The obtained Landsat images used for this study had been already corrected to the 
Universal Transverse Mercator UTM 30 N based on the WGS 84 projection. There 
was however no need to georeference the images.  
 
 
 
 
University of Ghana http://ugspace.ug.edu.gh
45 
 
Radiometric Calibration 
 
Radiometric calibration was used since it’s essential for the analysis of multi-temporal 
images acquired on different dates and by different sensors. Each sensor has its’ own 
calibration parameters in recording digital numbers. These digital numbers recorded 
will thus represent different radiance and reflectance values (Mantey and Tagoe, 2013). 
As mentioned earlier this study was meant to detect change by comparing multi-
temporal satellite images. For calibration of the obtained satellite images, the equations 
presented by Chandler et al. (2009) for converting calibrated Digital Numbers to 
absolute units of At-sensor spectral radiance and Top- of Atmosphere (TOA) 
reflectance were done. Firstly the calibrated digital numbers (DN) of both images (TM 
and ETM+) were converted to absolute units of At-sensor spectral radiance using 
equation (1). Secondly the At-sensor spectral radiance values (Lλ) were then converted 
to Top-of-Atmosphere (TOA) reflectance (pλ) using equation (2). 
Equation (1) 
Lλ = Grescale * Qcal + Brescale 
 
Equation (2) 
𝜌λ =  
𝜋 ∗ 𝐿λ ∗ 𝑑
2
𝐸𝑆𝑈𝑁λ ∗ cos𝜃𝑠
 
 
Where;  
Lλ  = The spectral radiance 
Grescale  = Band specific rescaling gain factor 
Brescale  = Band-specific rescaling bias factor 
QCAL  = The quantized calibrated pixel value in digital number of each band 
pλ  = The top-of-atmosphere reflectance 
University of Ghana http://ugspace.ug.edu.gh
46 
 
d  = The distance from the earth to the sun in Astronomical Units (AU) 
ESUNλ  = The mean solar exoatmospheric irradiance 
θs  = The solar zenith angle 
 
Layer Stacking and Spatial subset 
All bands (band 1-band 8) of the two satellite images (2003, 2015) were put together 
through a process called “layerstack”. This was performed using the ENVI software 
under the tools tab. 
“Spatial Subsetting”, was performed by using the “Resize data (spatial/spectral) tool in 
ENVI. This tool was used to extract the area of interest from the obtained satellite 
images.  
The spatial subset was carried out using the coordinates of what the VBRP used in 
defining the Lower Volta Basin. 
 
Analysis Using ENVI and ArcGIS software’s 
The methods of analysing the data included, image classification using ENVI, change 
in vegetation cover using the normalized difference vegetation index (NDVI), 
development of land-use land cover classes, change detection and other GIS based 
processes such as on screen digitizing of different land-use features from the satellite 
images.  
 
Band 4, 3 and 2 of the Landsat 8 image was loaded into the ENVI software to obtain 
the true/natural colour. Subsequently other bands were selected to give a false colour 
and/or infrared imagery to aid with classification and further analysis. Other band 
combinations which were used included (5 4 3 Color Infrared (vegetation), 6 5 2 
University of Ghana http://ugspace.ug.edu.gh
47 
 
Agriculture, 7 6 5 Atmospheric Penetration, 6 5 4 Vegetation Analysis) The Landsat 7 
ETM+ followed a similar process to aid with the image classification. 
 
    
Plate 3. 1: Feature verification using Plate 3. 2: Feature verification 
“kml” overlay in google earth using spear tool 
 
  
Plate 3. 3: False colour rendition image. Showing vegetation as bright red, crops as 
pink and water as black. 
 
Land-Use/Cover Classification and Change Detection 
Two image classification processes were used in this study (Supervised classification 
and Post classification). However unsupervised classification which is the identification 
of natural groups, or structures, within multispectral data (Rai et.al, 2010) was 
employed and compared to the supervised classified images. This was done to verify 
and select the method which yielded the best result.  
 
 
University of Ghana http://ugspace.ug.edu.gh
48 
 
Unsupervised classification 
The Iterative Self Organizing Data Analysis (ISODATA classification) technique was 
used under the unsupervised classification method. This technique automatically 
created as many classes as justified by the data. It merged all classes for which there 
was a minimum distance between means and split classes for that which had high 
standard deviation within a group.  
 
Unlike the K-means method which required that a specific number of classes is chosen, 
this method provides a range. A range of classes (15 minimum and 30 maximum) 
therefore was selected for classification. 27 classes were generated after running the 
statistical process. These classes were combined until 7 classes were left. For the 
purpose of this study, and to avoid having so many classes of vegetated land cover, a 
classification scheme adopted from (UNEP-GEF, 2010) was used (See Table 3.4). 
Supervised classification  
The supervised classification method is used to cluster pixels in a dataset into classes 
corresponding to user-defined training. This classification type requires that training 
areas are selected for use as the basis for classification. In selecting the training areas 
the ENVI software enables the user to make use of what is known as region of interest 
(ROIs).  
 
Different ROIs were carefully selected as training areas for both 2003 and 2015 Landsat 
images. Various comparison methods were then used to determine if a specific pixel 
qualifies as a class member. With the broad range of different classification options 
provided under ENVI’s supervised classification namely “Maximum Likelihood 
University of Ghana http://ugspace.ug.edu.gh
49 
 
classifier” was used because it produced the best result for the study area and it has 
been used and recommended by other researchers who have worked in the Volta Basin.  
Change Detection 
The Post classification method provides classified images with evaluation of 
classification accuracy through post processing. An extra feature of the post 
classification method is that it allows for the extraction of statistics from the image that 
were used to classify the satellite image (Exilis, 2014). Accuracy assessment was done 
for both classified images in this study by using the confusion matrix tool under this 
post classification process. Classes were combined, sieved and lumped using this 
process as well to clean up the classified land-use/cover images that were created. The 
minority/majority analysis in ENVI was conducted to remove all unclassified pixels 
from the final classified image.  
Change detection statistics 
A change detection statistics was run on the two final classified images. With the 2003 
image set as the initial state and the 2015 image set as the final state, the change 
detection computed class/pixel change, area change and percentage change for all 
classified land-use/cover types. The classes were then converted to vector layers and 
saved as a shape file. It was opened in ArcMap for finalization of the land-use/cover 
map. 
Ground Truthing/Verification 
Ground truth verifications was conducted in the study sites with the aid of a Global 
Positioning System (GPS) (Garmin Etrex GPS 20) to confirm and help check the 
accuracy of land cover and use classification process. All the six study areas were 
University of Ghana http://ugspace.ug.edu.gh
50 
 
visited with GPS coordinates taken at specific places to aid in accuracy of classification. 
Some of the coordinates taken of features include, cleared farmlands, cultivated farm 
lands, bare land, bushes, partially bare grounds, grass lands, burnt fields and built up 
areas (markets, residential areas). Some of the ground control points were taken at 
converging points of certain features such as the intersection points of roads (major 
road junctions). 
 
The use of Google Earth Pro was also used in confirming some land cover/use features. 
There are two procedures that were employed in using google earth as a post 
verification tool. The employment of the "SPEAR" tool in "ENVI" was used in the 
classification process. This tool takes the coordinates of a selected cover type in ENVI 
and plots to the exact location (coordinates) in google earth. This helps verify a feature 
found in the satellite image. After classification was done for both satellite images 
further verification was done by; 
 Exporting the image to ArcMap. 
 Created polygon using the draw tool over the area (class) to be verified. 
 Converted polygon feature to kml file type and saved. 
 The “kml” file type was opened in Google Earth.  
 
 
Normalised Difference Vegetation Index  
The analysis of vegetation cover changes was done using the Normalized Difference 
Vegetation Index (NDVI). This method is used to analyse the amount or extent of 
change in vegetation. It is calculated by  
NDVI =
𝑁𝐼𝑅 − 𝑅𝐸𝐷
𝑁𝐼𝑅 + 𝑅𝐸𝐷
 
University of Ghana http://ugspace.ug.edu.gh
51 
 
 
Where RED is the visible red light portion (band 3) (620-750nm) of the electromagnetic 
spectrum and NIR is the invisible radiant energy (band 4) (750nm -1mm). The NDVI 
itself has an index varying between -1.0 to +1.0. 
It was used to determine the extent to which vegetation cover had changed in the two 
satellite images.  
    
Figure 3. 2 NDVI map of the two Landsat images 
The NDVI generated map was compared to the generated LULC map of the study area.  
It was used as a check to assess which of the Landsat images had a higher vegetation 
cover. 
 
3.2.2  Climatic Study 
Monthly climatic data for the past five (5) decades (1960-2013) was acquired from the 
Meteorological Services Agency of Ghana (GMet). Units of climatic data used in this 
study were rainfall and temperature. This data was from weather stations of the various 
districts of the study area. 
 
University of Ghana http://ugspace.ug.edu.gh
52 
 
Thiessen Polygon Method  
In calculating the average rainfall and temperature for the Lower Volta Basin, the 
values for the various weather stations in the above mentioned districts were calculated 
by using the Thiessen Polygon method.  
The Thiessen Polygon interpolation algorithm embedded in ArcGIS 10.2.2 was 
employed to calculate the areal average annual temperature and rainfall for seven 
weather stations Kpong, Akuse, Sogakope, Ada, Asutuare, Asuogyaman and Adidome. 
This method is based on the true assumption that rainfall is never uniform over the 
entire area of the Basin or catchment. But it varies in intensity and duration from place 
to place. Thus the rainfall recorded by each rain gauge station should be weighted 
according to the area, it represents (http://www.yemenwater.org, 2015). The method is 
suitable under these conditions;  
 Areas of moderate size,  
 when rainfall stations are few compared to the size of the Basin and  
 In moderate rugged areas.  
The polygons were constructed in ArcMap and each area was calculated as well as the 
whole area. 
The formula used is presented below. 
Pav= 
(𝑃1×𝐴1)+(𝑃2×𝐴2)+⋯…..(𝑃7×𝐴7)
𝐴1+𝐴2+...𝐴7
 
Where Pav is Precipitation average 
P1 is Rainfall measured at station 1 and  
A1 is area in km2 of the area in which the weather station 1 is located. 
The formula below was used in calculating for the average temperature for the Lower 
Volta Basin as well. 
University of Ghana http://ugspace.ug.edu.gh
53 
 
Tav= 
(𝑇1×𝐴1)+(𝑇2×𝐴2)+⋯…..(𝑇7×𝐴7)
𝐴1+𝐴2+...𝐴7
 
Where Tav is temperature average 
T1 is the temperature measured at station 1 and  
A1 is area in km2 of the area in which the weather station 1 is located. 
Trend line analysis of climatic data was analysed using a linear regression model. 
Graphs were drawn, and rainfall and temperature anomaly analysis were performed. 
3.2.3  Crop Suitability Study  
EcoCrop Model 
The EcoCrop module in the DIVA GIS programme was used to create a crop suitability 
map for Zea mays (maize), Capiscum annuum longus (pepper), and, Manihot esculenta 
(cassava). These crops were selected because they are grown in all districts of the Lower 
Volta Basin. 
 
This EcoCcrop module uses FAO's database of environmental requirements for plant 
species which is used to identify possible crops that can grow in a particular 
environment or geographic area (Hijmans et al., 2005). 
In using the EcoCrop module in the DIVA GIS programme to determine the suitable 
areas for growing these food crops, precipitation and temperature were the two main 
climatic variables that were used. 
The temperature parameters considered in determining the suitability of a growing 
season for the crops are presented below;  
KTMP: absolute temperature that will kill the plant. 
TMIN: minimum average temperature at which the plant will grow 
University of Ghana http://ugspace.ug.edu.gh
54 
 
TOPMN: minimum average temperature at which the plant will grow optimally 
TOPMX: maximum average temperature at which the plant will grow optimally 
TMAX: maximum average temperature at which the plant will cease to grow 
The rainfall parameters considered are presented below; 
RMIN: minimum rainfall (mm) during the growing season 
ROPMIN: optimal minimum rainfall (mm) during the growing season 
ROPMAX: optimal maximum rainfall (mm) during the growing season 
RMAX: maximum rainfall (mm) during the growing season 
The climatic data used for this mapping was downloaded from the www.worldclim.org 
website. This is because the DIVA GIS program uses a specific file format known as 
“CLM” which it refers to for climate data.  
The source of the current (1950-2010) climate files downloaded was WorldClim 
version 1.4 from the www.worldclim.org website.  
The data resolution was 2.5 arc minutes  
The source was downscaled GCM outputs based on the HADCM3 model of the 
projected future conditions (~2050). Also at a resolution of 2.5 arc minutes. (Available 
at http://www.worldclim.org/futdown.htm)  
 
The source of the projected future conditions (~2050) climate files downloaded was 
Govindasamy et al., (2003). These data were subsequently downscaled and matched to 
the WorldClim estimates of current climate also at a resolution of 2.5 arc minutes. 
The downloaded files include (precipitation, minimum temperature, maximum 
temperature, and altitude) “BIL” files which are extracted and converted into “GRD” 
files using ArcGIS. These are also further converted into CLM files using the DIVA 
University of Ghana http://ugspace.ug.edu.gh
55 
 
GIS program. The process includes Importing and converting BIL files into GRD files 
in DIVA-GIS and Creating CLM files in DIVA-GIS (Ramirez and Bueno, 2009). 
 
"In EcoCrop, the growing period is defined in days between Gmin and Gmax (start of 
growth and end of growth respectively). In DIVA-GIS, 12 possible growing seasons 
are considered, starting on the first day of each month. The length of the growing season 
is defined as the average of Gmin and Gmax" (Hijmans et al. 2005). 
 
To predict the suitable areas for crop areas for the Lower Volta Basin using the Ecocrop 
model in GIS, the rain during growing season and interaction (temperature times 
precipitation scores). 
3.2.4  Data 
Ghana district map 
An outline of the Ghana district map and Volta River (“shp” file format) was obtained 
from the Remote Sensing and GIS laboratory of the Geography and Resource 
Development Department, University of Ghana.  
Crop production data 
Crop production data for rain fed agriculture over the past decade (2003-2013) was also 
obtained from the district offices of the Ministry of Food and Agriculture (MOFA) at 
Lower Manya Krobo and Ada East. 
Additional data was obtained from books, thesis, journals, articles and internet sources 
to review in-depth literature. Reports were also collected from various institutions such 
as the District offices, Volta River Authority (VRA) and Ministry of Food and 
Agriculture (MOFA).  
University of Ghana http://ugspace.ug.edu.gh
56 
 
3.3  Socio Economic Study  
A total of 200 questionnaires were administered specifically to farmers who engaged 
in rain-fed agriculture only using the Purposive and random sampling technique. The 
questionnaires was pretested in the study area before they were finally administered.  
A face to face interview was conducted with farmers from the Ada East and Lower 
Manya Krobo district, with the aid of structured questionnaires made up of open and 
closed ended questions.  
 
The purpose of this was to assess the farmer’s knowledge on land-use and cover 
changes, climate change (rainfall and temperature), on their crop yield. Other 
Information sought were related to the influence of rainfall variability on planting times, 
crop productivity for the previous years (2013 and 2014) and their view on what they 
(farmers) think has been the major influencing factor of crop yield was inquired. As 
part of the social appraisal, the educational levels of the respondents were sought. 
3.3.1  Statistical Analysis 
The variables and responses from the questionnaire were subjected to statistical 
analysis. Descriptive statistics including frequencies gave a result of comparisons, 
means, sums and standard deviations. Cross tabulations, regression and Chi square tests 
were performed as well. 
 
  
University of Ghana http://ugspace.ug.edu.gh
57 
 
CHAPTER FOUR 
RESULTS  
Results of the study focused on Land-use/cover change as well as climate conditions 
and its relationship to crop productivity and crop suitability conducted in the Lower 
Volta Basin are presented below.  
4.1  Land-Use/Cover Classification and Change Analysis 
A combination of methods were used to carefully classify both (2003 and 2015) Landsat 
images. The images (Figure 4.1 & 4.2) below shows the Landsat images (2003 and 
2015) of the Lower Volta Basin with region of interests (ROIs) selected for the process 
of supervised classification.  
 
Figure 4. 1: Selected ROI’s used for maximum likelihood supervised classification. 
 
2003 
University of Ghana http://ugspace.ug.edu.gh
58 
 
 
Figure 4. 2: Selected ROI’s used for maximum likelihood supervised classification  
 
The selected regions of interest were carefully done to produce an accurate 
classification representing what was on the ground. 
2015 
University of Ghana http://ugspace.ug.edu.gh
59 
 
4.1.1  Accuracy Assessment  
Statistics of Supervised Classification (2003 and 2015) 
Table 4. 1: Accuracy results for maximum likelihood classification 2003 Image 
Class Water 
Built 
Area Forest/P.A Farm lands 
Mixed 
Savannah Grass/Thicket 
Burnt/Bare 
land 
Salt 
wining 
Pixel 
Total 
User 
Acc.% 
Water 61169 0 0 5 0 35 0 30 61239 99.89 
Built Area 3 3454 1 2661 46 8446 30 40 14681 69.53 
Forest/P.A 0 0 6920 1650 2181 251 4 0 11006 62.87 
Farm lands 0 283 141 20998 417 8621 0 0 30460 68.94 
Burnt/Bare 156 63 0 33 2 1567 14777 845 17443 84.72 
Salt wining 1479 0 0 21 0 81 54 19469 21104 92.25 
Mixed 
Savannah 1 4 380 15099 16374 3118 0 0 34976 46.81 
Grass/Thicket 5 715 38 9171 516 51691 381 0 62517 82.68 
Total 62813 4519 7480 49638 19536 73810 15246 20384 253426  
Producer 
Acc.% 97.38 76.43 92.51 72.3 83.81 70.03 96.92 95.51   
Source: Author’s Computation, 2015  Overall Accuracy = (194852/253426) 76.8%   
Kappa Coefficient = 0.7185 
  
University of Ghana http://ugspace.ug.edu.gh
60 
 
Table 4. 2: Accuracy results for maximum likelihood classification 2015 Image 
Class Water 
Built 
Area Forest/PA Farm lands 
Mixed 
Savannah Grass/Thicket Bare/Burnt 
Salt 
wining 
Pixel 
Total 
User 
Acc.% 
Water 231929 0 27 35 0 0 0 1793 233784 99.21 
Built Area 0 18265 7 8440 5699 246 3932 9 36598 59.91 
Forest/PA 64 9 36794 3381 17610 2168 284 0 60310 61.01 
Farm lands 16 997 72 244026 5742 6278 1785 0 258916 94.25 
Bare/Burnt 1777 1093 271 9665 1766 1182 114706 1717 132177 86.78 
Salt wining 3838 0 0 4 0 0 33 182743 186618 97.92 
Mixed 
Savannah 8 1084 1416 15827 27243 5415 858 52 51903 52.49 
Grass/Thicket 0 618 368 59311 8769 133946 975 0 203987 65.66 
Total 237632 22066 38955 340689 66829 149235 122573 186314 1164293   
Producer 
Acc.% 97.6 82.77 94.45 71.63 40.77 89.76 93.58 98.08     
Source: Author’s Computation, 2015  Overall Accuracy = (989652/1164293)  85.03%   
Kappa Coefficient = 0.8191 
University of Ghana http://ugspace.ug.edu.gh
61 
 
The error/confusion matrix in Table 4.1 indicates classification accuracy of eight 
classes for the 2003 image classification. The producer’s accuracy percentage ranges 
from 97.38% (water body), 76.4% (built up), 92.5% (Forest/Protected Area), 72.3% 
(Farm lands), 83.8% (Mixed Savannah) 70.0% (Grass Thicket), 96.9 (Bare/Burnt land), 
95.51 (Salt wining, Songaw Lagoon). On the other hand, the percentage of the user’s 
accuracy varies from 99.89% (Water), 69.53% (Built Area), 62.8% (Forest/PA), 68.9% 
(Farm lands), 84.7% (Bare/Burnt Area), 92.5% (Salt wining), 46.8% (Mixed savannah 
vegetation) and 82.6% (Grass/Thicket). The overall accuracy of the classification is 
76.88% for the eight classes that were accurately classified with a Kappa statistic of 
0.7185. 
 
For the 2015 image classification, the error/confusion matrix in Table 4.2 indicates 
classification accuracy of eight features as well. The producer’s accuracy percentage 
ranges from 97.6% (water body), 82.7% (built up), 94.4% (Forest/Protected Area), 
71.6% (Farm lands), 40.7% (Mixed Savannah) 89.7% (Grass Thicket), 93.5 
(Bare/Burnt land), 98.0 (Salt wining, Songaw Lagoon). On the other hand, the 
percentage of the user’s accuracy varies from 99.2% (Water), 59.9% (Built Area), 
61.0% (Forest/PA), 94.2.9% (Farm lands), 86.7% (Bare/Burnt Area), 97.9% (Salt 
wining), 52.4% (Mixed savannah vegetation) and 65.6% (Grass/Thicket). The overall 
accuracy of the classification is 85.0% for the eight classes that were accurately 
classified with a Kappa statistic of 0.8191. 
4.2  Land-Use/Cover Change (Change Detection) 
The tables 4.3 and 4.4 shows change detections statistics for the initial state (2003) 
classes seen in the columns and the final state (2015) classes seen in the rows. This was 
used to determine the distribution of classified pixels that changed classes. For each 
University of Ghana http://ugspace.ug.edu.gh
62 
 
“initial state class” (in each column), the table indicates how these pixels were classified 
in the “final state image” (in each row). 
Farmlands (Rain-fed and Irrigation) 
As shown in Table 4.3 and Table 4.4, about 51.12% (29597.6ha) of farmlands area was 
maintained in the study area. A total of 14,699.9ha representing 25.39% of grass/shrub 
area changed into farmlands, indicating the extent to which land with vegetation cover 
i.e. grass, bush and thickets are being turned to farmlands. Also about 8.29% (4799ha) 
of mixed savannah vegetation had been turned to farmland. Furthermore about 6.31% 
(3.654.9ha) of bare land/burnt area (in 2003) changed to farmlands in 2015. Burning of 
vegetation in the LVB is done for two main reasons, to prepare land for cultivation and 
secondly to enhance the growth of new grass for grazing purposes. The area, bare/burnt 
area is more likely to change significantly to either farmlands or grasslands. Overall, a 
total area of 57896ha of land classified as farmland in 2003, whilst about 95,099ha of 
land was also classified as farmlands in 2015. The class change for the farmlands was 
48.9% (28298.7ha), implying that other land-use/cover categories had been converted 
to farmlands. The image difference between 2003 and 2015 was 64.3% (37202.5ha) 
which indicates that there has been a considerable positive change or gain in farmlands 
over the period of study.  
Waterbody 
Waterbody which consist mainly of the Kpong head pond and Volta River had 87% 
(14,801 ha) retained after the image time difference analysis. In table 4.3, (8.64%) 
representing 1,469ha of bare land was gained from water. This is due to the fact that 
the salt which is mined in the Songaw lagoon, reflected more like the bare/burnt area 
land cover type and was thus classified as such. In table 4.3 (2.13%) representing 
University of Ghana http://ugspace.ug.edu.gh
63 
 
362.3ha was converted to farmlands. Forest/PA, Grass/Shrub and Mixed Savannah had 
1.05% (177.9ha), 0.25% (42ha) and 0.65% (114ha) changed respectively to waterbody. 
A total class of 19,610ha ha was classified as waterbody in the year 2003 whilst about 
17,007ha had been classified as waterbody in 2015. The class change for the waterbody 
was 7,5603ha.4ha (13%). The image difference between 2003 and 2015 was -5.4% 
(2,206ha), which shows that there has been some change (reduction) of water coverage 
over the years mainly accounted for by changes in the Songaw lagoon. 
 
Bare/Burnt Area 
This feature had 4,024ha of land area maintained. A total of 77,042ha was gained. This 
was mainly due to an increase in the area of salt which was classified under Bare/Burnt 
area at the Songaw lagoon. Other reasons accounting for this change include the 
increase in the number of manmade (built) water reservoirs for irrigation farming. 6.9% 
of bare/burnt area (6,325ha) was gained from grass/shrub land-use/cover type. Also 
3.1% representing 2,925.3ha was changed to farmlands. 0.3% (299ha) and 0.3% 
(278ha) was changed to Built-up area and Forest/PA respectively. About 1% (840) was 
also gained from the Mixed savannah in 2015. A class total of (41163ha) was classified 
in 2003 and 91734ha was classified as Bare/Burnt area. The class change for the 
bare/burnt land was 55.1% (87,709.2) telling that this land-use/cover had gained from 
other land-use/cover types. The image difference between the two images was 55.1% 
(50571ha), which indicates that the bare/burnt area had increased in coverage over the 
period of study.  
 
Forest/PA 
From Table 4.3 and 4.4, 27.5% (2,982.6ha) was maintained as Forest/Protected areas.  
University of Ghana http://ugspace.ug.edu.gh
64 
 
The computed statistics show that Forest/PA had a class change of 48.8% (87,709.2). 
Generally, an overall total class of (12,535ha) in 2003 and (11,411.9) in 2005 was 
classified as forest and protected area. This feature lost 27.6% (3,462.5) to grass/shrub 
land-use/cover. 18.6% (2337.1ha) was also lost to the mixed savannah land-use/cover 
type. Also 23.7% (2,982.7) was lost to farmlands which show encroachment by farmers. 
Finally 1% (128.3ha) was converted to Built-up area. The image difference between 
the years was -8.9% (-1123.5) which shows a reduction or loss in Forest/PA, to other 
land-use/cover types, especially, grasslands and farmlands.  
 
Grass/Shrub 
40.3% (55,536.0) of grass land and shrub did not change for the period 2003 to 2015. 
A total of 27.1% (37,320ha) had changed or been converted to farmlands. As mentioned 
earlier, farmlands area has increased and grassland and shrub had decreased. 16.7% 
(22,968ha) is also lost to Bare/Burnt area. It is to be noted that there is burning done 
especially in grassland areas by Herder’s. 10% (13,810ha) though quite small has been 
transformed to Mixed Savannah land cover type. Again, 4.3% (5854ha) was converted 
to Built-up Areas in in 2015. 0.6% (121.7ha) had also been converted to water. Probably 
from water standing in less permeable areas. A total class of 137,690.5ha was classified 
as Grass land/Shrub land-use in 2003 and 103,128.9 in 2015. The class change for the 
grassland /shrub is 59.6% (82154.4ha). The image difference between the two years 
was -25.1% (82,154.43ha) showing that there has been a reduction in the grassland area 
in the LVB over the period 2003 to 2015.  
 
Mixed Savannah 
In the mixed savannah class classification, 13% which is (7,589ha) area coverage 
approximately had not change over the period of study (2003 to 2015). However 31% 
University of Ghana http://ugspace.ug.edu.gh
65 
 
representing (18,129ha) was lost to farmlands. 37.4% (21,924.5ha) of this cover type 
was also converted into grasslands and shrub. 8% (4634ha) of the mixed savannah land 
cover type had also been changed to forest areas. This is as a result of some smaller 
vegetation which have reflectance characteristics similar to that of forest type 
vegetation. Also, 8.3% (4,887ha), was changed to bare/burnt area. Finally 1.6% 
(967.8ha) and 0.6% (388.7ha) representing built area and water respectively also gained 
from the loss of some portions of mixed savannah. Though almost insignificant it 
contributes to the change occurred in the LVB. A total class of 5,8520ha was classified 
as mixed savannah in 2003 and 31,425.5 in 2015. Total class change for the mixed 
savannah was 87% (50,931.7ha). The image difference between 2003 and 2015 was -
46.3% (-27,095.9ha) which indicates that quite a considerable area had been lost to 
other land-use/cover change.  
Built- Up Area 
For the built up area, 32.7% remained unchanged in the change detection statistical 
analysis. The overall class total for the year 2003 was 16,647.48ha and 17,181ha in the 
year 2015. 23.9% representing (3,986ha) of Bare/Burnt area was changed into the Built 
up class. Indicating that, some form of construction had taken place on the bare land. 
6.8% (1,139ha) of grass lands/shrub had also been lost to built-up area. Farmlands lost 
to built-up area also represented 22.7% (3,781.8ha). Also 11.6% (1,935ha) of mixed 
savannah vegetation had been lost to the built area class. The image difference between 
2003 and 2015 was positive, indicating 3.21% (534.3ha) had been gained from other 
classes. 
University of Ghana http://ugspace.ug.edu.gh
66 
 
Table 4. 3: Land-Use/Cover Change Detection Statistics (%) 
F
in
al
 S
ta
te
 (
2
0
1
5
) 
Initial State (2003) 
 Water Built Area Forest/P.A Farm lands Burnt/Bare land Grass/Thicket Mixed Savannah 
Water  87.03 1.44 0.09 0.01 6.03 0.09 0.66 
Bare/Burnt  8.64 23.95 1.38 6.31 83.39 16.68 8.35 
Forest/PA  1.05 0.70 27.45 1.19 0.30 1.51 7.92 
Grass/Shrub 0.25 6.84 27.62 25.39 6.90 40.33 37.46 
Farm lands  2.13 22.72 23.79 51.12 3.19 27.10 30.98 
Mixed Savannah  0.67 11.62 18.64 8.29 0.92 10.03 12.97 
Built Area  0.24 32.72 1.02 7.68 0.33 4.25 1.65 
Class Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 
Class Changes 12.97 67.28 72.55 48.88 95.61 59.67 87.03 
Image Difference -5.45 3.21 -8.963 64.257 55.128 -25.101 -46.3 
Source: Author’s Computation, 2015 
  
University of Ghana http://ugspace.ug.edu.gh
67 
 
Table 4. 4: Land-Use/Cover Change Detection Statistics (Area/ha) 
 Initial State (2003) 
F
in
al
 S
ta
te
 (
2
0
1
5
) 
 Water Built 
Area 
Forest/P.A Farm 
lands 
Burnt/Bare land Mixed 
Savannah 
Grass/Thicket Row Total 
Water 14800.68 239.22 11.07 7.11 12.41 388.71 121.77 15580.97 
Bare/Burnt 1469.25 3986.55 172.71 3654.90 4024.89 4886.82 22968.00 41163.12 
Forest/PA 177.84 115.92 3440.79 688.41 278.19 4633.92 2076.84 11411.91 
Grass/Shrub 42.03 1139.04 3462.48 14699.97 6324.75 21924.54 55536.03 103128.84 
Farm lands 362.34 3781.80 2982.60 29597.67 2925.36 18129.06 37320.03 95098.86 
Mixed Savannah  114.12 1935.18 2337.12 4799.07 840.33 7589.25 13810.41 31425.48 
Built Area 40.77 5446.44 128.34 4445.10 299.07 967.86 5854.23 17181.81 
Class Total 17007.03 16647.48 12535.47 57896.37 91734.12 58520.97 137690.46 0.00 
Class Changes 2206.35 11201.04 9094.68 28298.70 87709.23 50931.72 82154.43 0.00 
Image Difference -14001.06 534.33 -1123.56 37202.49 50571.00 -27095.49 -34561.62 0.00 
Source: Author’s Computation, 2015 
University of Ghana http://ugspace.ug.edu.gh
68 
 
 
Figure 4. 3 Land-Use/Cover map of the LVB (2003 & 2015). Source: (Author’s Construct 2015). 
University of Ghana http://ugspace.ug.edu.gh
69 
 
The map (Figure 4.3) shows the land-use/cover map for the years 2003 and 2015. This 
is the result of the two carefully classified images of the Lower Volta Basin. By visually 
comparing both images, changes in classes are noticeable without difficulty. In 2003 
35.1% of the LVB was covered with grasslands/shrub whereas farmlands took 14.8% 
of the total area in the LVB. However In 2015 grasslands/shrub covered 32.3% and 
farmlands covered 29.8%. 
Figure 4.3 shows the area of change for the seven classified land-use/cover classes in 
the LVB over the 13 year period (2003-2015) 
 
Figure 4. 4: Land-use/cover change over the period of study (2003-2015) 
From the Figure 4.4, one can tell that farmlands have increased considerably over the 
years. Bare/Burnt areas have decreased also. The mixed savannah land cover type and 
the grass/thicket have also decreased extensively. From figure 4.4, it is noticed that all 
vegetation classes have decreased considering the time period under study. However 
farming area and crops have gained more coverage area over the years. Even though 
the classified forest/protected area has decreased, it is quite marginal. The same can be 
said about the built area as well as the waterbody land-use/cover features.  
0.00
20000.00
40000.00
60000.00
80000.00
100000.00
120000.00
140000.00
160000.00
A
re
a 
H
e
ct
ar
e
s
Initial state (2003) Final state (2015)
University of Ghana http://ugspace.ug.edu.gh
70 
 
 
Figure 4. 5: Land-Use/Cover change in Kpong area 
Source: Author’s Construct, 2015 
From the classified map figure (4.5), the built-up area (colour red) showing the Kpong 
Township has increased over the years. As it can be seen from the map, the built-up 
area replaced another land cover feature (vegetation). The 2015 image (right) also 
shows an increase in the area of bare land.  
Figure 4.5 is a classified map showing a section of the Lower Volta Basin which has 
undergone considerable change. This area is right below the Kpong head pond, between 
Akuse and Asutuare. 
The map shows how farmlands have developed along the stretch of the Volta river in 
2015 (Figure 4.6). Also apart from the statistical difference mentioned earlier, a visual 
comparison of the two images shows a change in farmlands and waterbody. As can be 
seen the natural ponds at the Lower right corner of the 2003 classified image has 
reduced in the 2015 classified image. 
University of Ghana http://ugspace.ug.edu.gh
71 
 
 
 
 
Figure 4. 6: Land –Use/Cover Change detection (2003 and 2015)  
 
 
University of Ghana http://ugspace.ug.edu.gh
72 
 
 
Figure 4. 7: Commercial irrigation farming area at Asutuare. 
 
  
Plate 4. 1:1 Rice farms at Asutuare  Plate 4. 2: Banana plantation 
 (Golden Exotics Co. Ltd). 
   
 
The Ada small farm irrigation scheme was established to provide small holder farmers 
on the stretch between Kasseh and Big Ada with an alternative source of water for 
farming. 
University of Ghana http://ugspace.ug.edu.gh
73 
 
        
Plate 4. 3: Astrium imagery showing Plate 4. 4: Zoomed image showing  
a section of irrigation scheme. some farms, reservoir and pump station. 
Source: Astrium image. Google earth, (2015) 
Table 4. 5: Area coverage of each land-use/cover type in the LVB 
Land-use/cover type Area (ha) Percentage covered in 
LVB 
Water 19610.09 6.1 
Bare/Burnt 41163.1 12.9 
Forest/PA 11411.9 3.6 
Grass/Shrub 103129 32.3 
Farm lands 95098.9 29.8 
Mixed Savannah  31425.5 9.9 
Built Area 17181.8 5.4 
 
Table 4.5 shows the classified types of land-use/cover types and their coverage area in 
the LVB for the year 2015. Currently Grass lands and farmlands constituting 32.3% 
and 29.8% respectively dominate more than half of the LVB. These two are the most 
dominant land-use types in the LVB currently. 
 
University of Ghana http://ugspace.ug.edu.gh
74 
 
  
Plate 4. 5: Astrium image showing a section of Dordoekorfe irrigation scheme. 
 Source: Astrium image. Google earth, (2015) 
                    
Plate 4. 6: Reservoir at Dordoekorfe  Plate 4. 7: Pump station at  
   Dordoekorfe 
 
Source: Author’s Photograph, (2015) 
 
An interesting phenomenon which began in the year 2005 has considerably changed 
or altered the land-use type in the Ada Foah Township. This phenomena could be 
referred to as urban gardening. Vegetable farming, which includes mainly carrots, 
pepper and tomatoes as well as watermelon fruit are grown using a simple irrigated 
University of Ghana http://ugspace.ug.edu.gh
75 
 
system. This has led to a reduction in the rate at which buildings are constructed. 
While some land owners are leasing out their parcels of land for farming, others are 
themselves farming on their lands which were initially meant for building (see Plate 
4.9 and 4.10). 
 
 Plate 4. 8: Satellite image of Ada Foah area  
Source: Astrium image. Google earth, (2015) 
   
Plate 4. 9: Irrigation farm at Ada Foah  Plate 4. 10: Irrigation farm at Ada Foah 
Source: Author’s photograph, (2015) 
 
The pictures (Plate 4.9 and 4.10) show irrigated (underground water) vegetable farm 
fields in the Ada Foah town of the LVB from. Plate 4.8 shows an overview of Ada 
University of Ghana http://ugspace.ug.edu.gh
76 
 
Foah Township. As can be seen from the satellite image (Plate 4.8), there are many 
small farmlands that have sprung up among this fast growing urban town.  
4.3  Incidence and Trend of Climate Factors 
4.3.1  Rainfall 
The average annual total rainfall in the Lower Volta Basin for the period under study 
(1960-2013) was 871.4mm (Appendix 2). The minimum and maximum rainfall 
recorded over the period was 633.3mm and 1,272.8mm respectively.  
Figure 4.8 shows the average annual rainfall pattern from 1960 – 2013 in the LVB 
calculated using the Thiessen Polygon method. From the chart, the first 10years (1960-
1970) experienced generally high rainfall above 1,000mm with the highest rainfall 
recorded (1272.8mm) for the entire 53years period which occurred in 1963. 
 
The regression equation of the rainfall data revealed between the periods 1960 and 2013 
(53years) shows that there is a reduction in rainfall amount at the rate of 3.59mm/year. 
From early 1970s to 1980s low rainfalls are noticed which fall below the average 
rainfall. Despite a considerable increase in 1981, the year 1983 recorded the lowest 
rainfall for the entire period. Up until the 1990s, the period between 1982 and 1992 
experienced shortfalls in rainfall. Rainfall amounts from this year (1992) to the early 
2000s has hovered around the average, with 1996 recording 1008.3mm and 1999 
recording 871.4mm. Most of the years from 1983 to the mid-2000s generally recorded 
low rainfall averages. From 2005, rainfall interestingly increases sharply and drops 
again sharply after 2008. 
University of Ghana http://ugspace.ug.edu.gh
77 
 
 
Figure 4. 8: Average annual rainfall totals, and trend in the Lower Volta Basin.  
 
 
Figure 4. 9: Rainfall anomaly in the Lower Volta Basin 
y = -3.5926x + 970.23
R² = 0.1317
0.0
200.0
400.0
600.0
800.0
1000.0
1200.0
1400.0
1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012
R
ai
n
fa
ll 
(m
m
)
Years
Average Annual Rainfall Annual Average Linear (Average Annual Rainfall)
y = -3.5926x + 98.795
R² = 0.1317
-300.0
-200.0
-100.0
0.0
100.0
200.0
300.0
400.0
500.0
1
9
6
0
1
9
6
1
1
9
6
2
1
9
6
3
1
9
6
4
1
9
6
5
1
9
6
6
1
9
6
7
1
9
6
8
1
9
6
9
1
9
7
0
1
9
7
1
1
9
7
2
1
9
7
3
1
9
7
4
1
9
7
5
1
9
7
6
1
9
7
7
1
9
7
8
1
9
7
9
1
9
8
0
1
9
8
1
1
9
8
2
1
9
8
3
1
9
8
4
1
9
8
5
1
9
8
6
1
9
8
7
1
9
8
8
1
9
8
9
1
9
9
0
1
9
9
1
1
9
9
2
1
9
9
3
1
9
9
4
1
9
9
5
1
9
9
6
1
9
9
7
1
9
9
8
1
9
9
9
2
0
0
0
2
0
0
1
2
0
0
2
2
0
0
3
2
0
0
4
2
0
0
5
2
0
0
6
2
0
0
7
2
0
0
8
2
0
0
9
2
0
1
0
2
0
1
1
2
0
1
2
2
0
1
3R
ai
n
fa
ll 
(m
m
)
Annual Avg Rainfall Linear (Annual Avg Rainfall)
University of Ghana http://ugspace.ug.edu.gh
78 
 
From the Figure 4.9, the anomaly in rainfall pattern of the Lower Volta Basin is 
evident that there were dry and wet periods in the area. Apart from 1963 being the 
wettest year, the period between 1960 and 1969 were relatively wet years with rainfall 
values well above the normal (annual average). The period 1970 to 1995 were dry 
years as most of the years for the period had rainfall values which fell below the 
normal average. Indicating that there were rainfall deficits in the area. 1996 to 2013 
has been characterized by an interesting mix of dry and wet periods. 
4.3.2  Temperature 
The average annual temperature of the Lower Volta Basin ranged from 28○C to 34○C. 
The minimum temperature was 28.3○C and the maximum temperature was 34.8○C. The 
value for the average annual temperature is also 32.7○C (Appendix 2). 
 
From Figure 4.10, the minimum temperature value of 28.3○C was recorded in 1979. 
The maximum temperature of 34.8○C was also recorded in 1965. The period between 
1962 and 1980 recorded temperatures below the average annual temperature (32.7○C), 
indicating that those years were relatively cold years. However in this same period 
(1962-1980) 1965 recorded the highest temperature value of 34.8○C. Again the period 
from 1962 to 1980 can be generally regarded relatively cold with three sharp drops in 
temperature for the years 1969 (29.2○C), 1970 (29.4○C) and 1979 (28.3○C). In the 
succeeding years, from 1982 to 2013 has experienced generally warm temperatures. 
Almost all the values in this period have recorded values above the annual average 
temperature, leading to an increasing level in temperature as indicated by the trend line 
in figure 4.10. There was a generally increasing trend in average annual temperature 
for the study area at a rate of 0.0292◦C/year. 
 
University of Ghana http://ugspace.ug.edu.gh
79 
 
 
Figure 4. 10: Average temperature in the Lower Volta Basin 
 
  
Figure 4. 11: Temperature anomaly in the Lower Volta Basin 
y = 0.0292x + 31.876
R² = 0.1366
24.0
26.0
28.0
30.0
32.0
34.0
36.0
1
9
6
0
1
9
6
1
1
9
6
2
1
9
6
3
1
9
6
4
1
9
6
5
1
9
6
6
1
9
6
7
1
9
6
8
1
9
6
9
1
9
7
0
1
9
7
1
1
9
7
2
1
9
7
3
1
9
7
4
1
9
7
5
1
9
7
6
1
9
7
7
1
9
7
8
1
9
7
9
1
9
8
0
1
9
8
1
1
9
8
2
1
9
8
3
1
9
8
4
1
9
8
5
1
9
8
6
1
9
8
7
1
9
8
8
1
9
8
9
1
9
9
0
1
9
9
1
1
9
9
2
1
9
9
3
1
9
9
4
1
9
9
5
1
9
9
6
1
9
9
7
1
9
9
8
1
9
9
9
2
0
0
0
2
0
0
1
2
0
0
2
2
0
0
3
2
0
0
4
2
0
0
5
2
0
0
6
2
0
0
7
2
0
0
8
2
0
0
9
2
0
1
0
2
0
1
1
2
0
1
2
2
0
1
3
Te
m
p
er
at
u
re
(◦
C
)
Years
Average Annual Temperature Mean Linear (Average Annual Temperature)
y = 0.0292x - 0.8023
R² = 0.1366
-5.0
-4.0
-3.0
-2.0
-1.0
0.0
1.0
2.0
3.0
1
9
6
0
1
9
6
1
1
9
6
2
1
9
6
3
1
9
6
4
1
9
6
5
1
9
6
6
1
9
6
7
1
9
6
8
1
9
6
9
1
9
7
0
1
9
7
1
1
9
7
2
1
9
7
3
1
9
7
4
1
9
7
5
1
9
7
6
1
9
7
7
1
9
7
8
1
9
7
9
1
9
8
0
1
9
8
1
1
9
8
2
1
9
8
3
1
9
8
4
1
9
8
5
1
9
8
6
1
9
8
7
1
9
8
8
1
9
8
9
1
9
9
0
1
9
9
1
1
9
9
2
1
9
9
3
1
9
9
4
1
9
9
5
1
9
9
6
1
9
9
7
1
9
9
8
1
9
9
9
2
0
0
0
2
0
0
1
2
0
0
2
2
0
0
3
2
0
0
4
2
0
0
5
2
0
0
6
2
0
0
7
2
0
0
8
2
0
0
9
2
0
1
0
2
0
1
1
2
0
1
2
2
0
1
3
Te
m
p
e
ra
tu
re
(◦
C
)
Years
University of Ghana http://ugspace.ug.edu.gh
80 
 
From figure 4.11 the average annual temperature anomaly shows negative and positive 
temporal patterns of change from the period 1960 to 1990. The period 1960 to 1962 
recorded positive values whereas there was a (-2.1○C) value in 1963. Then again a rise 
in 1964 and 1965. An increase and decrease trend is observed until 1990 when 
temperature becomes quite steady. The period 1971 to 1975 shows positive values with 
negative values before and after that period. So after the 1990s most years had positive 
values except for 2013 which had a -0.2○C value and 1999 which was very close to 0○C 
with a value of (0.1○C).  
 
4.4  Crop yield in the Ada East & Lower Manya districts 
Data on crop yields from Ada East and Lower Manya Krobo districts, are presented 
below. Cassava and maize were used because out of all the crop yield data collected 
from Ada East and Lower Manya Krobo, they are the only two crops which were grown 
in both districts. Figure 4.12 is shows crop production and area (ha) cultivated for 
cassava in the Ada East District for 2005 to 2013.  
 
Figure 4. 12: Cassava Production in Ada East 
0
500
1000
1500
2000
2500
3000
3500
4000
2005 2006 2007 2008 2009 2010 2011 2012 2013
0
5000
10000
15000
20000
25000
30000
H
EC
TA
R
ES
YEARS
YI
EL
D
 (
M
T)
cassava cropped area (ha)
University of Ghana http://ugspace.ug.edu.gh
81 
 
The figure 4.12 shows that the production/yield of cassava corresponds to the area 
cultivated. A correlation analysis reveals that there was a strong positive correlation 
(0.82) between cassava yields and area cultivated. 
Figure 4.13 also shows maize production in the Ada East for the same period (2005 to 
2013). A positive correlation (0.65) was recorded. It is observed in figure 4.13 also that 
production/yield had a corresponding relationship to area under cultivation.  The period 
of 2005 to 2009 showed a significant increase in area cultivated which corresponded 
with increase in yield (Mt/ha). However there was a maize production decline with 
cultivated area still increasing for the period (2011 to 2012). In 2013 there was an 
observed increase in yield as cultivated area also dropped.  
 
Figure 4. 13: Maize Production In Ada East 
 
 
 
 
0
50
100
150
200
250
300
350
400
2005 2006 2007 2008 2009 2010 2011 2012 2013
0
200
400
600
800
1000
1200
1400
1600
YI
EL
D
 (
M
T)
YEARS
H
EC
TA
R
ES
ha (maize) Maize
University of Ghana http://ugspace.ug.edu.gh
82 
 
Figure 4.14 and 4.15 also show yields of cassava and maize in the Lower Manya Krobo 
District  
 
Figure 4. 14: Cassava Production in Lower Manya Krobo 
 
In this district (Lower Manya Krobo), a similar relationship between crop 
production/yield and cultivated area is noticed as was in the Ada East district. A 
correlation coefficient of (0.63) was computed. It is therefore visually observed that 
crop yield increases with increased cultivated area and decreases with a decrease in 
cultivated area (see figure 4.14).  
0
2000
4000
6000
8000
10000
12000
14000
0
50000
100000
150000
200000
250000
2005 2006 2007 2008 2009 2010 2011 2012
H
EC
TA
R
ES
YI
EL
D
 (
M
T)
YEARS
Cassava Production (Lower Manya Krobo)
Cassava Cropped area (ha)
University of Ghana http://ugspace.ug.edu.gh
83 
 
 
Figure 4. 15: Maize Production in Lower Manya Krobo 
Again in figure 4.15, maize yields increases and decreases with an increase and 
decrease in cultivated area over the period. A correlation coefficient of (0.96) was 
computed, indicating a very high positive relationship between yield and area 
cultivated. A decrease in yield in 2006 followed by a steady increase from 2006 till 
2011 (peak) is observed.  
 
In the Ada East District, it is noticed in figure 4.12 and 4.13, that the yield of cassava 
and maize increase and decrease with the area of land cultivated. Furthermore, the 
Lower Manya Krobo Districts also experienced a similar trend in the relationship 
between crop yield and the area under cultivation. The yield per hectare for both crops, 
in both districts by year can be seen in Appendix 5 
 
  
0
2000
4000
6000
8000
10000
12000
14000
16000
2005 2006 2007 2008 2009 2010 2011 2012
0
5000
10000
15000
20000
25000
30000
35000
H
EC
TA
R
ES
YEARS
YI
EL
D
 (
M
T)
maize cropped area (ha)
University of Ghana http://ugspace.ug.edu.gh
84 
 
4.5  Crop suitability (EcoCrop Model) 
Below shows the crop suitability map of various crops (Maize, Pepper and Cassava) for 
current climatic conditions (1950-2010) and future climatic conditions, 2041-2060 
(~2050) in the Lower Volta Basin.  
The suitability map for both climatic periods was based on rainfall and temperature 
parameters only.  
 
Figure 4. 16: Crop suitability map for maize crop, current climatic conditions (1950-
2010). 
 
 
Maize crop (1950-2010) 
University of Ghana http://ugspace.ug.edu.gh
85 
 
 
Figure 4. 17: Crop suitability map for maize crop, future climatic conditions (~2050). 
 
A clear difference can be seen between Figure 4.16 and 4.17. Figure 4.17 shows the 
North Western part of the Lower Volta Basin having a wider coverage being very 
marginal, marginal and suitable as compared to Figure 4.16 which has almost the whole 
area being excellent for growing the maize crop. 
 
Figure 4.18 and Figure 4.19 show the crop suitability maps of the Lower Volta Basin 
for the pepper crop for both current and future climatic conditions respectively. 
Maize crop (~2050) 
University of Ghana http://ugspace.ug.edu.gh
86 
 
 
Figure 4. 18: Crop suitability map for pepper crop, current climatic conditions (1950-
2010). 
 
 
Figure 4. 19: Crop suitability map for pepper crop, future climatic conditions (~2050). 
 
Figure 4.18 shows that the Lower Manya Krobo (north western side) area has a small 
portion being marginal. Comparing it to Figure 4.19, which shows future suitability 
University of Ghana http://ugspace.ug.edu.gh
87 
 
conditions it is found that some areas have become non suitable for crop production. 
The area being excellent for current climatic conditions (Figure 4.18), has reduced for 
future climatic conditions as shown in Figure 4.19. However a very large portion of the 
LVB still remains excellent for pepper crop production based on rainfall and 
temperature parameters. 
 
Figure 4.20 and 4.21 also shows crop suitability maps for the cassava crop. Figure 4.20 
shows that the current climatic parameters provide suitable and excellent conditions for 
the cassava crop. Figure 4.21 also shows that the future climatic parameters provide 
excellent conditions for cassava. There is no much difference for the cassava crop for 
both current and future periods as compared to the other two crops (Figures 4.16, 4.17 
and 4.18, 4.19). 
 
 
Figure 4. 20: Crop suitability map for cassava crop, current climatic conditions (1950-
2010). 
 
 
University of Ghana http://ugspace.ug.edu.gh
88 
 
 
Figure 4. 21: Crop suitability map for cassava crop, future climatic conditions 
(~2050). 
 
  
University of Ghana http://ugspace.ug.edu.gh
89 
 
4.6  Socio Economic Study 
4.6.1 Background information from respondents 
This section presents the outcome of interviews and questionnaire administered to 
individuals (farmers) in communities within the study area. A total of 200 respondents 
from twelve (12) randomly selected communities namely; Dorgobom (20%), Tojeh 
(10%), Manaikpo (10%), Big Ada (10%), Bedeku (5%), Kasseh (5%), Dogo (5%), 
Korleykope (5%), Bueyonye (15%), Oborba West (5%), Yongase (5%) and Kabose 
(5%) were interviewed. The figure 4.22 illustrates the distribution of respondents in the 
study area.  
 
 
Figure 4. 22: Distribution of respondents in the study area 
With respect to sex of the respondents 67% were males whilst 33% were females. 
Figure 4.22 presents the age distribution of the respondents in the study area. Thirteen 
percent (13%) of the respondent were between the ages of 18-25 years, 17% and 38 % 
were between the ages of 26-35years and 36-45years respectively whilst 32 % were 
above 45 years at the time of the interview. 
Dorgobon
20%
Tojeh
10%
Manaikpo
10%
Big Ada
10%
Bedeku
5%
Kasseh
5%
Dogo
5%
Korleykope
5%
Buenyoye
15%
Oborba
5%
Yongase
5%
Kabose
5%
Distribution of respondents
University of Ghana http://ugspace.ug.edu.gh
90 
 
 
Figure 4. 23: Age groups of respondents 
 
The crops grown by farmers in the study area included vegetables such as pepper and 
tomatoes as well as cassava among others. Fourteen percent (14%) of the sampled 
population had had SHS/6th form education, Primary (33.5%), JHS (38%) with14.5% 
having no formal education at the time of the interview. The figure 4.24 illustrates the 
educational background of respondents in the study area.  
 
Figure 4. 24: Educational level of respondents in the study area. 
 
13.0
17.0
38.0
32.0
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
18-25yrs 26-35yrs 36-45yrs Above45yrs
P
er
ce
n
t
Age groups
14.5%
33.5%
38%
14%
0
5
10
15
20
25
30
35
40
None Primary JHS SHS/6th form
Level of education
University of Ghana http://ugspace.ug.edu.gh
91 
 
The household size of the respondents were grouped into three categories, 1-5, 6-10, 
and more than 10 people. Majority (57%) of the respondents had household size in the 
6-10 people category.  
Table 4. 6: Household size of respondents. 
Household size Frequency Percent 
1-5 34 17 
6-10 114 57 
10 and above 52 26 
Total 200 100 
 
With regards to land tenure system, 51% of the respondents owned their farm lands 
through inheritance from families whilst 37.5% are acquired their lands through lease. 
One hundred and twenty five the respondents owned their lands and 75 of them said 
they had rented the land. Out of this, 102 of them had inherited the parcels of land 
whiles 13 of them said they had bought them and 10 of them said they shared the land 
with others, thus group ownership. 
Table 4. 7: Land tenure type of respondents 
Land Tenure type Frequency Percent 
Inherited 102 51 
Bought 13 6.5 
Shared ownership 10 5 
Leased/Rented 75 37.5 
Total 200 100 
 
Table 4.8 shows grouped farm sizes and the number of respondents that have lands 
falling within these groups.  From the table 4.8, it is clear that majority of the farmers 
interviewed had farmlands with sizes ranging from 1.2 to 8.1ha. Very few had land over 
8.1ha.  
Table 4. 8: Respondents farm sizes. 
Farm size (Ha) Frequency Percent 
0-2.0 44 22.0 
2.4-4.0 60 30.0 
4.5-6.1 44 22.0 
6.5-8.1 47 23.5 
>8.1 5 2.5 
Total 200 100.0 
University of Ghana http://ugspace.ug.edu.gh
92 
 
Majority of the respondents have farmed these lands for 10 to 22 years, while a few 
(7.5%) had been farming for just less than 5 years. 
Figure 4.25 shows that most of the respondents that were interviewed had been farming 
for more than 10 years. A total of 148 respondents had been farming for more than 10 
years at the time of the interview. 
 
Figure 4. 25: Number of years respondents have been farming. 
 
Out of the 200 respondents, 159 of them said they had fallowed their farm lands for one 
of the following reasons; 
i. Because they had no labour to help cultivate their farm lands 
ii. So their lands could regain its soil nutrients. 
Moreover 49 of the respondents gave both of the above mentioned reasons which are 
lack of labour and to regain soil nutrients as the reasons why they fallow their farm 
lands. 
Table 4. 9: Reasons given for fallowing farm land. 
Reasons Total 
To gain soil 
nutrients 
Lack of labour Both Not 
Applicable 
97 13 49 0 159 
0 0 0 41 41 
97 13 49 41 200 
0
5
10
15
20
25
30
35
40
45
<5yrs 5-10yrs 10-20yrs 20-30yrs >30yrs
University of Ghana http://ugspace.ug.edu.gh
93 
 
Table 4.9 shows the number of respondents who practiced fallowing and the reasons 
why they did so. 
Twenty six of the respondents said they planted trees for the following reasons; 
providing shade on their farm land and for the purpose of harvesting the fruits for 
consumption.  
Moreover Azadirachta indica (Neem) and Acacia auriculiformis (Earleaf Acacia) trees 
are the two trees the respondents used for fire wood.  
Figure 4.26 shows the types of trees and the percentage of respondents who used these 
trees as fire wood.  
 
 
Figure 4. 26: Type of tree used as fire wood by respondents 
 
4.6.2   Population of the LVB 
The population (876,146) of the Volta Basin in 1960, rose sharply to 3,032,857 in 2000. 
It recorded an increase of 62% in the population between 1960 and 1970, compared to 
a 45% increase in 1970-1984; the period 1984-2000 experienced a 48% increase in 
population. Today the population calculated based on the districts that make up the 
LVB is 585,640. 
13%
22%
65%
Tree used as fire wood 
Acacia Acacia and neem tree Neem tree
University of Ghana http://ugspace.ug.edu.gh
94 
 
The Ada East district director explained, an increase in population over the years had 
contributed to the changes in land-use and land cover. He further mentioned that the 
influx of people in the district during the “Asafutufiami” festival which takes place in 
the month of August has led to the building of hotels, guest houses, restaurants and fuel 
stations in recent times. This he acknowledged as a form of land-use change, as the Ada 
area, especially Big Ada and Ada Foah is growing its coverage of built areas. The island 
communities are experiencing some cover changes as well in terms of clearing 
vegetation for farming and clearing of mangroves to be used as fuel wood and smoking 
of fish. 
 
4.6.3  Perception, Knowledge and Observed LUCC 
The interviews revealed that all respondents have knowledge on land-use/cover 
changes. All respondents were aware of changes taking place in their environment. 
They had observed land related changes such as increasing farmlands while decreasing 
vegetation cover and increasing infrastructure whiles decreasing vegetation cover. The 
observations made by the respondents with regards to land-use change shows that 
majority of them had observed an increase in farmlands and decrease in vegetation. 
52.5% of the then respondents agreed to the fact that they have observed increasing 
farm lands and decreasing vegetation cover as major land-use change. Figure 4.27 gives 
a representation of the people’s response to land-use changes.  
University of Ghana http://ugspace.ug.edu.gh
95 
 
 
Figure 4. 27: Land-use change as observed by respondents. 
 
116 farmers would seek to acquire land in a different area of the same district. Out of 
these 166, 123 would use it for farming while 43 claimed they would use it to build 
houses for themselves and for rent. 
 
According to the respondents, the major drivers of land-use change considered were; 
population, socio economic issues like (jobs and income) rainfall variability and soil 
condition. They ranked the drivers based on the knowledge they have on land-use/cover 
change in the chart (Figure 4.28).  
Land-use conflict is a challenge in the district. Lands on which people had their farms 
are being sold by family heads and chiefs to companies for purposes such as 
recreational centres. Furthermore the estuary had been sold to a real estate company 
which sought to develop hotel/housing units there. This he said has led to a 
disagreement between the people residing there and the land owners as they had not 
been relocated nor compensated (Source: District Director). 
52%
12%
34%
2%
0
20
40
60
80
100
120
Increasing farm lands
and decreasing
vegetation
Decreasing farmlands
and increasing
infrastructure
Increasing
infrastructure and
decreasing vegetation
Decreasing vegetation
and increasing bare
lands
N
u
m
b
er
 o
f 
R
es
p
o
n
d
en
ts
University of Ghana http://ugspace.ug.edu.gh
96 
 
  
Figure 4. 28: Drivers of land-use change 
 
Most respondents (54%) agreed that all four drivers were actually causing land-use 
change. However for the “socio economic” and “rainfall variability” drivers, none of 
the respondents disagreed. Further, a few respondents (10.7%) disagreed to population 
and soil condition as drivers of land-use change in the area. 
4.6.3.1  Statistical Analysis 
Regression analysis conducted to establish the causality between land-use change and 
drivers of land-use changes, revealed that there is a significant relationship between 
land-use change and Population, rainfall variability, socio-economic conditions and soil 
condition as shown by the global test of significance (F=21.5, P<0.05)(Table 4.5). This 
means population, poor soil condition, rainfall variability and socio-economic 
conditions jointly determines or influences the kind of land-use adopted. An adjusted 
R-Square of 0.794 obtained from the analysis means population, rainfall variability, 
soil conditions and socio-economic conditions jointly determines 79% of the variance 
in land-use change. 
However when the t-test was used to conduct an individual test of significance; 
population, rainfall variability and socio-economic conditions significantly influenced 
0
20
40
60
80
100
120
140
160
Population Rainfall Variability Soil condition Socio Economic
N
U
N
B
ER
 O
F 
R
ES
P
O
N
D
EN
TS
Strongly agree Agree Strongly disagree Disagree
University of Ghana http://ugspace.ug.edu.gh
97 
 
the type of land-use change (P<0.05) but soil condition did not (P>0.05) (see Appendix 
4). The regression analysis that was used to determine the drivers of land-use change 
ascertained from the socio economic data showed a significant relationship between 
land-use changes and the independent variables population, rainfall variability, socio 
economic conditions and soil condition. This analysis showed that there was 21% of 
other factors unaccounted for, which influenced land-use change. 
 
4.6.4  Knowledge and Perceptions on Climate Change 
Majority (64%) of respondents had knowledge in climate change whilst the remaining 
36% had no knowledge. The respondents level of knowledge on climate change were 
then sought to ascertain their level of understanding of issues related to climate change. 
36% of them had no knowledge, limited knowledge (53%) and 11% had some 
knowledge at the time of the interview. Figure 4.29 shows the knowledge levels of the 
respondents on climate change. 
 
Figure 4. 29: Knowledge level of respondents on climate change  
 
36%
53%
11%
No knowledge Minimal knowledge some knowledge
University of Ghana http://ugspace.ug.edu.gh
98 
 
All respondents had observed an increase in temperature. However 121 said they had 
observed a decrease in rainfall and 79 had observed a fluctuating pattern. 
According to the respondents, the earliest month in which they start planting is March. 
In early March, some of them grow watermelon in preparation for the planting major 
crops. Thus majority of the farmers start planting in March and April. Table 4.10 
illustrates this. 
 
Interestingly some older farmers, (32 of them) claimed they could tell if it was going to 
rain or not in the near future (days ahead) by looking up at how the sun appeared in the 
atmosphere. 
 
Table 4. 10: Month for planting crops  
Month Frequency Percent (%) 
April 52 26.0 
April and May 1 0.5 
June 21 10.5 
March 78 39.0 
May 48 24.0 
Total 200 100.0 
All respondents had changed the month in which they planted their crops in certain 
years. The major reason given for this change in planting time was because of the 
change in the onset of rains. Other reasons however included, not being able to purchase 
seeds and ploughing farmland (economic reasons).  
Even though there is a general planting month for the growing season every year, a fruit 
such as Citrillus lanatus sp. (water melon), which is mainly grown very close to the 
coastal areas has a different planting month. The Ada East district is the largest 
watermelon producer in the LVB. 
 
University of Ghana http://ugspace.ug.edu.gh
99 
 
Table 4.11 illustrates the month in which respondents considered beginning of the rainy 
season.  
 
Table 4. 11: Month in which rains begin (respondents view) 
Month Frequency Percent 
February 29 14.5 
March 62 31.0 
April 46 23.0 
April and May 2 1.0 
May 56 28.0 
June 5 2.5 
Total 200 100.0 
 
The respondents claimed that the onset of the rainy season was the months of February, 
March, April and May. 
 
Almost all of the respondents affirmed, that there was a change in the length of rainy 
season. 116 said it was decreasing, while 84 of them said it was fluctuating. Further, 
out of the 181 who responded yes to a change in their crop yield, 126 had experienced 
a decrease in their yield whilst 55 had experienced a fluctuating nature of crop yield 
over the past 10 years. The reasons the respondents gave for the changes in their crop 
yield included low soil fertility, less rain, diseases and application of fertilizers.  
The farmers who claimed had experienced a decrease in yield all said the changed 
rainfall pattern and quantity was a major contributing factor. Other farmers who had 
experienced a fluctuating nature claimed to have gotten high yields in certain years due 
to the application of chemical fertilizer. 
 
Figure 4.30 shows, the respondents view on how rainfall and temperature have affected 
their crop yield and soil condition.  
University of Ghana http://ugspace.ug.edu.gh
100 
 
 
Figure 4. 30: Respondents view on effects of rainfall and temperature on their yield 
and soil condition 
 
 
 
Figure 4. 31: Factors that influence farmer crop yield  
 
110 respondents had received professional advice from agricultural extension officers. 
Table 4.12 shows the interventions proposed by agricultural extension officers to the 
respondents.  
 
0
20
40
60
80
100
120
140
160
180
200
Reduced crop yield increased crop
yield
not changed my
yield
affected soil
negatively
affected soil
positively
FR
EQ
U
EN
C
Y
TRUE FALSE Not applicable
Less rain
16%
Application of 
fertilizer
17%
Soil fertility
5%
Less rain, temp & 
soil fertility
37%
Disease
3%
Soil fertility & Less 
rain
8%
Less rain, temp, soil 
fertility & disease
14%
Less rain Application of fertilizer
Soil fertility Less rain, temp & soil fertility
Disease Soil fertility & Less rain
Less rain, temp, soil fertility & disease
University of Ghana http://ugspace.ug.edu.gh
101 
 
Table 4. 12: Advice given to respondents by agricultural extension officers 
Advice Frequency Percent 
Mulching 66 33.0 
Proper use of fertilizer 30 15.0 
Use of drought resistant seeds 72 36.0 
Make sunken beds 26 13.0 
Construction of reservoir 6 3.0 
Total 200 100 
 
The table (4.12) shows that the most frequent advice given to farmers by extension 
officers are mulching and the use of drought resistant seeds.  
158 farmers had made changes to their farming methods.  
Figure 4.32 shows the different methods adopted to help reduce vulnerability to 
changing rainfall and temperature.  
 
Figure 4. 32: Farmer adaptation methods to changing rainfall and temperature 
The use of drought resistant seeds appears to be the most used method amongst the 
other methods the respondents engaged in. Some of the drought resistant crop varieties 
grown include “pectomech” and “nema giant” for tomato, “mamaba” for maize, 
“bankye hema” for cassava pepper and tomato. 
8%
58%
8%
6%
20%
Mulching Use od drought resistant seeds
Application of organic manure Frequent ploughing of the land
Application of chemical fertilizer
University of Ghana http://ugspace.ug.edu.gh
102 
 
4.6.4.1 Statistical Analysis on relationship between farmers knowledge of climate 
change and adaptation measures. 
The chi square test was used to establish the relationship between the farmers 
knowledge on climate change and the adaptation measures or changes they have made 
to their farming practices due to change in rainfall and temperature. The test revealed 
that there was no significant relationship (chi square =1.27, p>0.05).  
  
University of Ghana http://ugspace.ug.edu.gh
103 
 
CHAPTER FIVE 
DISCUSSION 
Studies on Land-use/cover change as well as climate conditions and its relationship to 
crop productivity and crop suitability was conducted in the Lower Volta Basin to 
provide additional information which could help solve challenges and bridge the 
knowledge gap in studies conducted in the LVB. This chapter presents further insight 
into the results obtained from this study and a linkage to other studies conducted in the 
LVB and Volta Basin as a whole. 
 
5.1  Land ±Use/Cover Change of Lower Volta Basin 
The land cover type of the Lower Volta Basin is generally savannah shrub characterized 
by many "neem" trees, shrubs and coastal savannah lands (Koku, 2002). This naturally 
has associated to it an interesting ecosystem. From this study, the land-use types that 
can be found in this area today includes grazing lands (pasture), farmlands (rain fed and 
irrigated), built-up areas which include houses, schools, hospitals, hotels guest houses, 
a few industries, roads lorry stations and markets.  
 
Despite the changes that have taken place in the Lower Volta Basin, these land-use 
changes have remained the same. Though there has been an increase in some and a 
decrease in some of the land-use types. According to Andah et al., (2003), the main 
land-use of the Lower Volta Basin is short bush fallow cultivation along the banks of 
the river, animal grazing and fishing. They also mentioned of how the cutting of wood 
had become an extensive economic activity in the southern transitional zone.  
 
University of Ghana http://ugspace.ug.edu.gh
104 
 
Using remote sensing techniques to develop land-use classification mapping for this 
study was useful and presented a detailed way to improve the selection of areas 
designed to agricultural, urban and/or industrial areas of a region as reported in (Rai et 
al., 2011). Despite the different methods for classifying satellite imagery, this study 
employed the maximum likelihood classifier (supervised classification) method. 
According to Stemmler and Su (1993), it is a primary algorithm for supervised 
classification of image data and offers best results. While some researchers such as Tutu 
(2008) and Plantier et al., (2006), concluded in their study that unsupervised 
classification had demonstrated improved accuracy of land-use/cover classification, 
others (Richards, 1995; Akgun et al., 2004), have also claimed that the maximum 
likelihood method is found to be more applicable and reliable for (especially Landsat) 
satellite image classification purposes. In this study, the maximum likelihood method 
under the supervised classification system produced very good results. In a similar 
research by Duadze (2004), in the Volta Basin of Ghana, supervised classification 
proved to derive accurate land-use and land cover maps for the savannah ecosystem in 
the Volta Basin of Ghana. Duadze went on to say that the maximum likelihood 
classification of LANDSAT images for a large area can be done successfully with high 
level accuracy.  
 
However in classifying the satellite images for this study, the ISO data unsupervised 
classification method was employed  prior to field visit, in order to determine strata for 
ground truth (Torahi and Suresh, 2011) and to analyse also, how the mixed pixels of 
certain features in the image would best fit when combined to other land-use features.  
 
University of Ghana http://ugspace.ug.edu.gh
105 
 
NDVI is mostly used as a greenness parameter which indicates biophysical activity of 
a land surface (Hansing and Wittich, 1995). It is well known that there is a direct 
correlation between NDVI and the amount of stress vegetation experiences (Travis, 
2006). Research conducted by Travis reveals that NDVI could be used to determine 
drought effects on specific land use/cover types.  
 
After classification was done a statistical analysis was run to check the accuracy of the 
classified image to the original Landsat image. Important features such as farmlands 
and built up-areas had 68.9% and 69.5% producer accuracy respectively and 93.65%, 
71.71% for farmlands and built-up areas user accuracy respectively. Duadze (2004), 
indicates that, an overall statement of accuracy is useful, but it does not say anything 
about the accuracy of the classified image as a whole and thus the need to look at the 
producer and user accuracies. Gumma (2010), reported that in African countries, it is 
generally difficult to map land-use and land cover classes because there are no large 
agricultural delineations in agriculture forms. Implying that agriculture and other land-
use and land cover areas are very similar because of the nature of small-holder 
agriculture which dominates. Furthermore, in Ghana and most parts of Africa, a single 
plot of land is usually occupied by different land-use types. This usually gives rise to 
mixed pixels. It is for this reason that ground truthing and supervised classification 
methods are suitable for this kind of study.  
 
Several change detection methods have been developed using conventional image 
differencing (consider change in reflection), using image ratio, normalized difference 
vegetation index, principal component analysis, multi-date image classification, post-
classification comparison, manual onscreen digitization (El-Raey et al., 1999). In a 
University of Ghana http://ugspace.ug.edu.gh
106 
 
comprehensive study conducted by Civco et al. (2002), it was concluded after studying 
all the change detection methods, that there is no single best way in which to perform 
change analysis. They further suggested that for any method being used, the accuracy 
depends heavily on each of the input maps. In this study the main method employed to 
identify land-use changes in the Lower Volta Basin was the image differencing and 
post classification comparison method. However Muttitanon and Tripathi (2005), 
reported that the most common approach used for change detection is digitizing and 
overlaying to compare the data sets from two dates in order to detect land-use changes 
of the first date data that have been replaced by the new land-use in the second date. In 
Turner et al. (1993), it was made clear that there are two forms of land-use/cover 
change. This they said was conversion and modification. In this study the change 
detection analysis was specifically focused on land-use/cover features that had been 
converted or replaced by another land-use/cover type rather than changes that affected 
the character of land cover and land-use type.  
 
The change detection analysis in this study revealed that, farm lands had increased. 
Implying that farming activity and/or the surface area of both rain fed and agricultural 
activities had increased over the time period studied. Increasing farmlands also meant 
that some type of land cover was being converted. Crop data obtained from Lower 
Manya Krobo and Ada East districts revealed that indeed the area cropped per farmer 
increased over the years (Figures 4.12, 4.13, 4.14 and 4.15).  This study showed that 
grass lands, mixed savannah and some burnt areas were changed into farming fields. 
This meant that grass lands and other sparsely vegetated areas had decreased as they 
had been converted to farmlands. On the contrary it should be noted that farmlands, 
when abandoned, could possibly be converted into grasslands or bushes as reported by 
University of Ghana http://ugspace.ug.edu.gh
107 
 
(Shindler, 2009). There is also no doubt that though the bush fallow system is still 
practiced in the area. An abandoned farm field could remain bare due to environmental 
conditions such as topology as well as changing rainfall patterns. Mantey and Tagoe, 
(2013) in their study, showed that there had been a decrease in sparse vegetation from 
3,604.71km2 (1985) to 2,784km2 (2013) around the Lower Volta River Basin which is 
similar to results obtained for mixed savannah vegetation in this study. 
 
With regards to burnt areas, the Lower Volta Basin is characterized by a random 
increase and decrease of burnt areas annually. This is mainly because of the herders 
who burn grass lands so as to gain fresh grass for their cattle to feed on and farmers 
who burn sparsely vegetated areas to prepare land for farming. So even though some 
burnt area has been lost to farmlands, it should be noted that these burnt areas keep 
increasing and decreasing. Bush burning is used to clear land for agricultural purposes, 
hunting, creating fire belts at the onset of the dry season, and inducing rapid re-growth 
of rangeland during the dry season, which often results in enormous damage to 
vegetation, wildlife, and properties because it is typically are not controlled (Boubacar, 
2005). Furthermore, increasing temperatures and dry spells as recorded from the 
climate data obtained could influence increased burnt areas and bare lands. 
 
The UNEP-GEF (2010), reports that regular burning, grazing of livestock and 
cultivation have resulted in the survival of relatively few trees. The vegetation in these 
areas is thus quite open and is dominated by short grasses. Andah et al. (2003), 
mentioned that annual bush burning takes place since grazing lands are poor. Forests 
and protected areas (densely vegetated areas) have reduced according to the change 
detection statistical analysis.  
University of Ghana http://ugspace.ug.edu.gh
108 
 
From the classified image (Figure 4.3), it is evident the forested area in the Kpong area 
has reduced. The protected area in and around the Shai hills which is quite a distance 
from the river can be said to have been maintained quite well. However a careful look 
at (Figure 4.3) shows that some bare and burnt lands have developed close to the forest 
reserve. Development of land by putting up structures is one land-use type which has 
also increased in the LVB. This is due to an increase in population and commercial 
activities in certain major towns like. Kpong, Aveyime, Akuse, Tefle and Big Ada. 
Barry et al. (2005), reported that urban land-use is limited to a few towns like Kpandu, 
Kwamekrom, Akuse, Sogakope and Ada Foah, however this study’s results shows that 
other towns in the Lower Volta Basin including Kpong, Tefle, Kasseh and Big Ada are 
growing in size. Urbanization can be ranked as one of the well-known LUCC. However, 
since urban areas occupy a relatively small fraction of the earth’s surface (Gruebler, 
1994), it may lead to a misconception that the growth of urban areas can be ignored in 
land-use studies (Heilig, 1994). Urbanization in reality affects land-use change 
elsewhere at large scale, especially through strong linkages between urban and rural 
areas (Lambin et al, 2001). 
 
Since farming is the main economic activity of the people in this region, majority of the 
people are farmers and would farm on more available lands if they have the means.  
Recent developments in land use, particularly below the Akosombo dam include 
irrigated rice, sugar cane and vegetable cultivation in the areas immediately adjoining 
the Volta River (Andah et al, 2003). They further proclaimed that, with increasing 
population and growing demand for food, the economy can no longer depend on rain 
fed agriculture. For the past twenty years Ghana has experienced drought periods and 
erratic rainfall. Irrigation is therefore the way forward in agriculture if the country is to 
University of Ghana http://ugspace.ug.edu.gh
109 
 
solve the food security problem. This study confirms that rain fed agriculture is not 
sustainable and thus the needed shift to irrigation to help keep and improve upon crop 
production. 
 
Moreover, private companies, such as Golden Exotics Co. ltd. have taken over lands 
from the locals and established large scale irrigated plantations. Even though local small 
scale farmlands have been overtaken by theses private companies, employment has 
been created for the people (locals) in the area. The government has invested in large 
scale irrigated rice farms which is leased to local farmers to help them improve upon 
their livelihood and reduce poverty. However this is arisen due to the lack of capacity 
of the local farmers to engage in large scale irrigated farming. Even though irrigation 
development in sub-Saharan Africa has remained well below its physical potential, 
international donors have shown renewed interest in irrigation investments due to the 
far reaching benefits of irrigation farming, several investments in sub-Saharan Africa 
have been directed towards irrigation development. These investments have been 
driven by government policies (colonial and postcolonial), multinational donor 
agencies, private investors, markets, and by technological innovations such as, drip, 
motorized pumps and treadle pumps (Ofosu, 2011). 
 
Further downstream, towards the Big Ada area to Ada Foah, farming activities have 
been affected. (Agbenyo 2009), says the hydropower schemes have made farming more 
difficult and expensive in the study communities. This problem was attributed to the 
seizure of regular seasonal floods that deposited silts on farmlands and wetted the 
farmlands for easy cultivation on the floodplains. Farmers in the communities now 
University of Ghana http://ugspace.ug.edu.gh
110 
 
maintain soil fertility through application of chemical fertilizers and also hire the 
services of tractors for tilling farmlands.  
 
In a study by Ofosu (2011), in the Upper East region of Ghana, it was reported that the 
major contributor to the expansion of irrigation in the study area is the rising demand 
for vegetable products in the urban centres of southern Ghana. This is very true as 
almost all the vegetable farmers in the Ada East area especially of the LVB claim to 
produce vegetables for the city (Accra). A social survey conducted in this study 
confirms that there is a high demand for vegetables in the city. This is one of the major 
reasons (high demand) why there is an upsurge of vegetable irrigation farming in the 
Ada Foah township (see plate 4.8, 4.9 and 4.10). 
 
Furthermore Ofosu (2011), points out that there are three major sources, of vegetables 
(especially tomatoes) for urban centres in Ghana (coastal area, middle-belt and Upper 
East Region). The increasing rate of urbanisation in Ghana, coupled with increasing 
population has both contributed to the increasing demand for vegetables (Ofosu, 2011). 
 
Interestingly, according to Rosegrant et al., (2001) and Hussain (2005), in areas such 
as South-East Asia and East Asia investments in irrigation have yielded significant 
impacts in terms of improving food security and poverty reduction, however the same 
cannot be said for sub-Saharan Africa. The FAO (2006), reports that Regions such as 
South-East Asia have almost exhausted their irrigation development potential, making 
the potential irrigable land in Sub-Saharan Africa a major hope for the world in terms 
of feeding the future population. 
 
University of Ghana http://ugspace.ug.edu.gh
111 
 
5.2  Incidence and Trend of Climate 
"Similar to the Sahel and Mid-Ghana, the Volta Basin have experienced significant 
downward trending rainfall since the 1970s" (Owusu, 2009). Owusu further mentions 
in his research that the southerly areas of the Volta Basin have experienced shifts in 
their bimodal regimes. He says, the dry spell (July-August) that separated the two rainy 
peaks has experienced increases in both the average total rainfall and the number of 
rainy days. And he goes on to discuss how food security could be threatened and 
household incomes eroded due to declining rainfall in nations of West Africa who have 
their economies being highly dependent on rain.  
 
It is known that a country such as Ghana, in the area of agriculture provides a high 
percentage of employment but with declining rainfall totals there is higher risks for 
rain-fed agriculture.  
The analysed rainfall data in this study is similar to what Lemoalle and Condappa 
(2010) reported. They reported that “The decades of the 1950s and 1960s were 
relatively wet, while the 1970s and 1980s were drier”. In Owusu (2007), there was a 
general observation of declining mean annual rainfall and the zone B in which the LVB 
falls was reported to have seen a reduction from around 1200mm to about 1000mm.  
 
It is to be noted however that, land surface changes can affect local precipitation and 
temperatures. Vegetation patterns and soil composition can influence cloud formation 
and precipitation through their impact on evaporation and convection (the rise of air) 
(De Sherbinin, 2002). 
 
University of Ghana http://ugspace.ug.edu.gh
112 
 
5.3  Perception and Knowledge on Land Use 
With all the respondents having observed changes in land-use/cover in their vicinity, 
majority of them had observed an increase in farmlands to which they had contributed 
to. The figure (4.23) shows clearly that majority had observed an increase in farmlands 
and decreasing vegetation. This suggested that many more lands were being cleared for 
cultivation of crops. 
The major drivers of land-use change considered included population, socio economic 
status (poverty/job), rainfall variability and soil conditions. The study revealed that the 
farmers did not consider poor soil quality as a major cause or drive land-use change in 
the area. According to Braimoh (2004), soil quality which is a biophysical variable, 
does not directly influence land cover/ use change but can influence land-use decisions. 
For instance a decline in soil fertility might cause farmers to move to uncultivated areas 
to farm. Rainfall variability and Socio economic factors were considered the highest by 
the farmers as the most influencing drivers of land-use change. This is not surprising as 
the main economic activity of the LVB is agriculture (rain fed) and majority of the 
population are into farming. 
 
5.4  Perception and Knowledge on Climate Change 
Normally the rainy season begins in May/June with some rains earlier in April. There 
is however a dry spell in August. The rainy season continues from September till the 
month of November. In the south, rainfall is distributed bi modally with long rains from 
April to July and short rains from September to November. The risk of within-season 
dry spells influences cropping choices (Lemoalle and Condappa, 2010). Owusu (2009), 
reported in his study that there are striking differences (including onset of rains, 
University of Ghana http://ugspace.ug.edu.gh
113 
 
intensity of rains and duration of rains) in the rainfall regimes and the regional strengths 
of the causal mechanisms of variability in the Sahel and the Guinea coast regions.  
 
The uni-modal regime of the Sahel seems to have suffered declines throughout the short 
rainy season (June-September), while the more southerly areas have experienced shifts 
in their bi-modal regimes. The short dry spell (July-August) that separated the two rainy 
peaks has experienced increases in both the average total rainfall and the number of 
rain days. The response from farmers showed that the earliest planting month was in 
March with some planting in April and others in May and June.  
 
All farmers involved in the survey reported that they change the month in which they 
plant sometimes. Some of the farmers planted early because they claimed they had 
noticed the rains starting early during the year in recent times. However many also 
planted (nursed) early in the month of March and April and transplanted the seedlings 
to their farms during the main rainy season month (June). The farmers interviewed also 
believed that the month of the start the rainy season had changed from early March to 
late March and April, with some others saying May. Recently, detailed atmospheric 
modelling over the region shows that in the near future too, the onset of rainy season 
will shift to later periods in the year, roughly from April towards May (Van de Giesen 
et al., 2010). Jung and Kunstmann (2007), also predicted that the change in the onset 
of rainy season may continue to move forward. Furthermore Laux et al. (2007), used 
principal component analysis to show that there is, indeed, a statistically significant 
shift forward in several components of 0.4–0.8 days/year. 
 
University of Ghana http://ugspace.ug.edu.gh
114 
 
It is clear that rainfall variability in recent times is making it difficult for farmers to 
develop a systematic means of growing crops with regards to planting (Van de Giesen 
et al., 2010). It was agreed by 58% of farmers that there was a decrease in the length of 
the rainy season whilst 42% said they had observed a fluctuating trend.  
 
5.5  Crop Suitability 
This study presents the first use of the EcoCrop model to map crop suitability in the 
Lower Volta Basin and Ghana as a whole. It has however been used in different 
countries to map out suitable areas for growing crops based on climatic factors. In a 
study by Parthasarathy et al. (2008), the EcoCrop model was used to map out suitable 
areas for growing ginger in some states (Orissa, West Bengal, Mizoram and Kerala) of 
India. Literature reveals that most of the studies conducted using the EcoCrop model 
have been done using large geographical areas such as states and especially countries.  
 
The model maps presented in Chapter 4 of this study reveal how future projected 
rainfall and temperature are going to affect areas suitable for growing crops (maize, 
cassava and pepper). For instance the Zea mays (maize) crop is suitable in almost the 
whole area of the LVB under current climatic conditions, but projected future 
conditions renders some areas marginal and very marginal (see figure 4.18 and 4.19). 
In each of the suitability maps, the coastal areas are shown to have excellent conditions 
for crops to grow. However it is to be noted that this model makes use of climate 
parameters only and thus does not take soil into consideration. Moreover the EcoCrop 
model provides an opportunity for agriculturist, spatial analysts and policy makers to 
analyse the effect of climate change on crop productivity to inform development of 
adaptive strategies and policy to minimize the negative impact of climate change on 
University of Ghana http://ugspace.ug.edu.gh
115 
 
crops (Yumbya et al., 2014). Yumbya further stated that this model provides 
opportunities for further potential research to fine tune the model and communicate 
model predictions better. 
 
The DIVA GIS programme was used in a study on model comparison for the 
distribution of crops by Villordon et al., (2006). It was reported in their study, that 
detailed information on the geographic distribution of a crop is important in planning 
efficient conservation strategies, but is often not available, particularly for minor crops. 
The modelling of species distribution has become an important tool in biogeography, 
evolution ecology, conservation and invasive species management (Anderson et al., 
2003 and Williams 2005). 
 
5.6  Socio Economic Study  
Agriculture is the main economic/livelihood in the Lower Volta Basin just as in the 
whole Volta Basin (Codjoe, 2004). It was noticed that about 86% of the farmers were 
educated. 
 
The population of the main Volta sub-Basin more than doubled between 1960 and 1984, 
a period of 24 years, and had been projected to reach 4,268,927 in 2010, based on annual 
inter censual growth rate of 2.5% during the period 1984-2000 (Codjoe 2004). In 
Mantey and Tagoe (2013), it was also stated that according to Ghana Statistical Service, 
the population in the Basin is expected to increase by 80% within a twenty-five year 
period (2000-2025) due to the high average population growth rate of 2.54%. Lemoalle 
(2007) said, the combination of demography and climate change defines new pressures 
on the environment. This sets the scene for an evolving and increasing food demand, 
University of Ghana http://ugspace.ug.edu.gh
116 
 
and for the necessary adaptation of the agricultural production, productivity and water 
uses. 
 
Household size mostly fell within 6-10 based on the survey conducted. The traditional 
rural settlement of Ghana usually has a larger household size as compared to urban 
settlements. A number of reasons, including the need to get more hands to work on 
farming fields as well as the extended family system (living together with relatives) 
leads to an increase in household size. A study conducted in the Volta Basin of Ghana 
by Braimoh (2004), showed that a change in house hold size influenced land-use change 
(farm size), confirming that land–use change is driven by increase in demand for food. 
 
Land ownership in the LVB is usually passed on from one generation to the next as 
most of the land is owned by families/clans and Chief’s. From the survey, 51% of 
respondents had inherited their lands. Furthermore, it was noticed during the field 
survey, through interaction with the local farmers and what was captured in the 
questionnaire interview that most of them did not own just one parcel of land. The 
reason being that the most of the land in the study area belongs to families and chiefs 
which have been passed on from generation to generation. This means families have 
very large plots of land. The availability of land as compared to the population tells 
why one family could have very large (larger than 10 hectors) plots of land. Farmers 
therefore have parcels of land in different locations.  
 
Most of the farmers interviewed had not less than 3 acres (2.1ha) for a parcel of land. 
With such large tracts of land, some of the farmers mentioned that one of the reasons 
for leaving their land to fallow is because of lack of labour to cultivate the fields. Also 
University of Ghana http://ugspace.ug.edu.gh
117 
 
the survey revealed, that all those who had not fallowed their farming fields before are 
those that had less than 5 acres of land. 
 
Field observations during field visits supported by responses to questions reveal that 
the Azadirachta Indica (neem tree) is very dominant in the LVB. It grows everywhere 
and could be considered the most dominant tree species in the LVB. In this region most 
thickets are made up of this kind of tree. However there has been very little mention of 
this in literature as most studies and research are much more concerned with the 
economically valuable trees such as Shea and Cocoa. Acacias were also another species 
the local people fetch for firewood. Most of the people did not plant trees, however 
efforts have been made through various programmes such as, Project for Improving 
Water Governance in the Volta Basin (PAGEV) and the Greening Basin Initiative 
(GBI) to encourage tree planting.  In some other parts of the Volta Basin, Barry et al 
(2005), reported of an afforestation project funded by the VRA, with the aim of 
encouraging fishermen to plant wood lots and adopt the use of improved stoves, which 
use less fuel wood for fish smoking. Almost all of these programmes have not been 
sustainable as bush burning and animal grazing had been identified as major causes of 
unsuccessful tree planting projects (Okine and Nyarko, 2008). 
 
5.7  Agriculture In the LVB 
Many of the interviewed farmers had experienced a decline in crop yield over a 10 year 
period whilst others said they had experienced an increase and decrease trend.  
From the figure 4.21, it is obvious that there wasn’t one person who claimed that rainfall 
and temperature had not affected their yield. Figure 4.22, shows the various factors the 
farmers said affected their crop yield. From this chart it is noticeable that a lot more 
University of Ghana http://ugspace.ug.edu.gh
118 
 
farmers didn’t think rainfall only was the reason for the change in their yields. Moreover 
the “application of fertilizer” factor was considered the best intervention method by 
those who had experienced an increase in crop yield. 
 
In each district of the LVB there are agricultural extension officers who are however 
usually not adequate in terms of number. Field visits revealed that the extension officers 
in the Lower Manya Krobo district and Ada East district were inadequate in number as 
compared to the number of farmers in the district. The extension officers, educate, and 
aid the farmers with challenges they face to help achieve higher yields.  More than half 
of the farmers engaged in the use of drought resistant seeds which cost a little higher 
than the normal seeds. The farmers claimed the extension officers encouraged them to 
use these seeds. Some other farmers explained, ploughing the field helped in keeping 
the soil moist for a longer period as run off was reduced. Furthermore some of the 
farmers said they could not afford to buy fertilizer and drought resistant seeds.  
 
Moreover from this study, farming practices adopted by farmers due to changes in 
temperature and rainfall patterns does not depend on their knowledge of climate change. 
Even though recent research has suggested a possible shift to groundwater for 
agriculture because of the unreliable nature of rainfall (Masiyandima and Giordano, 
2007; Tuinhof et al., 2011), efforts are being made to maintain crop yields and reduce 
poverty.  
 
Public and private owned irrigation systems have been setup in the Ada East district.  
The Ada small farm irrigation scheme was established to provide small holder farmers 
on the stretch between Kasseh and Big Ada with an alternative source of water for 
University of Ghana http://ugspace.ug.edu.gh
119 
 
farming. Moreover, as in most other regions of Africa, the irrigated area in the Volta 
Basin is only a small fraction (<0.5%) of the total cultivated area, and is also only a 
small fraction of the irrigation potential (2%) Lemoalle and Condappa, (2010). 
Furthermore, the authors pointed out that even though the areas for which irrigation 
water is available from the Akosombo and Kpong dams, they have not been fully 
exploited. It has become clear from this study that farmers do not have any laid down 
adaptation measures put in place to address the issue of changing climate and thus 
practice what the extension officers tell them.  
 
It is projected that agricultural production and access to food in many African countries 
would be severely affected (UNFCCC, 2013). The northern region of Ghana has started 
experiencing this phenomenon and if nothing is done about it, food security would 
seriously be affected and the problem of malnutrition would be exacerbated Source: 
(https://mofafoodsecurity.wordpress.com/food-security-situation-in-ghana). 
 
It is an issue of major concern (Obeng, 2000) that agricultural production in Ghana is a 
terrible recurring cycle of poor farmer incomes irrespective of production outcomes. 
Thus when the weather is good and output levels are high, gluts force prices down and 
farmers are confronted with low incomes. On the other hand, when bad rains lead to 
low production outcomes, there is nothing to sell and production outcomes are limited 
to subsistence. 
 
According to the Ghana Statistical Survey (GSS) (2014), a little over half (51.5%) of 
households in Ghana own or operate a farm. Even though, farming is mostly rural and 
University of Ghana http://ugspace.ug.edu.gh
120 
 
engages about 83% of rural household, it is much common in rural savannah areas with 
about 93% of households involved. 
 
Research by Sivakumar and Hatfield (1990), showed that rainfall is spatially 
heterogeneous, and so farmers scatter their fields around their vicinity to minimize risk. 
Field visit, interaction with the farmers as well as the classified satellite imagery 
confirmed what Sivakumar and Hatfield had stated in their study. Granaham (2012), 
reported that since agricultural practice is dependent on the availability and distribution 
of rainfall, farmers suffer significant loss when rain fails. The Volta Basin is also noted 
for livestock production as it happens to be almost entirely within the savannah 
grassland belt of the country. The natural grass serves as grazing fields providing food 
for cattle, sheep and goats (UNEP-GEF, 2010). GSS also informs that livestock rearing 
is concentrated in the rural savannah where 86% of drought animals can be found. Out 
of this, 63% they said, are cattle.  
 
A discussion with the older farmers of the Dorgobom community revealed that the area 
(Dorgobom) used to be the hub of grazing fields for cattle. One of the farmers explained 
that as a teenager he could clearly recall how cattle came all the way from as far as 
Ningo Prampram to graze in the Dorgobom area and neighbouring communities. 
According to the older farmers the district agricultural office was setup there in 
Dorgobom.  
 
Regardless of the rainfall variability and trend in both districts, yields are kept generally 
high. Apart from the large cropped area contributing to increased yields, Sagoe (2006), 
said, improved farming technologies which includes planting more than two types of 
University of Ghana http://ugspace.ug.edu.gh
121 
 
varieties of root crops and planting improved varieties that are drought tolerant help 
keep yields high. Also Lemoalle and Condappa (2010) claim, fertilizer and small-scale 
irrigation can improve the yields of rain fed crops and alleviate poverty. It is to be noted 
that rainfall is relatively abundant in the Lower Volta Basin, however the temporal and 
spatial distribution of rainfall influences agriculture more than total rainfall (Lemoalle 
and Condappa, 2010). 
 
The maize crop follows a very similar trend. A direct relationship between maize 
production and area cultivated is observed figure 4.13. Area cultivated for maize was 
generally higher in Lower Manya Krobo for the period (2005-2013) as compared to 
Ada East. The highest recording yield for the said period in Ada East was 25,387.2 in 
2009 and 28,887.1 in 2012 for Lower Manya Krobo district. Ada East district recorded 
a decline in yield for both crops (cassava and maize) in 2012. It was also the lowest 
recorded yield for the period (2005-2013). Lower Manya Krobo, also experienced a 
sharp decline in yield as well as cultivated area, for both crops in 2013. This is because 
the district was split into upper and Lower Manya Krobo.  
 
According to Terrasson et al. (2009), simulation modelling showed that fertilizer (N40, 
P13 kg/ha) can increase yield potential of most soils of the Volta Basin. Moreover 
Lemoalle and Condappa (2010), reported that if the climate trend remains constant, the 
greater production of food required cannot be achieved only by increasing area of 
cultivation. Furthermore an increase in area is not always in proportion with the 
increase in production (Parthasarathy et al., 2008). An increase in input systems to 
maximize production to necessary levels will be needed.  
 
University of Ghana http://ugspace.ug.edu.gh
122 
 
From analysed climate data and crop yield it is established that rainfall amounts could 
be high for a particular year, however low yields. This could be due to reasons such as, 
the rains not falling during the growing season but rather outside of it or the rains falling 
heavily and for long periods during the growing season, thus destroying some of the 
crops due to flooding. Interactions with the agricultural extension officers and farmers 
confirmed that there could be a high amounts of rainfall in certain periods which 
destroys crops. Information from farmers reveal that deceptive rains which start well in 
the beginning of a growing season and ends abruptly during the growing season 
destroys crops and thus low yields. Despite the unreliable rainfall pattern and intensity, 
the dry spells also pose a threat to yield as it has been observed to be increasing in 
certain areas and fluctuating in other areas, (Owusu, 2009) 
 
According to Barry et al. (2005), poor water resources management has resulted in an 
inability to bridge intra-seasonal dry-spells, which often causes reductions in yield. 
They further mentioned, low yields could be avoided with better water management for 
agriculture in the tropical drylands. 
 
An interview with the planning director of the district Ada East district informed that 
there was no recorded database nor maps of showing the nature, trend and rate at which 
the land cover and land-use is changing in the district. The land-use of the district, he 
claimed, is changing very fast from the coast as observed. The nature of change as 
observed is a decrease in farming activities and clearing of bushes (vegetation) for 
construction of buildings (houses and hotels mostly), fuel stations, tourist sites and 
roads.  
 
University of Ghana http://ugspace.ug.edu.gh
123 
 
The vegetable irrigation farming has got so much attention and has created part time 
jobs as well as full time employment for the unemployed youth. Because of this activity, 
cow dung which was a nuisance initially and a challenge with regards to how to dispose 
of it has now become useful to all those engaged in this farming activity. The cow dung 
is collected and composted. It is bagged and sold to those who need it. Moreover 
everyone engaged in this activity uses the compost because of the nature of the soil 
(sandy loam) in this area. With an initial start-up cost of approximately 2,500 Ghana 
cedis, almost all of those engaged in this activity greed to making enough profit out of 
the venture. 
 
Although efforts are being made in water management, groundwater resources are 
poorly understood and under exploited as indicated by Martin and Van de Giesen 
(2005), and Lemoalle and Condappa (2010). This study’s findings however tells 
otherwise in the Ada Foah area of the Ada East district.  
 
Challenges and Constraints  
A major challenges faced during the study included the type of satellite imagery used. 
Due to the nature and quality of satellite imagery used in this study, the details of certain 
land-use features fell under the mixed pixels category and were thus generalized and 
could not be mapped appropriately. This is because these features cover a very small 
area and the 30x30 resolution Landsat imagery used could not capture such detail in 
developing the land-use maps. 
 
However the cost involved in getting such imagery is very high and usually purchased 
by Institutions and/or companies.   
University of Ghana http://ugspace.ug.edu.gh
124 
 
CHAPTER SIX 
CONCLUSION AND RECOMMENDATIONS 
 
6.1  Conclusion 
From the two satellite images studied through LUCC mapping and NDVI analysis, it is 
observed that various land use types have been changed from one type to another. 
Rainfall especially impacts on vegetation cover in this area. In as much as changing 
climate is affecting vegetation cover and agricultural crop production, the quest to 
increase crop yield influenced by a growing population of farmers is the chief driver of 
land use change in the Lower Volta Basin. 
 
The largest and dominant conversion of land use in the LVB over the last 12 years has 
been the expansion of agriculture at the expense of grasslands/shrub (grazing land) and 
mixed savannah vegetation. Moreover the dominant land use/cover type in the LVB 
currently, is grasslands/shrub constituting 32.3%. Farmlands and grasslands together, 
cover 62.1% of the whole LVB. Structure and composition of vegetation of the Lower 
Volta Basin is changing rapidly. The patterns of land use in the LVB are highly 
determined by rainfall amounts and soil characteristics. However, vegetation, 
especially grasslands (grazing fields) increase and decrease annually due to the wet and 
dry seasons experienced every year in the LVB. 
 
This study has shown that the maximum likelihood classification method is indeed good 
for mapping Land use/cover changes that take place is large areas such as the LVB. 
 
University of Ghana http://ugspace.ug.edu.gh
125 
 
In the LVB, it is established from this study that there has been a decrease of rainfall 
over the years. This study also indicates how rainfall amounts and pattern could affect 
vegetation and agricultural crop production. Temperature, another climate parameter 
was shown to have increased over the years in the LVB with implications for vegetation 
cover and crop productivity as well.  
 
The decreasing trend of rainfall and increasing temperature will have adverse effects 
on all land based productive systems in the LVB. These changes are likely to cause big 
challenges on efforts to alleviate poverty and promote food security. It is therefore 
necessary to consider interventions to prepare the communities on how to adapt to 
climate change.  
 
This study also shows that crop yield from the two study districts in the Lower Volta 
Basin had experienced an increasing trend. However the production was associated to 
the increased tracts of farmlands. The increase in crop production was due to a number 
of reasons which includes, an increase in cropped area, expansion of private and state 
owned irrigation farming along the Volta River and the use of drought resistant seeds 
as well as the application of fertilizers. The growing of large scale irrigation farming 
especially along the river from Akuse to Asutuare has led to small scale farmers losing 
their farm fields. This growing phenomenon is however creating employment 
opportunities for many people in the area. 
 
The crop suitability models in this study shows that current climatic conditions provide 
excellent, very suitable and suitable conditions for crop production in the LVB. But 
projected future climatic conditions (~2050) showed that most of the southernmost 
parts of the LVB which includes the Ada East and South Tongu areas remain excellent 
University of Ghana http://ugspace.ug.edu.gh
126 
 
and suitable whilst the Lower Manya Krobo area (North western parts) is expected to 
have marginal, very marginal and not suitable conditions.  
  
University of Ghana http://ugspace.ug.edu.gh
127 
 
6.2  Recommendations 
Most of the land use/cover research that has been done in the Volta Basin of Ghana 
focuses on the Northern part of the Basin with very little attention given to the LVB. It 
is however important to track, frequently map and document land use/cover changes 
that occur in the LVB through remote sensing techniques for proper land management 
planning and other policy interventions. It would therefore be necessary for researchers 
to conduct Land Use/Cover change studies often using Landsat satellite images which 
do not cost yet gives accurate information.  
 
 A further detailed crop suitability modelling should be conducted by academic 
researchers. In a rain fed agricultural region, areas suitable for growing crops 
under changing climatic conditions are usually not known. There is therefore the 
need to use climate crop suitability models such as the one used in this study to 
determine which areas will be suitable for growing specific kinds of crops in the 
future. 
 
 The Ministry of Food and Agriculture (MoFA) should ensure records and 
databases of crop production are archived properly to help improve research 
findings. 
 
 There is also the need to increase the number of agricultural extension officers in 
the various districts of the Lower Volta Basin. The agricultural extension officers 
should educate the farmers on the issues of climate change, its impacts and adaptive 
measures as well as assist them. 
 
University of Ghana http://ugspace.ug.edu.gh
128 
 
  The cost of drought and disease resistant seeds should be subsidized by the 
government and made accessible to small holder farmers at moderate prices. 
 The Ghana Irrigation Development Authority should collaborate with local 
irrigation schemes within the LVB to assist small holder farmers so as to mitigate 
the impacts of climate change and increase crop production. 
 There is an urgent need for a regional framework and guidelines for sustainable 
land management including all sectors of land use like cropping, grazing and 
urbanization in the LVB. This should be carried out by all relevant stakeholders 
including the VRA, MoFA, Forestry Commission, Lands Commission, Local 
Government and Local Authority. 
  
University of Ghana http://ugspace.ug.edu.gh
129 
 
REFERENCES 
Africa water facility (AWF). (2010). Reoptimization and reoperation study of 
Akosombo and Kpong dams. 
Agbenyo, P. K. (2009). Effects of the Akosombo and Kpong power schemes on six 
selected Mafi-communities in the Volta region.  
Agyarko, T. (2010). Forestry Outlook Study for Africa (FOSA): Ghana, Accra: FAO 
http://ftp. fao. org/docrep/fao/003/AB567E/AB567E00. Date Accessed 21st. 
April 2015 
Agyarko, T. (2001). Country Report. In: G.M.A.L.A Forestry (Editor) Forestry 
Commission, Ghana, 2010. Annual Report.  
Akgun, A., Eronat, A., H. & Türk, N. (2004). Comparing Different Satellite Image 
Classification Methods: An Application In Ayvalýk District, Western Turkey. In 
International Congress for Photogrammetry and Remote Sensing, Istanbul, 
Turkey. 
Akubia, J. E., K. (2014). Geo-Information Technology Application to Changing 
Climate and Land-Use-Cover Impacts on Water Resources and Socio-Economic 
Conditions in the Lower Volta Basin, Ghana. Unpublished thesis submitted to the 
Institute for Environment and Sanitation Studies, for the award of a MPHIL 
degree. 
Amanor, K., S. & Pabi, O. (2007). Space, time, rhetoric and agricultural change in the 
transition zone of Ghana. Human Ecology, 35(1), 51-67. 
Amatekpor, J., K. (1999). Soils and land use in the Volta basin: State of the art. The 
Sustainable Integrated Development of the Volta Basin in Ghana Volta Basin 
Research Project, Accra. Gordon, C., and Amatekpor, J. A., (editors) 
Andah, W. E., van de Giesen, N. & Biney, C. A. (2003). Water, climate, food, and 
environment in the Volta Basin. Contribution to the project ADAPT, Adaptation 
strategies to changing environments. ADAPT, Accra.  
University of Ghana http://ugspace.ug.edu.gh
130 
 
Anderson, R. P., Lew, D. & Peterson, A. T. (2003). Evaluating predictive models of 
species. Distributions: Criteria for Selecting Optimal Models, Ecological 
Modelling, 162. 
Asante, F.A. (2007). Analysis of water-related poverty in the Volta Basin of Ghana. 
Accra: Institute of Statistical Social and Economic Research (ISSER), University 
of Ghana. 
Asner, G. P. (2004). Biophysical remote sensing signatures of arid and semiarid 
ecosystem, in: S.L. Ustin, (Ed.), Remote Sensing for Natural Resources 
Management and Environmental Monitoring. Third ed., Hoboken, New Jersey, 
pp: 53-109.  
Ayivor, S. J. (1999). Small-Farmer Management of Floodplain Distributed by 
Damming: A case study of the Lower Volta Basin of Ghana; An Unpublished 
Thesis submitted to the Department of Geography and Resource Development, 
University of Ghana, for the Award of MPhil Degree 
Barry, B., Obuobie, E., Andreini, M., Andah, W., & Pluquet, M. (2005). The Volta 
River Basin, Comprehensive Assessment of Water Management in Agriculture, 
Comparative study of river basin development and management. Report. IWMI 
Publication, 33. 
Batterbury, S. P. J. (1999). Environmental histories, access to resources and landscape 
change: an introduction. Land Degradation. Dev. 10(4):279–89 
Berger, T., & Ringler, C. (2002). Tradeoffs, Efficiency Gains and Technical Change: 
Modeling Water Management and Land Use within a Multiple-Agent 
Framework. Quarterly Journal of International Agriculture 41(½):119–144. 
Boateng, E. A. (1967). A Geography of Ghana. Cambridge University Press: 
Cambridge. 
Boko, M., Niang, I., Nyong, A., Vogel, C., Githeko, A. & Medany, M. (2007). Africa. 
In M. L. Parry, O. F. Canziani, J. P. Palutikof, P. J. van der Linden, & C. E. 
Hanson (Eds.), Climate change 2007: Impacts, adaptation and vulnerability. 
Contribution of working group II to the fourth assessment report of the 
University of Ghana http://ugspace.ug.edu.gh
131 
 
intergovernmental panel on climate change. Cambridge: Cambridge University 
Press. 
Braimoh, A. K. (2004). Modelling Land use Change in the Volta Basin of Ghana. 
Ecology and Development Series, No. 14, Cuvillier Verlag, Gottingen.  
Briassoulis, H. (2000). Factors influencing land-use and land-cover change. 
Encyclopedia of Life Support Systems (EOLSS). 
Butera, M. K. (1983). Remote sensing of wetlands. Geoscience and Remote Sensing, 
IEEE Transactions on, (3), 383-392. 
Butt, M. J., Waqas, A., Iqbal, M.F., Muhammad, G. & Lodhi, M.A.K. (2012). 
Assessment of urban sprawl of Islamabad metropolitan area using multi-sensor 
and multi-temporal satellite data. Arab. J. Sci. Eng. 37, 101–114. 
Canada Centre for Remote Sensing/Natural Resources Canada (CCSR), (1997). 
GlobeSAR2 Radar Image Processing and Information Extraction Workbook 
Version 1.2. Ottawa, Ontario, Canada. 
Challinor, A. J., Ewert, F., Arnold, S., Simelton, E. & Fraser, E. (2009). Crops and 
climate change: progress, trends, and challenges in simulating impacts and 
informing adaptation. Journal of Experimental Botany, 60 (10): 2775-2789.  
Challinor, A.J., Simelton, E.S., Fraser, E.D.G., Hemming, D. & Collins, M. (2010). 
Increased crop failure due to climate change: assessing adaptation options using 
models and socio-economic data for wheat in China. Environmental Research 
Letters, 5 (3): 034012.  
Civco, D. L., Hurd, J. D., Wilson, E. H., Song, M., & Zhang, Z. (2002). A comparison 
of land use and land cover change detection methods. In ASPRS-ACSM Annual 
Conference. 
Civco, D.L. & Hurd J.D. (1999). A hierarchical approach to land use and land cover 
mapping using multiple image types. Proc. 1999 ASPRS Annual Convention, 
Portland. pp. 687-698. 
University of Ghana http://ugspace.ug.edu.gh
132 
 
Codjoe, S. N. A, Eckart Ehlers & Paul L. G. Vlek. (2005). Effects of Change in 
Population, Household Conditions and Farming Practices on Agricultural Land 
Use in the Volta River Basin of Ghana, 1984–2000. 
Codjoe, S. N. Ardey. (2004). Population and land use/cover dynamics in the Volta 
River Basin of Ghana, 1960-2010. Ecology and development series no. 15. 
Crumley, CL, eds. (1994). Historical Ecology: Cultural Knowledge and Changing 
Landscapes. Santa Fe: Sch. Am. Res. 304 
CSIR-WRI (Council for Scientific and Industrial Research-Water Research Institute). 
(2000). Climate change vulnerability and adaptation assessment on water 
resources of Ghana. A UNFCC/EPA/WRI Accra report.  
d’Aquino, P., August, P., Balmann, A., Berger, T., Bousquet, F., Brondízio, E., & 
Warren, K. (2002). Agent-based models of land-use and land-cover change. In 
Proc. of an International Workshop (pp. 4-7). 
Daniel, L. Civco, James D. Hurd, Emily H. Wilson, Mingjun Song & Zhenkui Zhang 
(2002). A Comparison of Land Use and Land Cover Change Detection Methods. 
ASPRS-ACSM Annual Conference and FIG XXII Congress.  
Dawelbait, M. A., & Morari, F. (2011). LANDSAT, Spectral Mixture Analysis and 
Change Vector Analysis to Monitor Land Cover Degradation in a Savannah 
Region in Sudan (1987-1999-2008). Int. J. Water Resour. Arid Environ, 1, 
366377. 
De Sherbinin, A. (2002). A guide to land-use and land-use cover change (LUCC). A 
collaborative effort of SEDAC and the IGBP/IHDP LUCC Project. Columbia 
University, New York. 
Dean, A. M., & Smith, G. M. (2003), An evaluation of per–parcel land cover mapping 
using maximum likelihood class probabilities, International Journal of Remote 
Sensing, 24 (14), pp. 2905-2920. 
Degu, A. M., F. Hossain, D. Niyogi, R. A. Pielke Sr., J. M. Shepherd, N. Voisin, & T. 
Chronis (2011). The influence of large dams on sur- rounding climate and 
precipitation patterns, Geophys. Res. Lett., L04405, doi: 10. 1029/2010GL04648.  
University of Ghana http://ugspace.ug.edu.gh
133 
 
Duadze, S. E. K. (2004). Land use and land cover study of the savannah ecosystem in 
the Upper West Region (Ghana) using remote sensing. Ecology and development 
series No. 16. 
Eastman, J. R., Fulk M. (1993). Long sequence time series evaluation using 
standardized principal components. Photogrammetry. Eng. Remote Sens. 
59(8):991–96 
El-Kawy, O. R. A., Rod, J. K., Ismail, H. A. & Suliman, A.S. (2010). Land use and 
land cover change detection in the Western Nile Delta of Egypt using remote 
sensing data. Applied. Geography. 31 (2011), 483–494. 
El-Raey, M., Sahraf El Din S.H., Kafagy, A.A. & Abo Zed, A.I. (1999). Remote sensing 
of beach erosion and accretion patterns along Damietta-port side shoreline, 
Egypt. International Journal of Remote Sensing, 20, pp. 1087–1106.  
FAO, (2000). The Ecocrop database. In: FAO (Editor), Rome, Italy.  
FAO, (2000). Two essays in climate change and agriculture. Rome: Food and 
agricultural organization. 
FAO, (2006). World Agriculture: Towards 2030/2050. Prospects for Food, Nutrition, 
Agriculture and Major Community Groups. Interim report. Global perspective 
studies unit. Rome 
Fisher, B., Anderson, S., Bryant, J., Margolese, R. G., Deutsch, M., Fisher, E. R. and 
Wolmark, N. (2002). Twenty-year follow-up of a randomized trial comparing 
total mastectomy, lumpectomy, and lumpectomy plus irradiation for the treatment 
of invasive breast cancer. New England Journal of Medicine, 347(16), 1233-
1241. 
Fobil, J. N. and Attaquayefio D. K. (2003). Remediation of the environmental impacts 
of the Akosombo and Kpong dams. Volta Basin Research Project (VBRP). 
Food and Agricultural Organization of the United Nations (FAO). (1995). Planning for 
Sustainable Use of Land Resources. FAO Land and Water Bulletin 2. 472 pp. 
Rome: Food and Agricultural Organization of the United Nations. 
University of Ghana http://ugspace.ug.edu.gh
134 
 
Food and Agriculture Organization (FAO). (2008) Climate and Climate Change in 
West Africa, 2008. http://www.fao.org/nr/clim/abst/clim_080502_en.htm. 
Accessed 12 May 2015. 
Geerts, S., Raes, D., Garcia, M., Del Castillo, C., & Buytaert, W. (2006). Agro-Climatic 
Suitability Mapping For Crop Production In The Bolivian Altiplano: A case study 
for quinoa. Agricultural and forest meteorology, 139(3), 399-412.  
Geker, J. (1999). The Effects of the Volta Dam on the People of the Lower Volta. The 
Sustainable Integrated Development of the Volta Basin in Ghana. VBRP, Legon, 
119-121. 
Ghana Statistical Service, (2014). Ghana Living Standards Survey 6 (GLSS 6), Main 
report. 
Gleick, P. H. (2009). The World’s Water. Island Press. ISBN, 10, 1-59726.  
Global Environment Facility (GEF)-United Nations Environment Programme (UNEP). 
(2002). Project Development Facility (PDF-B) 
Gonzales, C. A. (2009). Assessing the Land Cover and Land Use Change and Its Impact 
on Watershed Services in a Tropical Andean Watershed of Peru. A Master of 
Science thesis to the Department of Biological and Environmental Science, 
University of Jyvaskyla. 
Gordon, C. (1999). An overview of the fish and fisheries of the Volta basin. Gordon, 
C., & Amatekpor, J. A., (editors) The Sustainable Integrated Development of the 
Volta Basin in Ghana Volta Basin Research Project, Accra. 
Govindasamy, B, P., Duffy B. & Coquard J.  (2003). High-resolution simulations of 
global climate, part 2: effects of increased greenhouse cases. Climate Dynamics 
21: 391–404. 
Gregory, T. A., (2008). Reoptimization of West African Dams to Restore Downstream 
Ecosystems & Livelihoods. 
Grübler, A. & Nakićenović, N. (1994). International burden sharing in greenhouse gas 
reduction. International Institute for Applied Systems Analysis. 
University of Ghana http://ugspace.ug.edu.gh
135 
 
Gu, G., & Adler, R. F. (2003). Seasonal rainfall variability within the West African 
Monsoon system. CLIVAR Exchanges, 8, 11-15. 
Guide, E. U. S. (2008). ENVI on-line software user’s manual. ITT Visual Information 
Solutions. 
Gumma, K. M., Prasad, T. S., & Bubacar, B. (2010). Delineating shallow ground water 
irrigated areas in the Atankwidi Watershed (Northern Ghana, Burkina Faso) using 
Quickbird 0.61-2.44 meter data. African Journal of Environmental Science and 
Technology, 4(7), 455-464. 
Gyasi, E. A. (1999). The Sustainable Integrated Development of the Volta Basin in 
Ghana Volta Basin Research Project, Accra. 
Gyasi, E. A., Karikari, O., Kranjac-Berisavljevic, G., & Vordzogbe, V. V. (2006). 
Study of Climate Change Vulnerability and Adaptation Assessment Relative to 
Land Management in Ghana. University of Ghana, Legon, Accra, 1-91. 
Hansing, O. & Wittich, K. P. (1995). Area averaged vegetation cover fraction estimated 
from satellite data. Int. J. Biometeor., 38 
Hathout, S. (2002). The use of GIS for monitoring and predicting urban growth in East 
and West St Paul, Winnipeg, Manitoba, Canada. J. Environ. Manage. 66, 229–
238. 
Heilig, G. K. (1994). Neglected dimensions of global land-use change: reflections and 
data. Population and Development Review, 831-859. 
Hijmans, R.J., Guarino, L., Cruz, M. & Rojas, E. (2001). Computer tools for spatial 820 
analysis of plant genetic resources data: DIVA-GIS. Plant Genetic Resources 
Newsletter, 127: 15-19.  
Hijmans, R. J., Guarino L., Jarvis A., O’Brien R., Mathur P., Bussink C., Cruz M., 
Barrantes I. & Rojas E. (2005). DIVA-GIS – Version 5.2 – Manual. Available at 
http://www.diva-gis.org/Materials.htm.  
Hoffhine, E., (2000). Forest Change Detection, Harvest and Cover Type Distinction 
Using Multiple Dates of Landsat Thematic Mapper. M.S.-degree Thesis, 
University of Maine, Orono. 
University of Ghana http://ugspace.ug.edu.gh
136 
 
https://mofafoodsecurity.wordpress.com/food-security-situation-in-ghana/: date 
accessed 11th July 2015.  
Huette, A. R., W. J. D. Van Leeuwen, G. Hua, J. Qi & A. Chehbouni, (1992). 
Normalization of multidirectional red and NIR reflectance’s with the SAVI. 
Remote Sensing of Environment, 41: 143-154.  
Hurd, J.D., D.L. Civco, C. LaBash, & P. August. (1992). Coastal wetland mapping and 
change detection in the northeast United States. in Proc. 1992 
ASPRS/ACSM/RT'92 Convention, Washington, D.C. 1:130-139. 
Hurd, J.D., E.H. Wilson, S. Lammey & D.L. Civco. (2001). Characterization of Forest 
Fragmentation and Urban Sprawl using Time Sequential Landsat Imagery. Proc. 
2001 ASPRS Annual Convention, St. Louis, MO. 
Hussain, I. (2005). Pro-poor Intervention Strategies in Irrigated Agriculture in Asia. 
Poverty in Irrigated Agriculture: Issues, Lessons, Options and Guidelines 
(Bangladesh, China, India, Indonesia, Pakistan and Vietnam). Project Final 
Synthesis Report. IWMI, Colombo. 
IPCC. (2007). Climate Change 2007: The Physical Science Basis. Contribution of 
Working Group I to the Fourth Assessment Report of the Intergovernmental 
Panel on Climate Change. IPCC. 
IPCC. (2007). Climate Change: Impacts Adaptation and Vulnerability. Contribution of 
working group II to the Fourth Assessment Report of the IPCC. 
Iqbal, M. F., & Khan, I. A. (2014). Spatiotemporal land use land cover change analysis 
and erosion risk mapping of Azad Jammu and Kashmir, Pakistan. The Egyptian 
Journal of Remote Sensing and Space Science, 17(2), 209-229. 
Iqbal, M. F., Khan, M. R. & Malik, A.H. (2013). Land use change detection in the 
limestone exploitation area of Margalla Hills National Park (MHNP), Islamabad, 
Pakistan using Geo-spatial techniques. J. Himalayan Earth Sci. 46 (1), 89–98. 
IWMI (International Water Management Institute). (2013). “Water Policy Briefs.” 
[Available Online]: http://www.iwmi.cgiar.org/publicatios/briefs/water-policy-
briefs.  
University of Ghana http://ugspace.ug.edu.gh
137 
 
Jung, G. and Kunstmann, H. (2007). High-resolution regional climate modelling for the 
Volta region of West Africa.  
Klepeis, P., Turner B. L. (2001). Integrated land history and global change science: the 
example of the Southern Yucat´an Peninsular Region Project. Land Use Policy 
18(1):27–39 
Koku, J. E., (2002). Environment, Livelihood and Natural Resource Management In 
The Lower Volta Basin Of Ghana.  
Lachaud, J. P., (1998). Modelling the determinants of poverty and the labor market in 
Africa: the case of Burkina Faso. Working Papers CED 32. Bordeaux: Bordeaux 
University  
Lambin, E. F. and Geist H. J. (2006) Land-Use and Land-Cover Change, Local 
Processes and Global Impacts, Springer 2006 
Lambin, E. F., Baulies X, Bockstael N, Fischer G, Krug T, Leemans R, Moran EF, 
Rindfuss RR, Sato Y, Skole D, Turner B. L, Vogel C, Nunes C & Augè J. I (1999) 
Land-use and land-cover change implementation strategy. Stockholm, Sweden: 
IGBP Report No. 48 
Lambin, E. F. & Geist H. J. (2002) Global land-use and cover change: What have we 
learned so far? Global Change Newsletter 46, p. 27-30. 
Lambin, E. F., Geist H. J. & Lepers E. (2003) Dynamics of land-use and land-cover 
change in tropical regions. Annual review of environment and resources 28, p 
205 
Lambin E. F., Turner BL, Geist HJ, Agbola SB, Angelsen A, Bruce JW, Coomes OT, 
Dirzo R, Fischer G, Folke C, George PS, Homewood K, Imbernon J, Leemans R, 
Li X, Moran EF, Mortimore M, Rama krishnan PS, Richards JF, Skanes H, 
Steffen W, Stone GD, Svedin U, Veldkamp TA, Vogel C & Xu J (2001) The 
causes of land-use and land-cover change: moving beyond the myths. Global 
Environmental Change 11, p. 261-269. 
Lambin, E. F. (1997). Modelling and monitoring land-cover change processes in 
tropical regions. Progress in physical geography, 21(3), 375-393. 
University of Ghana http://ugspace.ug.edu.gh
138 
 
Lane, A., & Jarvis, A. (2007). Changes in climate will modify the geography of crop 
suitability: agricultural biodiversity can help with adaptation. SAT ejournal, 4 (1), 
1-12.  
Laux, P., Kunstmann, H. & Bárdossy, A. (2007). Predicting the regional onset of the 
rainy season in West Africa. International. Journal of Climatology. 
doi:10.1002/joc.1542. 
Leemans, R., Lambin E. F, McCalla A., Nelson J., Pingali P. & Watson B. (2003). 
Drivers of change in ecosystems and their services. In Ecosystems and Human 
Well-Being: A Framework for Assessment, ed. H Mooney, A Cropper, W Reid. 
Washington, DC: Island. 
Lemoalle, J. and Devaraj de Condappa. (2010) Farming systems and food production 
in the Volta Basin, Water International, 35:5, 655-680, DOI: 
10.1080/02508060.2010.510793 
Lemoalle, J. (2007). Global trends in the Volta basin. BFP Volta Reports Series.  
Lillesand, T. M. & Kiefer R. W. (1994). Remote sensing and image interpretation. John 
Wiley and Sons. 
Liu, X. H., Skidmore, A. K., & Osten, V. H. (2002), Integration of classification 
methods for improvement of land-cover map accuracy, ISPRS Journal of 
Photogrammetry & Remote Sensing, 56, pp 257-268. 
Lo, C. P., and Watson, L. J. (1998). The influence of geographic sampling methods on 
vegetation map accuracy evaluation in a swampy environment. Photogrammetric 
Engineering and Remote Sensing, 64(12), 1189-1200. 
Lobell, D. B., Burke, M. B., Tebaldi, C., Mastrandrea, M. D., Falcon, W. P., & Naylor, 
R. L. (2008). Prioritizing climate change adaptation needs for food security in 
2030. Science, 319(5863), 607-610.  
Loveland, T. R., Zhu Z., Ohlen D. O., Brown J. F., Reed B. C. & Yang L. M. (1999). 
An analysis of the IGBP global land-cover characterization process.  
University of Ghana http://ugspace.ug.edu.gh
139 
 
Lunetta, R.L., Johnson, D.M., Lyon, J.G. & Crotwell, J. (2004). Impacts of imagery 
temporal frequency on land-cover change detection monitoring. Remote Sensing. 
Environment. 89(4), 444-454. 
Lunetta, R.L., Knight, F.K, Ediriwickrema, J., Lyon, J.G., & Worthy, L.D. (2006). 
Landcover change detection using multi-temporal MODIS NDVI data. Remote 
Sensing of Environment, 105, 142-154. 
Mallapattu, P. K., Reddy J. R & Sreenivasula. (2013). Analysis of Land Use/Land 
Cover Changes Using Remote Sensing Data and GIS at an Urban Area, Tirupati, 
India. The Scientific World Journal, vol. 2013, Article ID 268623. 
doi:10.1155/2013/268623 
Manakos, I., Scheneider T. & Ammer U. (2000). A Comparison between the ISODATA 
and the eCognition Classification Methods on Basis of Field Data. IAPRS, 
33(Supp. B7): 133-139. 
Martin, N., & Van De Giesen. N. (2005). Spatial distribution of groundwater production 
and development potential in the Volta River Basin of Ghana and Burkina Faso. 
Water International 30.2 (2005): 239-249 
Masiyandima, M. & Giordano, M. (2007). Sub-Saharan Africa: Opportunistic 
exploitation In: the agricultural groundwater revolution: opportunities and threats 
to development. 
McCarthy, J. J., Canziani, O. F., Leary, N. A., Dokken, D. J., & White, K. S, (2001). 
Climate change 2001: Impacts, Adaptation and Vulnerability. Cambridge: 
Cambridge University Press. 
Mei, X. & Qing, R., (1999). Change detection based on remote sensing information 
model and its applications on coastal line of Yellow River Delta. Earth 
Observation Centre, NASDA, China 
Ministry of Environment and Science (MES). (2002). National biodiversity strategy for 
Ghana. CBD. http://www.cbd.int/doc/world/gh/gh-nbsap-01-en.pdf 
Ministry of Food and Agriculture (MOFA). (2011). Agriculture in Ghana, facts and 
figures. Statistics, research and information directorate. 
University of Ghana http://ugspace.ug.edu.gh
140 
 
Ministry of Food and Agriculture. (2015). North Tongu District Background. 
http://www.mofa.gov.gh/site date accessed, 12/05/15 
Mona, A. A., Dawelbait, & Francesco, M. (2011). LANDSAT, Spectral Mixture 
Analysis and Change Vector Analysis to Monitor Land Cover Degradation in a 
Savannah Region in Sudan (1987-1999-2008) International Journal of Water 
Resources and Arid Environments 1(5): 366-377  
Mortimore, M. (1998). Roots in the African dust: sustaining the Sub-Saharan drylands. 
Cambridge University Press. 
Muttitanon W., & Tripathi N.K. (2005). “Land use/ land cover changes in the coastal 
zone of Ban Don Bay, Thailand using Landsat 5 TM data”, International Journal 
of Remote Sensing, Vol. 26, No. 11, 2311-2323. Taylor and Francis. (ISSN: 
1366‐5901, Impact Factor 1.182). 
Natural Heritage Institute (NHI). (2007). Opportunities for Environmentally 
Hydropower Dam Re-optimization In The West African Power Pool Network of 
ECOWAS. 
Nunes, C. & Auge J. I. (1999). Land-use and Land-cover Implementation Strategy 
(Stockholm: IGBP) 
Nyamekye, C. (2010). Classification of time series of NDVI for Assessment of land 
cover change in Ghana using Noaa/A VHRR data (Doctoral dissertation, 
Department of Geomantic Engineering, Kwame Nkrumah University of Science 
and Technology). 
Ofosu, E. A. (2011). Sustainable Irrigation Development in the White Volta sub-Basin: 
UNESCO-IHE PhD Thesis. CRC Press. 
Okine, C. B. & Nyarko, J. A. (2008). Volta Basin Residents Train in Tree Management. 
Opoku-Ankomah, Y., & Cordrey, I. (1994). Atlantic sea surface temperatures and 
rainfall variability in Ghana. Journal of Climate, 7, 551-558. 
Osman-Elasha, B. (2010). Mapping of Climate Change Threats and Human 
Development Impacts in the Arab Region. United Nations Development 
University of Ghana http://ugspace.ug.edu.gh
141 
 
Programme, Regional Bureau for Arab States. Arab Human Development Report; 
Research Paper Series.  
Ostrom, E. (1990). Governing the Commons: The Evolution of Institutions for 
Collective Action. Cambridge, UK: Cambridge Univ. Press. 280 pp. 
Owusu, K. (2009). Changing rainfall climatology of West Africa: Implications for rain-
fed agriculture in Ghana and water sharing in the Volta Basin.  
Ozesmi, S. L & Bauer, M. (2002), Satellite Remote Sensing of Wetlands, Wetlands 
Ecology and Management, 10, pp 381-402. 
Pal, M. & Mather, P. M. (2003). An assessment of the effectiveness of decision tree 
methods for land cover classification. Remote Sensing of Environment, 86, 
pp.554–565. 
Parker, D. C., Manson, S. M., Janssen, M. A., Hoffmann, M. J. & Deadman, P. (2003). 
Multi-agent system models for the simulation of land-use and land-cover change: 
a review. Ann. Assoc. Am. Geogr. 93(2):314–37 
Parthasarathy, U., Jayarajan, K., Johny, A. K., & Parthasarathy, V. A. (2008). 
Identification of suitable areas and effect of climate change on ginger-a GIS 
study.  
Parthasarathy, U., Nandakishore, O. P., Jayarajan, K., Saji, K. V., & Nirmal, K. (2014). 
Prediction of Crop Suitability of Certain Indian Spices–A GIS Approach. 
Peter, D. T., Michael, L. L. (2003). Measuring praise and criticism: inference of 
semantic orientation from association. ACM Trans. Inf. Syst. 21, 315–346. 
Petit, C. (2001). Quantifying processes of land-cover by remote sensing, resettlement 
and rapid land-cover change in south-eastern Zimbabwe Remote Sensing 22(17): 
3435-3456 
Plantier, T. M., Loureiro, P., Marque, M. C. (2006). Spectral analyses and classification 
of ikonos images for forest cover characterisation. Proceedings of the 2nd 
Workshop of the EARSeL SIG on Land Use and Land Cover, Center for Remote 
Sensing of Land Surfaces, Bonn, 28-30. 
University of Ghana http://ugspace.ug.edu.gh
142 
 
Porter, J. R., Jamieson, P. D., & Grace, P. R. (2007). Wheat Production Systems and 
Global Climate Change. In Terrestrial Ecosystems in a Changing World (pp. 195-
209). Springer Berlin Heidelberg. 
Prakasam, C. (2010). Land use and land cover change detection through remote sensing 
approach: A case study of Kodaikanal taluk, Tamil nadu, International Journal of 
Geomatics and Geosciences 1(2): 32-44 
Purevdor, T. S., Tateishi, R., Ishiyama, T. & Honda, Y (1998). Relationship between 
percent vegetation cover and vegetation indices. International journal of remote 
sensing, 1998, 19 (18), 3519-3535.  
Rai, P. K., Gupta, S., Mishra, A., & Onagh, M. (2011). Multi-seasonal Irs-1c Liss III 
Satellite Data for Change Detection Analysis and Accuracy Assessment: A Case 
Study. Journal of GIS Trends, 2 (1), 13-19. 
Ramirez, J., & Bueno-Cabrera, A. (2009). Working with climate data and niche 
modeling I. Creation of bioclimatic variables. International Center for Tropical 
Agriculture (CIAT), Cali, Colombia, 1-6. 
Ramirez-Villegas, J., Jarvis, A., & Läderach, P. (2013). Empirical approaches for 
assessing impacts of climate change on agriculture: the EcoCrop model and a case 
study with grain sorghum. Agricultural and Forest Meteorology, 170, 67-78.  
Reilly, J., Baethgen, W., Chege, F. E., Geijn, V. D. S., Erda, L., Iglesias, A., & Howden, 
M. (1996). Agriculture in a changing climate: impacts and adaptation. In: RT 
Watson, MC Zinyowera & RH Moss (eds.), Climate change 1995; impacts, 
adaptations and mitigation of climate change: scientific-technical analyses. 
Cambridge (UK), Cambridge Univ. Press, 1996, pp. 427-467. 
Richards, J. A. (1995). Remote sensing digital image analysis. An introduction. Second 
revised and enlarged edition. Springer. Verlag. Berlin.  
Rodgers, C., van de Giesen, N., Laube, W., Vlek, P. L. G., & Youkhana, E. (2007). The 
GLOWA Volta Project: A framework for water resources decision-making and 
scientific capacity building in a transnational West African Basin. Water 
Resources Management, 21(1), 295–313. doi:10.1007/s11269-006-9054-y 
University of Ghana http://ugspace.ug.edu.gh
143 
 
Rosegrant, M. W., Paisner, M. S., Meijer, S. & Witcover, J. (2001). Global Food 
Projections to 2020: Emerging Trends and Alternative Futures. Washington D.C.: 
International Food Research Institute. 
Rosenzweig, C., Parry, M.L., Fischer, G. & Frohberg, K. (1993). Climate change and 
world food supply. Oxford: Environmental Change Unit, University of Oxford, 
UK.  
Sagoe, R. (2006). Climate change and root crop production in Ghana. Crop Research 
Institute: Kumasi, Ghana (2006):2-37 
Saxena, K. R. (2005). Dams: Incidents and accidents, A. A. Balkema, Leiden, 
Netherlands.  
Schindler, J., & Ghana, U. E. (2009). Ecology and Development Series No. 68, 2009. 
Schultz, B. (2002). Role of dams in irrigation, drainage and flood control. International 
Journal of water resources development, 18(1), 147-162 
Sivakumar, M. V. K., & Hatfield, J. L. (1990). Spatial variability of rainfall at an 
experimental station in Niger, West Africa. Theoretical and Applied Climatology, 
42(1), 33-39. 
Stemmler, J. & Su, Z., (1993). In Su, Z. 2000. Remote sensing of land use and 
vegetation for mesoscale hydrological studies. Int. j. remote sensing, 2000, vol. 
21, no. 2, 213-233.t 
Sunnar, F. & Musaogilu, N. (1998). An analysis of changes in multi-date data set: A 
case study in the Ikitelli area, Instabul, Turkey. International journal of remote 
sensing, 19 (2), 225-235.  
Tansey, K. T. & Millington, A. C. (2006). Land use/land cover change detection in 
Metropolitan Lagos (Nigeria): 1984-2000. ASPRS 2006 Annual Conference 
Reno, Nevada, May 1-5 
Tayyebi, A., Delavar, M. R., Saeedi, S., Amini, J., & Alinia, H. (2008). Monitoring 
land use change by multi-temporal Landsat remote sensing imagery. The 
International Archives of the Photogrammetry, Remote Sensing and Spatial 
Information Sciences, 37(B7), 1037-42.  
University of Ghana http://ugspace.ug.edu.gh
144 
 
Terrasson, I., Fisher, M. J., Andah, W., & Lemoalle, J. (2009). Yields and water 
productivity of rain-fed grain crops in the Volta Basin, West Africa. Water 
International, 34(1), 104-118. 
Thornton, P.K., Jones, P.G., Ericksen, P.J. & Challinor, A.J. (2011). Agriculture and 
905 food systems in sub-Saharan Africa in a 4°C+ world. Philosophical 
Transactions of the Royal Society A: Mathematical, Physical and Engineering 
907 Sciences, (1934): 117-136.  
Titriku, P. K. (1999). Agriculture in the Volta basin: Problems and prospects. Gordon, 
C., & Amatekpor, J.A., (editors) The Sustainable Integrated Development of the 
Volta Basin in Ghana Volta Basin Research Project, Accra. 
Torahi, A. A. & Suresh, C. R. (2011). Land Cover Classification and Forest Change 
Analysis, Using Satellite Imagery - A Case Study in Dehdez Area of Zagros 
Mountain in Iran Journal of Geographic Information System, 2011, 3, 1-11 
doi:10.436/jgis.2011.31001  
Travis, R. (2006). Drought and its effect on vegetation, comparison of Landsat NDVI 
for drought and non-drought years related to land use land cover classifications. 
Tucker, C.J. (1979). Red and photographic infrared linear combinations for monitoring 
vegetation. Remote Sensing of Environment, 8: 127-150.  
Tufuor, K. (2012) Ghana Forestry Report – draft Towards ECOWAS Forest 
Convergence Plan. 
Tuinhof, A.; Foster, S.; van Steenbergen, F.; Talbi, A. & Wishart, M. (2011). 
Appropriate groundwater management policy for sub-Saharan Africa in face of 
demographic pressure and climatic vulnerability. Strategic Overview series 
Number 5. Washington, DC, USA: World Bank. 
Turner, B. L. II, Clark C., Kates R. W., Richards J. F., Mathews J. T., & Meyer W. B. 
(eds.) (1990). The Earth as Transformed by Human Action: Global and Regional 
Changes in the Biosphere over the past 300 years. Cambridge: Cambridge 
University Press.  
University of Ghana http://ugspace.ug.edu.gh
145 
 
Turner, B. L. II, Skole D., Sanderson S., Fischer G., Fresco L., & Leemans R. (1995). 
Land-Use and Land-Cover Change; Science/Research plan. IGBP (International 
Geosphere/Biosphere Programme). 
Tutu, B. D. (2008). Assessing The Effects of Land-use/cover change on ecosystem 
services in Ejisu-Juaben District, Ghana: The case study of carbon sequestration. 
Enschede, ITC. 
UNEP-GEF Volta Project. (2010). Volta Basin Transboundry Diagnostic Analysis: 
National Report Ghana. UNEP/GEF/Volta/NR GHANA 1/2010 
UNFCCC (2013). Climate Change. Impacts, Vulnerabilities and Adaptation in 
Developing Countries. Bonn: United Nations Framework Convention on Climate 
Change. 
United Nations Environment Programme (UNEP) Project Development Facility 
(2002). Volta River Basin Preliminary Trans boundary Diagnostic Analysis. Final 
report.  
UN-Water. (2009). Climate Change Adaptation is Mainly About Water. Key message 
prepared for the 15th session of the Conference of the Parties to the United 
Nations Framework Convention on Climate Change. New York, UN-Water. 
Available from http://www.unwater.org/downloads/ 
UN-Water. (2011). Water and Climate Dialogue. Adapting to Climate Change: Why 
We Need Broader and ‘Out-of-the-Box’ Approaches.  
USGS, http://www.landsat.usgs.gov/panchromatic_image_sharpening_php, date 
accessed; 14/05/15. 
Ustin, S.L., P.J. Zarco-Tejada S. Jacquemoud & G. P. Asner. (2004). Remote sensing 
of environment: state of science and new directions. In: S.L. Ustin (ed.): Remote 
Sensing for Natural Resources Management and Environmental Monitoring, pp: 
679-728.  
Van de Giesen, N., Liebe, J., & Jung, G. (2010). Adapting to climate change in the 
Volta Basin, West Africa. Current science, 98 (8), 1033-1037. 
University of Ghana http://ugspace.ug.edu.gh
146 
 
Villordon, A., Njuguna, W., Gichuki, S., Ndolo, P., Kulembeka, H., Jeremiah, S. C. & 
Mwanga, R. O. (2006). Using GIS-based tools and distribution modeling to 
determine sweet potato germplasm exploration and documentation priorities in 
sub-Saharan Africa. HortScience, 41(6), 1377-1381. 
Vitousek, P.M. (1992). Global environmental change: An introduction. Ann. Rev. 
Ecological Evolution System. 23, 1–14. 
Weldeab, S., Lea, D. W., Schneider, R. R., & Andersen, N. (2007). 155,000 years of 
West African monsoon and ocean thermal evolution. Science, 316, 1303 – 1307 
Williams, K. A. (2005). An overview of the US National Plant Germplasm System's 
exploration program. HortScience, 40(2), 297-301. 
Woldemichael, A. T., Hossain, F., Pielke, R., & Beltrán‐Przekurat, A. (2012). 
Understanding the impact of dam‐triggered land use/land cover change on the 
modification of extreme precipitation. Water Resources Research, 48(9). 
Wong, R. K., Fung, T., Leung, K. S & Leung, Y. (1997). The Compression of a 
Sequence of Satellite Images Based on Change Detection. International journal 
of remote sensing, 18 (11), 2427-2436.  
Woodwell, G. M., Hobbie J. E, Houghton R. A, Melillo J. M. & Moore B. (1983). 
Global deforestation: contribution to atmospheric carbon dioxide. Science 222: 
1081–86 
World Water Assessment Programme (WWAP), (2011). Water and Climate Dialogue 
- Adapting to Climate Change: Why We Need Broader and 'Out-of-the-Box' 
Approaches 
WWAP (World Water Assessment Programme). (2009). World Water Development 
Report 3: Water in a Changing World. Paris/ London, UNESCO/Earthscan 
Xiao, J., Shen, Y., Ge, J., Tateishi, R., Tang, C., Liang, Y., & Huang, Z. (2006). 
Evaluating urban expansion and land use change in Shijiazhuang, China, by using 
GIS and remote sensing. Landscape and Urban Planning, 75, 69–80. 
Yates, D. N., & Strzepek, K. M. (1998). An assessment of integrated climate change 
impacts on the agricultural economy of Egypt. Climatic Change, 38(3), 261-287. 
University of Ghana http://ugspace.ug.edu.gh
147 
 
Yeboah, F. K. (1999). Mitigative Actions Taken by the VRA on Dam Affected 
Yumbya, J., De Vaate, M. B., Kiambi, D., Kebebew, F., & Rao, K. P. C. (2014). 
Geographic Information Systems for assessment of climate change effects on teff 
in Ethiopia. African Crop Science Journal, 22, 847-858.  
Yusuf, K. O., and Salami, A. W. (2009). Influence of Jebba Hydropower Dam on 
statistical distribution of hydro-meteorological variables in Niger River Basin, 
Nigeria. In 1st Annual Civil Engineering Conference, University of Ilorin, 
Nigeria (Vol. 26, pp. 32-40). 
Zhou, J. & Civco D.L. (1998). A wavelet transform method to merge Landsat TM and 
SPOT Panchromatic Data. International Journal of Remote Sensing 19(4):743-
757. 
Zubair, A. O. (2006). Change Detection in Land Use and Land Cover Using Remote 
Sensing Data and GIS. A project submitted to the department of geography, 
University of Ibadan in partial fulfilment for the award of Master of Science 
(MSc) Degree in Geographical Information Systems. 
 
  
University of Ghana http://ugspace.ug.edu.gh
148 
 
APPENDICIES 
APPENDIX 1: SOCIO-ECONOMIC SURVEY QUESTIONNAIRE 
The data collected for this survey is for academic purposes only and shall be kept 
strictly confidential as your identity shall also not be required. The purpose of this 
survey is to improve our understanding of the drivers of land use change and the 
influence of rainfall variability on the livelihoods of farmers.  
.LQGO\7LFN¥ZKHUHDSSURSULDWHDQGSURYLGHDQVZHUVWRRSHQ-ended 
questions. 
Interview Date..................  District………………...........               
Community ………........................ 
SECTION A: ACKGROUND INFORMATION 
1. Farmer‘s Gender: [  ] Male [  ] Female 
2. Age:  18-25[  ]  26-35[  ]  36-45[  ]  45 and above[  ]............. 
3. Marital Status:   Single [  ]  Married [  ] Divorced [  ] 
4. Years of residence of respondent in the community  
0-5 [  ]  5 -10 [  ]  10-15 [  ]  15-30[  ]  >30[  ] 
5. Highest educational level: No Formal Education[  ]  Primary[  ]  Middle(JHS)[  ]                             
Secondary(SHS) [  ]       Tertiary [  ] 
6. Household size? ...................... 
7.  What types of crops do you cultivate on your farm? 
.......................................................................................................................................... 
 
SECTION B: AGRICULTURE LAND USE AND DRIVERS OF LAND USE 
CHANGE 
8. Do you have knowledge on land use changes? 
[  ] Yes  [  ] No 
 
9. If yes what is your general observation on land use change? (please tick appropriate) 
University of Ghana http://ugspace.ug.edu.gh
149 
 
1 Increasing farm lands and decreasing vegetation  
2 Decreasing farmlands and increasing infrastructure  
3 Increasing infrastructure and decreasing vegetation  
4 Decreasing vegetation and increasing bare lands  
5 Decrease in size of wetlands and marshy areas and increasing 
infrastructure 
 
6. 
 Other.....................................................................................................................
.......... 
 
10. What kind of land tenure system do you have? 
i) Owned[  ]  ii) Leased[  ]  iii) Communal land[  ]  iv) Other........................... 
 
11. If owned how? 
i) Inherited[  ]   ii) Bought[  ]  iii) given by government[  ]  iv) 
Other...................... 
 
12. In what year did you acquire the farming field? ................................ 
 
13. What was on your farm (field) when you acquired it?  
i) Forest[  ]  ii) Grazing land[  ]  iii) Bush[  ]  iv) Woodlot[  ]  v) Crops[  
]  
vi) Other .............................................. 
14. How large is your farm? ....................................  
 
15. How long have you been farming? ....................................................  
i)  <5yrs[  ]  ii) 5-10yrs[  ]  iii) 10-20yrs[  ]  iv) 20-30yrs[  ] v) >30yrs[  
] 
 
16. Have you fallowed (rested) this field before? 
i) Yes[  ]  ii) No[  ] 
 
University of Ghana http://ugspace.ug.edu.gh
150 
 
17. If yes 
why?.................................................................................................................................
. 
 
18. Do you plant trees to replace trees that have been taken down? 
i) Yes[  ]  ii) No[  ] 
 
19. Why? 
..........................................................................................................................................
. 
 
20. What common tree species do you use for fuel wood? 
............................................................... 
 
21. Would you seek to acquire land in a different area of this district ? 
i) Yes[  ]  ii) No[  ] 
 
22. What will you use this land parcel for? 
....................................................................................... 
23. What will you consider as the major drivers of land use change?  
(3OHDVHWLFN¥IRUHDFK) 
 Strongly 
agree 
Agree Strongly 
disagree 
Disagree Don't 
know 
Population      
Rainfall 
variability 
     
Poor Soil 
condition 
     
Socio 
economic 
(job, food) 
     
Other: 
......................................................................................................................................... 
University of Ghana http://ugspace.ug.edu.gh
151 
 
 
KNOWLEDGE ON CLIMATE CHANGE AND VARIABILITY 
24. Have you ever heard of climate change or are you aware of it? 
i) Yes [  ]      ii) No[  ] 
25. Please tick the number that you feel best represents your knowledge of climate 
variability and change? 
i) No knowledge[  ]     ii) Minimal knowledge[  ]    iii) Some knowledge[  ]     iv) 
Extensive knowledge[  ]       v) Expert knowledge[  ] 
26. Have you ever received any professional advice (e.g. from agricultural officers) 
regarding the effect of a changing rainfall patterns on crop yield?    i) Yes [  ]  ii) 
No[  ]   
27. If yes, what was that advice? 
....................................................................................................... 
..........................................................................................................................................
............ 
28. Have you noticed any changes in the climate (temperature and rainfall) over the years?
      i) Yes[  ]   ii) No[  ]   
29. What pattern do you experience in temperature now compared to some years ago? 
i). Increasing  ii). Decreasing  iii). Fluctuating   iv). No change  v). No 
Observation 
30. What pattern of rainfall do you experience now compared to some years ago? 
i) Increasing[  ]  ii) Decreasing[  ]  iii) Fluctuating[  ]   iv) No change[  ]  
v) No Observation[  ] 
31. Has the length of dry spells during the rainy season increased? 
i) Yes[  ]  ii) No[  ]  iii) can't tell[  ] 
32. When (month) does the rain begin? 
University of Ghana http://ugspace.ug.edu.gh
152 
 
33. In what month do you start planting? ................................................................ 
34. Has there been a change in the month of planting during the rainy season? 
1. Yes  2. No 
35. If yes 
Why?................................................................................................................................
.. 
36. Has there been a change in the length of rainy season over the years? 
1. Yes  2.  No  3. can't tell 
37. If yes what is the nature of change?  
1. Increasing  2. Decreasing  3. Fluctuating   4. No change  5. No Observation 
 
Impact of Change in rainfall and temperature  
38. Has there been a change in your crop yield?  
Yes[  ]  No[  ]  Don't Know[  ] 
 
39. How do you compare your crop yield situation with the past 10 years? 
1) Increasing [  ]   2) same [  ]    3)  declining [  ]  4. Other 
........................................ 
 
40. How is your crop yield situation affecting your affecting your standard of living? 
..........................................................................................................................................
............ 
 
 
41. What do you think are the causes of these changes in your crop yield? 
University of Ghana http://ugspace.ug.edu.gh
153 
 
i) Less rain[  ]  ii). high temperature[  ]  iii). low soil fertility[  ]  iv). i-iii [  ] 
      v). Other ....................... 
42. What can you say about these statements: Rainfall and temperature have (please tick) 
[1] reduced my crop yield {T} {F} {N/A} 
[2] increased my crop yield {T} {F} {N/A} 
[3] not changed my crop yield {T} {F} {N/A} 
[4] affected my soil negatively {T} {F} {N/A} 
[5] affected my soil positively {T} {F} {N/A} 
 
ADAPTATION STRATEGIES 
43. Have you made any changes to your farming practices because you have noticed 
changes in temperature and rainfall?    
i) Yes[  ]    ii) No[  ] 
44. If yes, what adjustments have you made to suit these changes and variability? 
i)....................................................................................................................................... 
ii)...................................................................................................................................... 
iii).……............................................................................................................................ 
45. Do you receive information on weather forecast or early warning to enable your 
household to respond to extreme climatic conditions in a timely fashion?  
i) Yes[  ]  ii) No[  ] 
Thank You.  
University of Ghana http://ugspace.ug.edu.gh
154 
 
APPENDIX 2: AVERAGE CLIMATE DATA VALUES FOR THE LOWER 
VOLTA BASIN 
Years $YJ7HPSHUDWXUH◦& Avg. Annual Rainfall 
(mm) 
1960 32.9 972.6 
1961 33.2 1134.8 
1962 33.1 1258.0 
1963 30.6 1272.8 
1964 32.7 1073.7 
1965 34.9 1100.8 
1966 31.8 1001.4 
1967 31.9 847.0 
1968 32.5 1250.4 
1969 29.2 797.3 
1970 29.4 867.1 
1971 33.2 858.5 
1972 32.8 715.1 
1973 33.4 937.9 
1974 33.4 841.6 
1975 33.1 730.7 
1976 32.2 787.9 
1977 30.0 657.2 
1978 30.6 758.5 
1979 28.3 764.2 
1980 33.5 864.5 
1981 33.8 996.9 
1982 32.8 946.2 
1983 32.6 633.3 
1984 32.1 878.5 
1985 32.9 865.3 
1986 32.6 671.5 
1987 32.2 896.5 
1988 32.0 791.3 
1989 32.2 806.9 
1990 32.2 687.8 
1991 33.2 832.6 
1992 33.4 634.8 
1993 33.5 979.1 
1994 33.7 803.5 
1995 33.6 847.1 
1996 33.3 1008.3 
1997 33.3 866.9 
1998 33.6 666.6 
1999 32.8 904.2 
2000 33.0 726.9 
2001 32.9 876.2 
2002 33.6 841.6 
University of Ghana http://ugspace.ug.edu.gh
155 
 
2003 33.4 763.3 
2004 33.2 791.4 
2005 33.4 716.0 
2006 33.5 960.4 
2007 33.4 1015.0 
2008 33.4 1124.9 
2009 33.5 878.6 
2010 33.8 878.8 
2011 33.2 818.2 
2012 33.4 663.4 
2013 32.4 794.1 
 
  
University of Ghana http://ugspace.ug.edu.gh
156 
 
APPENDIX 3: THIESSEN POLYGON ALGORITHM MAP SHOWING 
WEATHER STATIONS 
 
 
Source: Author’s Construct 
  
Legend 
Weather station 
University of Ghana http://ugspace.ug.edu.gh
157 
 
APPENDIX 4: REGRESSION RESULTS FOR CAUSALITY RELATIONSHIP 
BETWEEN LAND USE CHANGE AND DRIVERS OF LAND USE CHANGE 
(POPULATION, RAINFALL VARIABILITY, SOIL CONDITION AND 
SOCIO-ECONOMIC CONDITION) 
 
Variable Beta Std. Error t-value p-value 
(Constant) 1.705 0.12 14.08 0.00 
Population -0.120 0.34 -3.530 0.001** 
Rainfall variability 0.136 0.069 1.967 0.051* 
Poor soil condition 0.056 0.037 1.508 0.133 
Socio-economic(Job, 
food) 
-0.154 0.071 -2.164 0.032** 
Adj. R-Square=0.794 
           R=0.912                                         F-statistics= 21.549 
Note:* significant at p<0.1; ** significant at p<0.05 
 
  
University of Ghana http://ugspace.ug.edu.gh
158 
 
APPENDIX 5: CROP YIELD DATA VALUES FOR STUDY DISTRICTS AND 
CORRELATION TABLES 
ADA EAST DISTRICT 
Years 
ha 
(cassava) 
Cassava 
yield (MT) 
Yield/
ha ha (maize) 
Maize yield 
(MT) 
Yield/ha 
2005 2350 18800 8 1188 150.4 0.8 
2006 2900 20010 6.9 1213 170.4 0.8 
2007 2942 20594 7 1286 228.8 0.8 
2008 3485 24395 7 1380 304 0.8 
2009 3526 25387.2 7.2 1388 310.4 0.8 
2010 2600 18200 7 1390 312 0.8 
2011 2603 18221 7 1425 340 0.8 
2012 2650 13250 5 1450 180 0.4 
2013 2915 20405 7 1374 336.6 0.9 
 
  ha (cassava) cassava 
ha (cassava) 1  
cassava 0.8209 1 
 
  ha (maize) Maize 
ha (maize) 1  
Maize 0.6535 1 
 
LOWER MANYA KROBO DISTRICT 
Years 
ha 
(cassava) 
Cassava 
yield (MT) 
Yield/
ha ha (maize) 
Maize yield 
(MT) 
Yield/ha 
2005 12520 159004 12.7 13602 23123.4 1.7 
2006 9190 117632 12.8 8990 14384 1.6 
2007 9194 117683.2 12.8 9840 16728 1.7 
2008 9198 131531.4 14.3 9812 18642.8 1.9 
2009 10001 156015.6 15.6 10640 19152 1.8 
2010 10998 186966 17.0 11102 21093.8 1.9 
2011 10770 215400 20.0 13650 28665 2.1 
2012 10850 217000 20.0 13751 28877.1 2.1 
2013 4350 87000 20.0 6045 12694.5 2.1 
 
  ha (cassava) cassava 
ha (cassava) 1  
cassava 0.63204 1 
 
  ha (maize) maize 
ha (maize) 1  
maize 0.9647 1 
University of Ghana http://ugspace.ug.edu.gh
159 
 
 
University of Ghana http://ugspace.ug.edu.gh