Using Dssat-Ceres Maize Model to Estimate District-Level Yields in Northern Ghana

Abstract

Maize farmers accrue low yields due to poor weather, lack of certified seeds and farm inputs. The crop cutting method used by MoFA in yield estimation is tedious, labour intensive and most information are not adequately captured for decision making and policy. This study explores an additional method of scaling farmer-level yield to Tolon District level yield. The DSSAT-CERES maize model was used to simulate farmer-level yields where weather, soil and crop management data from selected sites in Tolon District were used. The simulated farmer-level yields were compared to the observed yields using a 1:1 plot, R2 (0.82) and Wilmott-d index (0.95). The validation confirmed the goodness of fit. To estimate the district yields, an aggregation procedure was used. Three different fertilizer rates and three sowing dates were interrelated then obtained as a weighted sum of management categories. The probability distribution of the District-level simulated yield was compared with the observed using cumulative frequency distribution. The trends agreed well. Analysis of the long-term climate in Northern Ghana showed that the frequencies of hot/dry regimes increased in the more recent periods. The validated model was used to investigate the usefulness of improved management such as addition of 3,000 kg/ha of farm yard manure to improve fertility, reduce run off and increase the water holding capacity of soil to increase yield. The results showed the mean District yield of 1935 kg/ha and 1482 kg/ha with application of high inputs (HI) under improved and under normal management respectively. Mean yield of 1386 kg/ha and 1190 kg/ha were obtained for improved and for normal management respectively with application of medium inputs (MI). Under low inputs (LI), the mean yields of 415 kg/ha and 225 kg/ha were obtained for improved and for normal management respectively. The District yield under improved management practices were 30.6% higher than under normal management. The yield gain is attributed to addition of appropriate inorganic fertilizers. Information on climate variability are crucial guiding principles for all stakes in sustainable future planning to enhance food security.