Abstract:
Efficient use of water resources for irrigation requires an accurate estimation of crop water requirement. Understanding changes in evaporation rates is expected to be of great importance for water resource planning and management. This study estimate the crop water use of rice and cassava crops using four ET models (Samani-Hargreaves, Jensen-Haise, Blaney-Morin (BMN) and FAO-56 Penman-Monteith) in Lokoja, Nigeria. In addition, the study also estimates crop water requirements of Cassava and Rice. A forty years (1997-2010) climatic data, used in computing reference evapotranspiration (ETo) for Lokoja, were obtained from Nigeria Meteorological Institute (NIMET), Lokoja. A non-parametric Mann-Kendal trend and Sen’s slope tests were performed on the respective meteorological variables. Also, five statistical measures were used to test the efficiency of the models. The results obtained during the study showed that annual rainfall showed an annual increase of 5.2 mm/yr2. Monthly and inter-annual sunshine hour significantly reduced for the period examined. Maximum temperature at Lokoja was found to increase slightly while minimum temperature and mean temperature increased significantly (p < 0.001). Vapour pressure increased significantly (p > 0.01) while relative humidity showed no significant change. Compared to FAO-56 valurs, BMN model yielded the best predicting equation with the lowest root-mean-square-error (RMSE) and Mean Bias Error (MBE) of 0.72mmday-1 and 0.52mmday-1, respectively. Samani-Hargreaves (SH) model on the other hand gave the highest RMSE and MBE values of 1.14mmday-1 and 1.29mmday-1, respectively. The relative error (RE) estimate shows that BMN model over-estimated the ET0 values by 11% while SH model overestimate it by 34% and Jensen-Haise (JH) under-estimated the value by 22%. Monthly conversion factors were derived for the models in relation to FAO-56 values. Based on BMN estimates, the seasonal water requirements of Rice and Cassava for the study area are estimated.
