PERFORMANCE OF TWO AGRO-HYDROLOGICAL MODELS IN SIMULATING SOIL WATER BALANCE OF A RAINFED MAIZE FIELD

PERFORMANCE OF TWO AGRO-HYDROLOGICAL MODELS IN SIMULATING SOIL WATER BALANCE OF A RAINFED MAIZE FIELD

Abstract:

Roles of water balance components in any agro-ecological system are indispensible for most physical and physiological processes within the soil-plant-atmosphere system. The performance of two agro-hydrological models: the Soil-Water-Atmosphere-Plant (SWAP) and Irrigation Scheduling Impact Assessment model (ISIAMOD) in simulating soil water balance components (SWBC) of a cropped field under rainfed condition, was studied in a sandy clay loam soil at the Research field of the Department of Agricultural Engineering, Ahmadu Bello University, Samaru, Nigeria. A field experiment consisting of nine sets of weighing-type mini-lysimeter installed in a field of size 0.053 ha was used in the study. The experiment consisting of three treatments replicated three times. The three treatments comprise of cropped lysimeter set-up covered with plastic mulch (polythene) to measure transpiration (T) process, no-mulch cropped lysimeter setup to measure evapotranspiration (ET) process and no-cropped lysimeter set-ups to measure the process of evaporation (E). The lysimeters and the surrounding field were planted with Sammaz-28 Maize variety. The components of soil water balance which include ET, T, E, runoff (Roff) and deep percolation (Dp) were measured directly from the lysimeters. To complement the measurement of ET from the lysimeter, soil moisture depletion study was also carried on the field by installing gypsum blocks at different depths to measure soil moisture. The results showed that the performance of the two models in simulating soil water balance components as compared to the field measured values was satisfactory based on the outcome of the statistical indicators used. The statistical indicators used to compare the performance of the models are coefficient of residual mass (CRM), modelling efficiency (EF) and root mean square error (RMSE). CRM showed that ISIAMOD has the tendency of underestimating the ET, T, and Ecropby a value which ranges from 2.5 to 6.0% while SWAP has the tendency of overestimating the same components which ranges from 2.0 to 9%. The modeling efficiencies of the two models range from 84 to 90%, except for evaporation processes which ranges from of 54 to 62%. The RMSE of the two models ranges from 0.29 to 0.86. They both simulated the seasonal run-off and drainage well. The results show that two models can be used for determination of soil water balance components of cropped soil and for analyzing a better water management option for agricultural production.

PERFORMANCE OF TWO AGRO-HYDROLOGICAL MODELS IN SIMULATING SOIL WATER BALANCE OF A RAINFED MAIZE FIELD