Performance of simplified water balance models under maize in a semiarid subtropical environment

Performance of the soil-plant-atmosphere-water (SPAW) model and water balance model (WBM) was assessed in simulating water balance components under maize. The two models differed in their structure and input data requirements. A few modifications involving the estimation of green canopy based on the SPAW model procedures were used in a modified WBM and its performance was also tested. Simulated sum of interception evaporation, soil evaporation, plant transpiration and deep drainage from seeding to harvest with the WBM, modified WBM, and SPAW model matched well with the measured sum of profile water depletion, rain and irrigation. The rms of deviations of the simulated sum were 3, 4 and 3 cm for measured values ranging between 42 and 123 cm. Estimates of drainage with the WBM and modified WBM using empirical coefficients agreed well with those calculated using the SPAW model. These results suggest that drainage at water contents below field capacity needs to be included in water balance studies. Soil evaporation was overestimated with the three models, more so with the SPAW model, as indicated by comparison with measured values from column studies. The slope of the relation between above-ground biomass measured at harvest and estimated transpiration with the three models (transpiration efficiency) varied between 0.44 and 0.51 t ha(-1) cm(-1). It suggests that the performance of the modified WBM and SPAW model was comparable to the WBM which requires measured values of leaf area index to infer green canopy. Thus, the modified WBM and SPAW model, which have a mechanism for estimating green canopy, are more useful in assessing water balance components in cropped soils.