Abstract
The soil moisture sub-model incorporated in CASA model (Potter et al. 1993), estimating net primary production (NPP), is validated with soil moisture observations at a dryland site in Inner Mongolia. The results showed that the soil moisture sub-model of the CASA model did not simulate absolute values and seasonal changes observed in Inner Mongolia. In order to simulate a realistic soil moisture estimate, the time interval on which the model is run to be daily instead of monthly with soil parameters from field experiments. The soil moisture model proposed by present study will be useful for estimating better water stress for precise NPP.