A factorial study of the energy and moisture transfer processes at the land surface

Abstract

Interactions of the energy and moisture transfer processes at the land surface are investigated using factorial experimentation on a simple non-vegetated model. The model is constructed from energy and moisture balance equations at the land surface. Parameterizations from general circulation models are used to describe the model components. The factorial analysis is performed on two differently parameterized versions of the model in both steady-state and dynamic models. Results indicate that the runoff parameterization plays a critical role in determining soil moisture. Also, the inclusion of a second soil layer appears to have an attenuating effect on the model's temperature and soil moisture response times, whereas the choice of top layer depth substantially affects the diurnal surface temperature variation. It is also found that the temperature response time is correlated with and a strong function of the moisture response time. These results are useful in understanding the variation in regional climate predictions documented in recent general circulation model studies. In addition, factorial designs are shown to be an effective tool for analyzing individual parameter sensitivity and parameter interactions in models with a large number of factors.