In order to estimation the future drought we need "available soil water content (ASWC)". In the traditional models, there is only the term of "soil water" not the ASWC.
It is possible but you can choose the good software. If you use Modflow you can make two simulations. The first concern the hydrodynamic parameters and the second the characteristics of water or soil.
GCMs use water balance in estimating soil water component. Available soil water could be modeled as a submodule using moisture in the root zone, type of plant uptake, and water retention in the soil. I recommend coupling the GCM with crop models (like EPIC) to evaluate available soil moisture and potential for deficits during droughts.
Another option is to use the soil water time series and use empirical coefficients to arrive at available soil water based on your soil type and land use.
It is possible to simulate the “available-soil- water- content,” ASWC, in GCMs.
As I recall the term “soil water” is actually a first approximation to ASWC . . . . all the land-surface-grid points in the GCM were assigned a fixed ASWC to be used in "evapotranspiration" calculations. . .this seemed to suffice for “weather-forecasting.”
Would this first approximation suffice for a climate-prediction model?
The ASWC parameterization in a "climate" GCM should fit the purpose of the GCM not its shortcomings.