Optimization of Soil Texture to Improve the Soil Moisture in the Land Surface Model

Soil moisture is a key variable in the hydrologic cycle and affects weather and climate; thus, accurate soil moisture prediction is essential in the land surface modeling. In this study, we have used a land surface model, called the University of Torino land surface Process model for Interaction in the Atmosphere (UTOPIA), to predict the soil moisture. The UTOPIA is a one-dimensional model representing the interactions among atmosphere, land surface, vegetation and soil layers. Being UTOPIA a multilayer soil model, the user can discretize the soil into a certain number of layers: each layer needs specific physical properties related to soil moisture and temperature depending on the soil texture type. The soil texture allows to infer a useful information about the soil, such as the wilting point, field capacity, heat capacity, and eventually soil moisture. However, it is hard to obtain the accurate information of soil texture, especially in deep soil layers, due to insufficient and/or uncertain observation. Therefore, we have implemented the micro-genetic algorithm (micro-GA) within UTOPIA to optimize the soil textures by comparing the model-generated soil moistures versus the in-situ observations. The micro-GA is a global optimization algorithm based on the natural selection or survival of fitness to evolve the best potential solution. As a preliminary result, we anticipate that the optimal soil textures within the multiple layers lead a substantial improvement on soil moisture prediction. Furthermore, we will investigate the changes in latent and sensible heat fluxes which can be affect to the atmosphere model from the soil moisture improvements.