Indicators of Soil Physical Quality: From Simplicity to Complexity

Abstract

Soil moisture and soil water storage play a significant role in lumped and distributed hydrological simulation, both for model initialization and in later time steps to control and to correct model performance. On the other side, rainfall-runoff models still need to be improved to simulate reasonably well the vertical exchanges of heat and water between the soil and the atmosphere, that may result in inconsistent soil moisture fields. Therefore, many issues remain to be adequately addressed, such as how to include new data sources as well as how to improve methods for calibration, validation, parameterization and upscaling of hydrological models. This work focuses on relatively new data sources, addressing the supply of soil moisture (Soil Moisture Experiment 2003 – SMEX03) and soil water storage information (GRACE – Gravity Recovery and Climate Experiment) to a typical lumped rainfall-runoff model (SMAP – Soil Moisture Accounting Procedure running at daily and monthly steps) from in situ measurements and remotely sensed imagery for the São Francisco and Amazon basins in Brazil. In particular, a sensitivity analysis is conducted to evaluate jointly both hydrological simulations and data collected and acquired for the studied areas highlighting soundly based and good results and also pointing out some of the challenges to be faced in the near future. We should mention that much more work on soil physics is still necessary for applications regarding rainfall-runoff models to work properly at the watershed scale.