Simultaneous Estimation of both Hydrologic and Ecologic Parameters in the New Eco-hydrological Model Assimilating Microwave Signal

Dynamics of both soil moisture and vegetation are very important because these variables affect an atmospheric circulation through land-atmosphere interactions and our environmental resources. To model soil moisture and vegetation dynamics accurately and simultaneously, parameter optimization in the complicated hydrologic-ecologic coupling model remains an issue. We develop the new system that consists of a land surface model (LSM) to estimate soil moisture, surface runoff, sub-surface runoff, groundwater dynamics, and turbulent fluxes, a dynamic vegetation model (DVM) to estimate vegetation growth and death, and a radiative transfer model (RTM) to estimate the microwave brightness temperature. In addition, the auto-calibration system assimilating microwave signal from AMSR-E is developed. The system minimizes the difference between simulated brightness temperature and observed brightness temperature by adjusting system parameters. The system is validated at 14 in-situ observation stations including Japan, China, Mali, Niger and the United States. Landuse types of these stations are cropland, grassland, or bare soil. Firstly, we confirm the sensitivity of each parameter in the system using the variance-based sensitivity analysis and then the system is validated by in-situ observed soil moisture and satellite observed leaf area index (LAI) from the Moderate Resolution Imaging Spectroradiometer (MODIS). The sensitivity analysis shows both of hydrologic and ecologic parameters have some sensitivities to multi-frequency and multi-polarization microwave brightness temperatures and this analysis contributes to reduce the number of parameter we should calibrate. Optimized model can accurately simulate both of soil moisture and LAI at the same time in the stations which have various hydroclimatic conditions. This study suggests new auto-calibration system of the complicated model that includes precise hydrological and dynamic vegetation schemes by using microwave signal and it can be applied to other state-of-art land surface models.