Effective Soil Hydraulic Parameters at Multiple Scales Using Data Assimilation Schemes and Remote Sensing Data.

In recent years we have adopted a multi-facet approach for estimating effective soil hydraulic property at different scales using different modeling and assimilation techniques in conjunction with air/space based remotely sensed soil moisture data. In this presentation we will discuss: (1) a bottom-up approach, where larger-scale effective parameters are calculated from the Bayesian distribution of insitu hydraulic property measurements, remotely sensed soil moisture data, and Markov Chain Monte Carlo algorithm, (2) a top-down approach, where effective soil hydraulic parameters are estimated by inverse modeling with genetic algorithm using time series of remotely sensed soil moisture measurements, and (3) a neural network approach with bias correction, where effective soil hydraulic parameters were estimated by exploiting the correlations with soil texture, topographic attributes, and vegetation characteristics at multiple spatial resolutions. Numerical and experimental results using these various effective soil hydraulic parameter estimation approaches including some comparisons between the approaches will be presented for various hydro-climatic regions across the globe.