A Reduced-Adjoint Variational Data Assimilation for Estimating Soil Moisture Profile and Soil Hydraulic Parameters from Surface Soil Moisture Observations

Abstract:

Soil moisture and soil hydraulic parameters play an important role in the global water cycle and has an important impact on weather and climate, energy fluxes at the land surface, agricultural and irrigation management practices, and food production. Soil hydraulic parameters are by far the most important land surface parameters to govern the partitioning of soil moisture between infiltration and evaporation fluxes at a range of spatial scales. Soil moisture is highly variable in space and time owing to the dynamics in soil hydraulic properties, precipitation, vegetation cover and topography. Therefore, accurate estimation of soil moisture pattern and the unknown hydraulic parameters of the soil is of critical importance for land surface and land- atmosphere interaction modeling. Satellite based surface soil moisture observations (upper few centimeters of soil column) can be obtained globally with well-defined spatial and temporal resolutions. These data however do not provide information on soil moisture through depth, as required by the land surface models. On the other hand, to estimate soil hydraulic parameters field measurements are required, which in large scale needs extensive measurements. In this work, the potential of using surface soil moisture measurements to retrieve soil moisture profile and the soil hydraulic parameters will be explored in a synthetic study, using reduced-order variational data assimilation and a 1D soil water model (HYDRUS-1D model) that simulates soil water dynamics. Proper orthogonal decomposition (POD) is a model reduction technique, which is used to approximate the gradient calculation in Variational Data Assimilation (VDA). Two distinct approaches are explored in this study when using POD in the VDA. In the first approach, an optimization algorithm is applied in order to minimize cost function entirely in the POD-reduced space. The second approach uses POD to approximate only the adjoint model. The accuracy and feasibility of the proposed approaches to retrieve initial soil moisture profile and finally estimate the unknown parameters of the soil will be investigated in this study. Our findings will help to assess the value of periodic spaceborne observations of topsoil moisture for soil moisture profile and soil hydraulic parameter estimation.