H51G-1469
Data Assimilation of Unsaturated Flow: Considering the Vertical Measurement Scale of Soil Moisture

Friday, 18 December 2015
Poster Hall (Moscone South)
Qiuru Zhang and Liangsheng Shi, Wuhan University, Wuhan, China
Abstract:
Soil moisture data measured by various techniques has been widely used in unsaturated flow data assimilation, while the effect of vertical measurement depth (or scale) of soil moisture on model prediction and parameter estimation has not been quantitatively investigated. This study adopts the ensemble Kalman filter (EnKF) to assimilate surface soil moisture data of different observation depths into a one-dimensional variably-saturated water flow model, to estimate soil hydraulic parameters and soil moisture profiles. Through a few synthetic experiments, two factors are considered: soil heterogeneity and upper boundary conditions. Results show that as the heterogeneity of soil increases, the affecting depth of surface soil moisture decreases, and the observation with a larger depth is advantageous. For homogenous soil, soil moisture data with a smaller scale leads to better estimation of soil hydraulic parameters and soil moisture profiles. For heterogeneous soil, surface soil moisture exerts less or no influence on deeper layers. It is thus not able to estimate multiple parameters simultaneously by only assimilating surface soil moisture data in heterogeneous soil. It is found that regardless of the measurement scale, the estimated average soil moisture within the measurement scale matches well with the measurements at the corresponding scale. The accuracy of soil moisture profiles retrieval and soil hydraulic parameter estimation also depends on the upper boundary conditions. Flow regimes controlled by upward fluxes are found less successful, as the observed surface data may be no longer useful to infer the hydrologic processes in the deep soil, and, the observation with a larger scale tends to be more useful to reveal the relationship between surface soil moisture and deeper layer soil moisture. Our results show that the soil moisture with smaller depths is more effective to estimate soil moisture and soil hydraulic parameters of surface layer, while the soil moisture with larger depths is more helpful to interpret the correlation between surface soil moisture and deeper layer soil moisture.