Variational Assimilation of Land Surface Soil Moisture Observations for the Estimation of Key State and Parameters of the Diffuse Recharge Flux
Open AccessDiffuse recharge flux represents a significant source of groundwater and has been widely assessed worldwide for water resources management. Diffuse recharge refers to water from precipitation that infiltrates and percolates through the unsaturated zone. Groundwater is usually the main source of water supply in arid and semiarid regions, therefore, to understand the vulnerability of the water resources, for efficient and sustainable management of groundwater, and to preserve the natural environment, an accurate assessment of diffuse recharge flux is crucial.Currently, available methods for recharge estimation are either direct or indirect. Lysimeters, stream gaging, and chemical tracers are methods for directly measuring recharge. While the first two techniques provide specific measurements, they may not accurately represent the overall conditions on a large scale, and installing a network of these instruments for mapping purposes may not be feasible. The chemical tracer method is capable of providing recharge estimates for extensive areas, but it is labor-intensive and can only be carried out experimentally. This approach is not scalable for large-scale mapping or providing global estimates. Indirect methods for the estimation of recharge mainly rely on linking recharge to other measurements (e.g., precipitation, stream discharge, etc.) using models, such as the zero-flux plane (ZFP) method and the water balance methods. The Zero-Flux Plane (ZFP) is a plane that separates two zones in the soil where water is moving upwards and downwards simultaneously. The ZFP method requires a relatively high density of monitoring wells, which makes it a relatively expensive technique. In the water balance method recharge has been estimated indirectly as the residual of the surface or subsurface water balance. A major disadvantage of the water balance approaches is that the accuracy of the estimated recharge relies on the accuracy of other components of the water balance equation. Thus, even slight inaccuracies in those variables can lead to uncertainties in the recharge rate. One widely used numerical model for simulating the movement of flow in the unsaturated porous media based on water balance equation and darcy’s law is the HYDRUS-1D model. Hydrus-1D is used to simulate soil water movement and estimate various components of water balance, including infiltration, soil evaporation, transpiration, and diffuse recharge. HYDRUS-1D requires fewer input parameters and have been shown to provide satisfactory results by several studies, especially when provided with accurate initial soil moisture profile, soil hydraulic parameters and forcing data. However, one major limitation of the current suite of indirect approaches, including Hydrus-1D model is their lack of capability to develop spatial mapping. In the proposed research a state-of-the-art data assimilation technique based on a low/ reduced order variational approach is introduced to combine the horizontal coverage and spatial resolution of remote sensing of near surface soil moisture data with the vertical coverage and temporal continuity of a soil moisture simulation model (HYDRUS-1D) to produce estimates of the effective soil hydraulic properties and soil moisture profile with minimal uncertainty. Given maps of effective soil hydraulic parameters and the profile of soil moisture, spatial patterns of diffusive recharge flux are estimated. The proposed framework for estimation of the effective soil hydraulic parameters and soil moisture profile (hence groundwater recharge) is tested at the point scale (through a set of synthetic and field site experiments) and at the large scale (over an area of 11,340 km2 in the U.S. Southern Great Plains) using SMAP (Soil moisture Active Passive) observation and reanalysis forcing data. Results demonstrate the capability of proposed technique to map groundwater recharge, soil moisture and other water balance components with good accuracy across a range of spatial and temporal scales and shows promise in using this technique to conduct surface-subsurface studies at large scale.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.