Evapotranspiration and Recharge Mapping Using Land Integrated Data Assimilation (LIDA-2) System and Quantifying Their Dependencies on Land/Atmosphere Variables

Evapotranspiration and recharge constitute a major portion of terrestrial water cycle thus influencing agriculture, water supply, weather, and climate predictions. Ground observations of these fluxes are sparsely available and could not be used as accurate representation of large areas due to land surface heterogeneity and dynamic nature of transfer process. Development and implementation of innovative techniques to estimate these fluxes using widely available remotely sensed data are essential for large scale study of their dynamics and accurate assessments of the water resource availability. The developed LIDA-2 is an observation driven variational data assimilation system to estimate the key parameters (evaporative fraction, bulk heat transfer coefficient, Brooks-Corey parameter) of evapotranspiration and recharge fluxes by assimilating GOES land surface temperature and SMAP surface soil moisture observations into a coupled water and dual-source energy balance model. Second order information is used to estimate the uncertainty and guide the model toward a well-posed estimation problem. The algorithm is implemented in part of the US southern great plain (SGP) and Oklahoma panhandle regions to map evapotranspiration and recharge in this region and its performance is evaluated through comparison tests using available ground measurements datasets. Soil moisture estimations are validated against in-situ observations from USCRN and MESONET sites. Evapotranspiration estimates are validated using land surface fluxes from Atmospheric Radiation Measurement (ARM) stations. Both soil moisture and evapotranspiration estimates demonstrate significant improvement from open loop estimates due to data assimilation. The annual recharge map is compared with the maps from literature and showed good agreement in spatial patterns. Estimated annual recharge values are also within the historical range as reported in the literature. The monthly recharge estimates follow the trend shown in the groundwater wells data from Oklahoma Water resource Board (OWRB) and USGS. Two applications of these high-resolution dataset generated from LIDA-2 are demonstrated in this study. First, evaporative fraction and soil moisture estimates are used to study the impact of land surface heterogeneities (i.e. vegetation, soil type) on the coupling strength of the terrestrial leg of the land-atmosphere interaction in southern great plains. Second, explicit easy to use equations of the evapotranspiration and recharge fluxes based on their major atmospheric and land/environmental drivers are derived using Genetic Expression Programming (GEP) to aid the water resource managers and decision makers to better monitor these fluxes. The use of parsimonious models of water and energy balance and widely available data along with the quality of indirect methods of evaluation in assessing the accuracy, adds to the uniqueness of the approach by making it applicable to data scarce regions. The proposed framework can be used as a tool to map evapotranspiration and recharge and assess water availability at regional scale, in different regions of the world.