A Land Data Assimilation System for Mapping of the Linkage between the Terrestrial Water, Energy and Carbon Cycles

The exchange of heat (energy), water and carbon between the terrestrial (land) biosphere and atmosphere play a key role in the Earth’s past and future climate. The rate of exchange of heat (energy), water and carbon in and out of the land surface, known as land surface fluxes, determine the local temperature, moisture states and carbon exchange with the atmosphere. The terrestrial or land component of water, energy and carbon cycles are strongly coupled and operate in a consistent manner. All Land Surface Models (LSMs) used in hydrologic, ecological and climate models - directly or indirectly - account for the linkages between the terrestrial cycles. How various models perform is highly dependent on how these linkages are represented in the LSMs. Lack of proper representation of these linkages currently results in wide range of uncertainties and variations in regional simulated land surface fluxes, climate, and climate projections. There is a gap in real-world (true) representation of the linkages/ coupling between the terrestrial water, energy, and carbon cycles at regional scale. The primary limitation is the lack of direct observation of key variables (land surface state and fluxes) that can quantify these linkages with the required spatial and temporal resolution and continuity. Our study focuses on development of novel observation-driven approaches to indirectly capture/map these linkages from implicit information contained in land surface state observations (i.e., surface soil moisture, temperature, and vegetation index) which are widely available from remote sensing and across a range of spatial and temporal scales. Specifically in this study a Land Integrated Data Assimilation (LIDA) Framework is introduced that assimilates the GEOS land surface temperature and SMAP surface soil moisture observations into a coupled water and energy balance model to estimate the key parameters of turbulent heat and moisture fluxes. An uncertainty quantification algorithm is introduced which uses second order information to estimate the uncertainty of the estimates, guide the model toward a well-posed estimation problem and evaluate the accuracy of estimation. Direct (comparison with in-situ measurements) and indirect measures are used to evaluate the accuracy of the estimation. The method is implemented over Southern Great Plain and Part of Oklahoma Panhandle region and the mapped states and fluxes are then used to map the key link between the terrestrial cycles (soil moisture control on evapotranspiration), across different biomes, soil types and seasons.