Utility of Assimilating Surface Radiometric Temperature Observations for Evaporative Fraction and Heat Transfer Coefficient Retrieval

Recent advances in land data assimilation have yielded variational smoother techniques designed to solve the surface energy balance based on remote observations of surface radiometric temperature. These approaches have a number of potential advantages over existing diagnostic models, including the ability to make energy flux predictions between observation times and reduced requirements for ancillary parameter estimation. Here, the performance of a recently developed variational smoother approach is examined in detail over a range of vegetative and hydrological conditions in the southern U.S.A. during the middle part of the growing season. Smoother results are compared with flux tower observations and energy balance predictions obtained from the two-source energy balance model (TSM). The variational approach demonstrates promise for flux retrievals at dry and lightly vegetated sites. However, results suggest that the simultaneous retrieval of both evaporative fraction and turbulent transfer coefficients by the variational approach will be difficult for wet and/or heavily vegetated land surfaces. Additional land surface information (e.g. leaf area index (L_AI) or the rough specification of evaporative fraction bounds) will be required to ensure robust predictions under such conditions. The single-source nature of the variational approach also hampers the physical interpretation of turbulent transfer coefficient retrievals. Intercomparisons between energy flux predictions from the variational approach and the purely diagnostic TSM demonstrate that the relative accuracy of each approach is contingent on surface conditions and the accuracy with which L_AI values required by the TSM can be estimated.