A framework for the specification of modeling and observation uncertainties for land data assimilation of remote sensing retrievals

Authors: Crow, Wade; Bolten, John

Submitted to: American Meteorological Society
Publication Type: Abstract Only
Publication Acceptance Date: 12/20/2007
Publication Date: 1/23/2007
Citation: Crow, W.T., Bolten, J.D. 2007. A framwork for the specification of modeling and observation uncertainties for land data assimilation of remote sensing retrievals [abstract]. American Meteorological Society Meeting. 2007 CDROM.

Interpretive Summary:

Technical Abstract: Recent advances in the development of sequential land data assimilation techniques have demonstrated that remote sensing observations of surface soil moisture can improve the dynamic representation of root-zone soil moisture in hydrologic models. However, much of the available evidence is based on identical twin experiments using synthetically generated, and artificially perturbed, measurements. These experiments, while extremely useful diagnostic tools for evaluating filter efficiency, typically simplify or avoid a number of key complexities facing operational efforts to assimilate spaceborne observations. One typical assumption in synthetic experiment is that magnitude of both modeling background and observation errors are known in a statistical sense. In reality, adequate error information is almost never available in operational settings and the miss-speciation of error parameters can easily lead to filter divergence. Fortunately, the statistical analysis of filter innovations, defined as the observed difference between model predicted and actual observations, provides a valuable tool for diagnosing the miss-specification of model error. Several on-line procedures for estimating model error parameters - based on the analysis of model innovations - have been introduced for geophysical models. However, they have not been widely applied to hydrologic models. Using a series of synthetic twin experiments and the statistical analysis of filtering innovations, this paper will present a framework for diagnosing the magnitude and auto-correlation of both observation and background modeling error in hydrologic models and/or hydrologic remote sensing retrievals. The potential benefits of applying such tools to the Ensemble Kalman filter-based assimilation of remotely sensed soil moisture into a land surface model will be examined.