Hydrology is experiencing rapid changes as a result of improved scientific understanding and measurement technology.  Advances in global modeling and remote sensing are likely to provide large amounts of new information in the coming decades.  There will be an increased need for efficient methods to process and interpret all of this information.  Many of the most promising data processing options combine observations with model predictions, a process commonly known as data assimilation.   The data assimilation methods which have been most successful in practical applications are based on either variational or recursive estimation concepts.  Each of these approaches has distinctive advantages and limitations but neither provides a satisfactory solution for very large applications (e.g. applications which work with large amounts of remote sensing data over continental-scale regions).

In this project we propose to develop a computationally efficient and robust approach to hydrologic data assimilation which combines the best aspects of variational and recursive estimation.   This work will be methodological in nature but its overall goal is to advance scientific understanding of large-scale hydrologic processes.  The data assimilation methods we develop in this project will be tested on a case study which will provide insight about scientific questions of hydrologic interest. The testing and application phase of our project will rely on our previous experience with data assimilation techniques and on methods which have been successfully applied in meteorology and oceanography.