Advances in Parameter Estimation and Data Assimilation for Hydrologic Modeling

In the past two decades, the availability of new data sources (particularly remotely sensed information) and improved computational tools has resulted in significant developments in the field of hydrologic modeling, from simple flow models to complex numerical models which simulate the coupled behavior of multiple fluxes (hydrologic, chemical, energy, etc.). At the same time, significant improvements have taken place in the data assimilation and parameter estimation methods. Although the increasing complexity of models has outpaced the development of appropriate systems identification methodologies, there is a need to design models that are properly constrained by observational data. Independent and collaborative research efforts by various groups worldwide have led to improved modeling techniques, optimization methods for parameter estimation, methods for estimating predictive uncertainty, and methods for evaluating the relative merits of competing models. This talk will review some of the key developments during the past 20 years and speculate on future directions.