Watershed Reanalysis: Towards a National Strategy for Model-Data Integration

Abstract

Reanalysis or retrospective analysis is the process of re-analyzing and assimilating climate and weather observations with the current modeling context. Reanalysis is an objective, quantitative method of synthesizing all sources of information (historical and real-time observations) within a unified framework. In this context, we propose a prototype for automated and virtualized web services software using national data products for climate reanalysis, soils, geology, terrain and land cover for the purpose of water resource simulation, prediction, data assimilation, calibration and archival. The prototype for model-data integration focuses on creating tools for fast data storage from selected national databases, as well as the computational resources necessary for a dynamic, distributed watershed prediction anywhere in the continental US. In the future implementation of virtualized services will benefit from the development of a cloud cyber infrastructure as the prototype evolves to data and model intensive computation for continental scale water resource predictions.