Collaborative Research: ITR: Framework for Optimal Merging of Multi-sensor Spatial Data and Multi-model, Multi-analysis Ensemble Forecasts of Heavy Precipitation and Floods

Harvard University, Cambridge MA

Investigators

Abstract

The aim of this project is to integrate satellite and ground-based data with numerical weather prediction models, hydrological models and neural network models of precipitation and flooding, in order to develop techniques for forecasting heavy precipitation and flooding that are tailored to local regions. The focus regions will be river basins in Appalachia, Nepal, Mozambique and Brazil where there are significant orographic influences on precipitation. The project is a collaboration between a hydrologist and a meteorologist and involves support personnel with expertise in information technology. The work combines scientific computing and machine learning techniques with meteorology and hydrology. The forecasting of heavy floods in areas adjacent to mountainous regions is an important problem in many places around the world. If successful, the research would develop an approach that could then be adapted to operational use for such forecasting.