Authors: Jonathan A. Quebbeman – RTI International; David Yates and Tom Hopson – NCAR
Title: Extended Hydrologic Prediction Project (EHP) – a multi-model sub-seasonal forecasting system in India
Abstract: Funded through the World Bank National Hydrology Project in India, RTI is leading the development of a new operational sub-seasonal (4-week) forecasting system in three major basins of India with the Central Water Commission. As part of the development process, RTI has developed and tested four separate modeling methods including a) Lumped-Conceptual model [WEAP], b) Spatially-distributed model [WRF-Hydro], c) Statistical Model [AR-based], and an AI model [LSTM]. Each model was validated at twelve locations using a range of performance metrics to determine when and where a given approach provides maximum skill. This lightning talk will introduce the EHP project and highlight the results, challenges, and conclusions of the multi-model evaluation for forecasting system development.