I-Corps: Real-Time Big Data Based Decision Support System for Water Use

San Jose State University Foundation, San Jose CA

Investigators

Abstract

The broader impact/commercial potential of this I-Corps project is to provide accurate and timely precipitation forecasts for water resource management during rainfall and hydrological extremes including intense flooding and droughts events. This I-Corps team has developed a unique high resolution weather forecasting model using a high-performance computer framework that employed all available local weather observations and state-of-the-art local doppler radar data to improve local precipitation forecasts on a short-term basis. In conjunction with precipitation forecasts, our system will support water resources related decision-making. A web-based service to provide these pinpoint rainfall forecasts and other related hydrological data products will be available. Operations managers of businesses and government agencies (i.e., water districts, agricultural farms, insurance companies, energy companies) who rely on very accurate and precise rainfall forecasts to make economic-based decisions will be able to have quick access to customized rainfall forecasts and hydrological products at their user-defined locations. During short-term hydrological extremes, these pinpoint accurate rainfall forecasts and products will improve the efficiency and decision-making abilities within the companies and agencies themselves, and it will help them to better communicate their decisions to their end-users and the public. This I-Corps project will employ a high-resolution weather forecasting model. It can have a horizontal resolution of 1 km or finer and uses specific 3D-Variational data assimilation methods which incorporate all available real-time weather observations from multiple sources. These input sources include satellite data (e.g., GOES-R), weather station and rain gauge data, quantitative precipitation estimates from local doppler radars, and other global and regional forecasting models (e.g., NOAA?s High Resolution Rapid Refresh and North American Model). The model can be executed separately for any specific region and location of interest to the end-user, can extract pinpoint rainfall forecasts, and can be updated every 3 or 6 hours depending on the circumstance. Other hydrological products available from the project that can also assist in decision support such as drought conditions (e.g., soil moisture and drought severity indices) and water levels of local reservoirs. It is anticipated that the products from this high-resolution modeling and big-data system will provide guidance to support water management and consumption decisions for government agencies and the private sectors.