GIS-based NEXRAD Stage III precipitation database: automated approaches for data processing and visualization

This study develops a geographical information system (GIS) approach for automated processing of the Next Generation Weather Radar (NEXRAD) Stage III precipitation data. The automated processing system, implemented by using commercial GIS and a number of Perl scripts and C/C + + programs, allows for rapid data display, requires less storage capacity, and provides the analytical and data visualization tools inherent in GIS as compared to traditional methods. In this paper, we illustrate the development of automatic techniques to preprocess raw NEXRAD Stage III data, transform the data to a GIS format, select regions of interest, and retrieve statistical rainfall analysis over user-defined spatial and temporal scales. Computational expense is reduced significantly using the GIS-based automated techniques. For example, 1-year Stage III data processing (similar to9000 files) for the West Gulf River Forecast Center takes about 3 days of computation time instead of months of manual work. To illustrate the radar precipitation database and its visualization capabilities, we present three application examples: (1) GIS-based data visualization and integration, and ArcIMS-based web visualization and publication system, (2) a spatial-temporal analysis of monsoon rainfall patterns over the Rio Grande River Basin, and (3) the potential of GIS-based radar data for distributed watershed models. We conclude by discussing the potential applications of automated techniques for radar rainfall processing and its integration with GIS-based hydrologic information systems. (C) 2004 Elsevier Ltd. All rights reserved.