WinDB2, a Geospatial Database for WRF

The Wind Database (WinDB2) is an open source, geospatial database for environmental data. The project contains an extensible PostgreSQL/PostGIS database schema, along with WRF post-processing tools for field extraction and height interpolation written in Python. Initially developed to store years of high-resolution WRF wind fields, the project is being generalized to store many types of WRF variables. WinDB2 is currently being used as the backend for a daily, operational 200-m resolution wind forecast for San Francisco Bay. Hydrodynamic tidal model output and a plethora of observational data have also been stored and analyzed using WinDB2. WinDB2 also solves the problem of reprojecting WRF data into nearly any other map projection with little effort.

Some ideal use cases for WinDB2 are: 1) web mapping using Geoserver, a power web-mapping utility capable of creating custom web-mapping tiles for time-varying model data, 2) wind energy resource assessments that require the storing of many years of WRF wind fields at various heights over a wind farm or region, 3) storing many years of WRF data in a single database, with the ability to write simple SQL queries for analysis that can span years of data, 4) easily connecting analysis computers to large amounts of WRF data on remote-computing resources by running a simple Python script, and 5) cost-effective, long-term archival of WRF output, as WinDB2 allows you to specify a masked region-of-interest via a Shapefile (e.g. a wind farm, field site, observation locations, over water/land, etc...).

This paper gives an overview of the capabilities of WinDB2 for rapid analysis and mapping, and shows multiple examples of the utility of storing your environmental data in a geospatial database.