Verification Tools for Air Quality Models

Environment Canada's VAQUM (Verification for Air QUality Models) system was built to evaluate the performance of air quality model forecasts. It is based on PostgreSQL, an open-source object-relational database management system, and PostGIS, which adds support for geospatial data. This system has been under development since 2008 and there have been a number of new products developed in the last few years.

Air quality model forecasts for the surface level are extracted from output files at specific locations where air quality observations are available. Similarly, observations are added to the database. The system is then able to produce a variety of regional or global analyses with georeferenced data pairs of observations and forecasts that are coupled when needed. Some of the verification products now available from VAQUM include statistics based on hourly forecasts and daily maximum forecasts, categorical scores, statistics per forecasted hour, statistics by observed value category, maps of per-station statistics, and metropolitan area time series. In this poster we will present the essential VAQUM products used for the near-real-time evaluation of Environment Canada's air quality forecast systems. Current developments and future plans will also be described.