Simulated visibility at the Nile River based on particulate matter forecasts for Cairo, Egypt (facing south).
STI provided a range of services to develop PM2.5 forecasting programs for 22 cities that report forecasts to AIRNow and USA Today. We developed objective forecasting tools to predict 24-hr average PM2.5 concentrations, forecasted fine particles daily during the winter, and evaluated forecast performance. We also developed a web-based program, called AQForecaster, which guides air quality agency staff through the process of issuing PM forecasts. A Forecasting Resource Center was established that offered air quality forecasters phone and e-mail access to STI scientists and an informative web site. Through the Forecasting Resource Center, STI scientists performed case studies for a variety of regional PM2.5 events, assisted state and local agencies with forecast tools, and provided analysis support for the NOAA/EPA Air Quality Forecast Modeling effort. STI also designed, developed, and taught forecasting workshops at the EPA's 2003 and 2004 National Air Quality Conferences.
STI scientists developed a statistical technique to predict next-day PM2.5 and PM10 concentrations for Cairo, Egypt. We developed a software program that imports observed and forecasted meteorological data and computes forecasts. The program also simulates visibility in Cairo using observed or predicted PM2.5 and PM10 concentrations. The program was created in Visual Basic and stores observations and forecasts in a Microsoft Access database. STI staff worked with local air quality and meteorological experts to develop and implement this forecasting system and trained them on its operation.
STI developed a guidance document to assist air quality agencies in developing and operating ozone forecasting programs and then updated the document to include PM2.5 forecasting programs. We surveyed experts in the air quality forecasting community, conducted interviews with public outreach coordinators to determine their requirements, and used information gained from a PM forecasting pilot study performed for the U.S. EPA. This information was used to develop a comprehensive ozone and PM2.5 forecasting guidance document (http://www.epa.gov/airnow//aq_forecasting_guidance-1016.pdf).
STI developed ozone forecasting methods that allow California's Northern Sierra Air Quality Management District (AQMD) staff to predict whether air quality in their region is coupled or de-coupled with air quality in nearby Sacramento and San Francisco. Five years of air quality and meteorological data were analyzed using a variety of regression techniques to identify statistically robust forecasting methods. The forecasting system was implemented using forecasted weather variables (surface winds, surface and aloft temperatures, etc.) to predict maximum next-day ozone concentrations 24 hours in advance. The forecasting system helps Northern Sierra AQMD staff anticipate when their air basin will be impacted by ozone transported from upwind regions.