Abstract

During 2019 ACOM and RAL set up the model WRF-Chem to calculate chemical forecasts for the FIREX-AQ field campaign. WRF-Chem produces 49 hours of forecasts at one-hour intervals, and the flight planning team uses these forecasts every morning to plan flight routes for the NASA DC-8 and NOAA Twin ... Show moreDuring 2019 ACOM and RAL set up the model WRF-Chem to calculate chemical forecasts for the FIREX-AQ field campaign. WRF-Chem produces 49 hours of forecasts at one-hour intervals, and the flight planning team uses these forecasts every morning to plan flight routes for the NASA DC-8 and NOAA Twin Otter aircraft. There are 4,998 plots of CONUS/Colorado/Front Range to make by 7:00 AM each morning, and model results do not begin arriving until after 4:00 AM. The first runs indicated that plotting with Python, matplotlib, and Cartopy would take 8+ hours to complete. The author will describe how parallel processing and pipelining were used to decrease this time to about 1 hour. A dispatcher script monitors incoming WRF-Chem output and assigns sub-tasks to derive additional variables and create US maps. The result is a practical, efficient, and reliable near-real-time forecasting system that is ready for flight planning every morning. The WRF-Chem forecasts and plotters continue to run daily, and are displayed on the ACOM web site: https://www.acom.ucar.edu/firex-aq/forecast.shtml. The WRF-Chem forecasts are part of the EU AQ-WATCH project (https://www.aq-watch.eu/), which has the objective to develop prototypes for air quality forecasting and analysis tools. Show less