Author(s)

Roland Stull, Josh Hacker & Henryk Modzelewski

Abstract

Ensembles of many runs of a numerical forecast model provide better meteorological information than a categorical forecast from a single run. The ensemble average gives more accurate plume track and wind speed. The spread of ensemble members defines likely bounds of possible plume tracks, while clustering of members gives information on track probabilities. Tracer transport examples from numerical mesoscale model ensemble runs are presented. 1 Introduction Forecast errors in wind direction, wind speed, static stability, and precipitation cause first-order errors in calculation of pollution concentration at receptors. Yet this meteorological information is traditionally obtained from a single categorical w

Keywords