Evaluation of the HIRLAM and HARMONIE numerical weather prediction models during an air pollution episode over greater İstanbul Area

We have evaluated the performance of operational versions of the numerical weather prediction (NWP) model HIRLAM and HARMONIE in severe air pollution atmospheric conditions in İstanbul in November 18-20 of 2009. During the episode, a high pressure system over northwest Turkey, including İstanbul, lead to the formation of an exceptional ground-based temperature inversion. The forecasts generated by HIRLAM and HARMONIE models were compared to observations at meteorological stations in the area. Both HIRLAM and HARMONIE 24h's forecasts are successful to predict possible episode conditions due to meteorological circumstances, and thus both models provide reliable weather input to an air pollution warning system. Results of this study are important especially for those involved in NWP, in air pollution modeling and decision-making where high resolution NWP model outputs are becoming a valuable product.