Validation of Retrieval of Annual Air Quality Means on a Limited Number of LES Model Simulations

Abstract

ZÁKLADNÍ ÚDAJE: PMC26 Book of Abstracts. Hannover: Leibniz University Hannover, German Weather Service, 2026. s. 11-11. [PMC 26: PALM Model Conference. 23.02.2026-26.02.2026, Hannover]. ABSTRAKT: Annual means of selected pollutants are important legislative air quality statistics with controlled limits. Current computational capabilities enable the application of the PALM model to assess areas of several square kilometres at meter-scale resolution. However, the computational requirements remain too high to simulate a full year, which is required to calculate the annual mean. The newly developed method for retrieving annual air quality statistics from a limited number of simulations was previously presented. The method identifies “typical” days using k-medoids clustering. Validation using monitoring stations and a pilot experiment for retrieving annual-statistics fields from LES models (presented at the last PALM Model Conference) was promising, so the validation process has moved to the next phase – validation using model outputs. For this purpose, fields of annual means were reconstructed over a large domain that includes the city of Prague. For the retrieval, a set of “typical day” simulations from PALM in 10 m resolution was used. This relatively coarse 10 m resolution was motivated by the need to model most of the Prague area and to enable the validation of as many air quality monitoring stations as possible.