Sensitivity study and comparative evaluation of WRF-Chem over Iran: Available and embedded dust emission schemes

Abstract

In this study, a severe dust storm has been simulated using the WRF-Chem model to investigate the accuracy of different dust emission schemes in estimating dust concentration in Tehran, the capital of Iran. First, by comparing the results of the GOCART, AFWA, UOC-2001, UOC-2004, and UOC-2011 dust emission schemes in WRF-Chem model, it is observed that the GOCART and AFWA schemes have the best performance compared to the in-situ measurements. The GOCART scheme with a better prediction of the trend of the dust concentration profile, and the AFWA scheme with a better prediction of the dust concentration peak value, are selected as the preferred schemes for this region. In the following, by using a two-way nested grid and embedding a new computational code for calculating the dust emission into the numerical model, an attempt has been made to increase the accuracy of the numerical model, especially in Tehran. Compared to the base model, the new dust emission scheme has decreased the overestimation of peak dust concentrations from observations by 24.9% and has enhanced the correlation coefficient by 0.12. Finally, by focusing on identifying dust sources, it is observed that external sources, namely Iraq and eastern Syria dry lands, have a significant influence on dust emissions. It is shown that the amount of mineral dust originating from these areas on February 20 and 21 is 1.5–1.9 Mg/km2. Also, the total dust particles lifted from the entire computing domain and Iran are 95.1 and 1.3 Tg, respectively.