The Development of METAL-WRF Regional Model for the Description of Dust Mineralogy in the Atmosphere

The mineralogical composition of airborne dust particles is an important but often neglected parameter for several physiochemical processes, such as atmospheric radiative transfer and ocean biochemistry. We present the development of the METAL-WRF module for the simulation of the composition of desert dust minerals in atmospheric aerosols. The new development is based on the GOCART-AFWA dust module of WRF-Chem. A new wet deposition scheme has been implemented in the dust module alongside the existing dry deposition scheme. The new model includes separate prognostic fields for nine (9) minerals: illite, kaolinite, smectite, calcite, quartz, feldspar, hematite, gypsum, and phosphorus, derived from the GMINER30 database and also iron derived from the FERRUM30 database. Two regional model sensitivity studies are presented for dust events that occurred in August and December 2017, which include a comparison of the model versus elemental dust composition measurements performed in the North Atlantic (at Izaña Observatory, Tenerife Island) and in the eastern Mediterranean (at Agia Marina Xyliatos station, Cyprus Island). The results indicate the important role of dust minerals, as dominant aerosols, for the greater region of North Africa, South Europe, the North Atlantic, and the Middle East, including the dry and wet depositions away from desert sources. Overall, METAL-WRF was found to be capable of reproducing the relative abundances of the different dust minerals in the atmosphere. In particular, the concentration of iron (Fe), which is an important element for ocean biochemistry and solar absorption, was modeled in good agreement with the corresponding measurements at Izaña Observatory (22% overestimation) and at Agia Marina Xyliatos site (4% overestimation). Further model developments, including the implementation of newer surface mineralogical datasets, e.g., from the NASA-EMIT satellite mission, can be implemented in the model to improve its accuracy.