Sensitivity of the WRF-Chem (V3.6.1) model to different dust emission parametrisation: assessment in the broader Mediterranean region
- 1National Observatory of Athens, Athens, Greece
- 2USDA-ARS Jornada Experimental Range, Las Cruces, NM, USA
- 3LATMOS/IPSL, UPMC Univ. Paris 06, Sorbonne Universités, UVSQ, CNRS, Paris, France
Abstract. In this study we aim to assess the WRF-Chem model capacity to reproduce dust transport over the eastern Mediterranean. For this reason, we compare the model aerosol optical depth (AOD) outputs to observations, focusing on three key regions: North Africa, the Arabian Peninsula and the eastern Mediterranean. Three sets of four simulations have been performed for the 6-month period of spring and summer 2011. Each simulation set uses a different dust emission parametrisation and for each parametrisation, the dust emissions are multiplied with various coefficients in order to tune the model performance. Our assessment approach is performed across different spatial and temporal scales using AOD observations from satellites and ground-based stations, as well as from airborne measurements of aerosol extinction coefficients over the Sahara.
Assessment over the entire domain and simulation period shows that the model presents temporal and spatial variability similar to observed AODs, regardless of the applied dust emission parametrisation. On the other hand, when focusing on specific regions, the model skill varies significantly. Tuning the model performance by applying a coefficient to dust emissions may reduce the model AOD bias over a region, but may increase it in other regions. In particular, the model was shown to realistically reproduce the major dust transport events over the eastern Mediterranean, but failed to capture the regional background AOD. Further comparison of the model simulations to airborne measurements of vertical profiles of extinction coefficients over North Africa suggests that the model realistically reproduces the total atmospheric column AOD. Finally, we discuss the model results in two sensitivity tests, where we included finer dust particles (less than 1 µm) and changed accordingly the dust bins' mass fraction.