WRF-chem sensitivity to vertical resolution during a saharan dust event. (August 2016)

• Record Type: Journal Article
• Authors: Teixeira, J.C.; Carvalho, A.C.; Tuccella, Paolo; Curci, Gabriele; Rocha, A.
• Is Part Of: Physics and chemistry of the earth. Volume 94(2016)
• Page Start: 188
• Page End: 195
• Publication Date: 2016-08
• Subjects: Dust event -- Model sensitivity -- Aerosol -- WRF-Chem -- Vertical resolution; Geophysics -- Periodicals; Geochemistry -- Periodicals; Earth sciences -- Periodicals; Geodesy -- Periodicals; Astrophysics -- Periodicals; 550
• DOI: 10.1016/j.pce.2015.04.002
• Languages: English
• ISSNs: 1474-7065

Abstract

Abstract: The Saharan dust event that occurred between the 22nd and 30th of June 2012 influenced the atmospheric radiative properties over North Africa, the Iberian Peninsula, the Western Mediterranean basin, extending its effects to France and Southern England. This event is well documented in satellite imagery, as well as on the air quality stations over the Iberian Peninsula and the AERONET NASA network. In order to assess the effect of the model vertical resolution on the extinction coefficient fields, as a proxy to the particulate matter concentrations in the atmosphere, the WRF-Chem model was applied during this period over a mother domain with a resolution of 18 km, covering Europe and North Africa. To this end five model setups differing in the number of vertical levels were tested. Model skills were evaluated by comparing the model results with CALIPSO and EARLINET LIDAR data. Results show that the model is able to simulate the higher level aerosol transport but it is susceptible to the vertical resolution used. This is due to the thickness of the transport layers which is, eventually, thinner than the vertical resolution of the model. When comparing model results to the observed vertical profiles, it becomes evident that the broad features of the extinction coefficient profile are generally reproduced in all model configurations, but finer details are captured only by the higher resolution simulations. Highlights: We have studied the model sensitivity to the choice of the number of vertical levels. All simulations where able to broadly reproduced the temporal and vertical patterns. The increase of model vertical resolution better depicts the small scale patterns.