Comparison of space-based lidar observations (CALIOP) with regional model results over Europe

A regional transport model (COSMO-MUSCAT: Consortium for Small-scale Modeling - MultiScale Atmospheric Transport Model) is used for simulations of aerosol optical properties within Europe. Based on the simulated concentrations of specific aerosol types together with mass extinction efficiencies, vertical profiles of the extinction coefficients (EXTs) are calculated from the model results. Vertical profiles of backscatter coefficients (BSCs) are computed using a lidar ratio. In the model version used here marine and dust aerosol types are not computed. Model simulations are performed for two short time periods: 19-26 July 2006 and 16-26 February 2007. The summer period is characterized by low wind speeds and a persistent meteorological situation. This caused the accumulation of particles within the atmosphere. For this time period an average aerosol optical depth (AOD) of 0.19 is simulated, whereas due to particle removal by several precipitation events during the winter period the average AOD is lower (0.14). In this investigation space-based observations of vertical profiles of BSCs are used for model evaluation. CALIOP (Cloud-Aerosol Lidar with Orthogonal Polarization) performs global aerosol profile measurements since April 2006 continuously. In addition to the BSC profiles information about the occurrence of particular aerosol types (smoke, polluted dust, clean continental, polluted continental, dust and clean marine) is available. Comparisons between observed and simulated BSC profiles for these two time periods show a relative model bias of -23%- -10% during summer and of -77%- -53% during winter between surface and 2 km altitude. Especially, during the winter period the bias is quite large caused in part by the tendency of the model to remove particles too efficiently due to wet deposition. Further, during 19-26 July 2006 Saharan dust was transported to Europe, which is not described by the model, and marine aerosol is only identified by CALIOP over sea surface. Eliminating BSC profiles where marine and dust aerosol types were detected by CALIOP improves the agreement between observed and simulated BSC profiles. A relative bias of -20% - -2% during summer and of -72% - -46% during winter is determined. The relatively large discrepancy during winter time can also be caused by the misinterpretation of the occurrence of marine aerosol. Marine aerosol was dominant during that time but CALIOP detects marine aerosol only over sea surface, whereas the transport of marine aerosol into continental regions is not considered. Additionally, differences regarding day- and night-time are found. In general comparisons between CALIOP and COSMO-MUSCAT lead to better results for night-time compared to day-time observations. This difference has also been found in comparisons of CALIOP BSC profiles with ground-based lidar profiles. The results point to the need to implement dust and marine aerosol type for characterizing aerosol distribution within the regional model for the European domain. Additionally, by using CALIOP data it is possible to analyze model ability to simulate the occurrence, the transport as well as the optical properties of specific aerosol types.