The objective of this study is the integration of satellite and in-situ measurements of particulate matter (PM10) to provide PM10 maps in Switzerland and South Tyrol (Italy) on an operational daily basis. Satellite retrieval of PM has been widely investigated in the past years, showing moderate potential (uncertainty of ∼30%) but also a number of severe limitations (e.g., due to cloud and snow cover or unknown aerosol extinction profiles). Its actual effectiveness can only be tested by a comparison with the mapping capability of ground-based measurements from existing air-quality networks. Moreover, the integration of both observational systems (assimilation) can improve PM mapping. Herein, we apply a linear model including aerosol optical depth (AOD) from the Spinning Enhanced Visible and InfraRed Imager (SEVIRI) and meteorological boundary layer height (BLH) to estimate spatially homogeneous maps of PM10 over the study region in 2008-2009. AOD from MODIS is used to compare the results with those of similar studies. The validation of the satellite maps reveals higher accuracy in flat areas (r ∼ 0.6, RMSE ∼ 10 μg m-3) than in alpine valleys and elevated sites. In contrast, the inverse distance interpolation of in-situ measurements is able to produce more accurate (r > 0.8, RMSE < 6 μg m-3) PM10 maps. An assimilation schema was developed considering the interpolation of ground measurements as a background field, updating it with satellite observations wherever they are available. The accuracy of the assimilated maps is assessed and compared to the background fields. It is found that satellite data is of limited benefit in the considered region due to the good spatial coverage of the ground networks and the difficulties inherent to the satellite PM retrieval over rugged topography. The results of the assimilation are positive (∼1 μg m-3 improvement in RMSE) when a number of ground sites (80%) are excluded. It is concluded that satellite data are of higher interest for regions with a sparser distribution of measurement sites (e.g., distance > 100 km between sites). © 2011 Elsevier Ltd.
Emili, E., Popp, C., Wunderle, S., Zebisch, M., Petitta, M. (2011). Mapping particulate matter in alpine regions with satellite and ground-based measurements: An exploratory study for data assimilation. ATMOSPHERIC ENVIRONMENT, 45(26), 4344-4353 [10.1016/j.atmosenv.2011.05.051].
Mapping particulate matter in alpine regions with satellite and ground-based measurements: An exploratory study for data assimilation
Petitta M.
2011-01-01
Abstract
The objective of this study is the integration of satellite and in-situ measurements of particulate matter (PM10) to provide PM10 maps in Switzerland and South Tyrol (Italy) on an operational daily basis. Satellite retrieval of PM has been widely investigated in the past years, showing moderate potential (uncertainty of ∼30%) but also a number of severe limitations (e.g., due to cloud and snow cover or unknown aerosol extinction profiles). Its actual effectiveness can only be tested by a comparison with the mapping capability of ground-based measurements from existing air-quality networks. Moreover, the integration of both observational systems (assimilation) can improve PM mapping. Herein, we apply a linear model including aerosol optical depth (AOD) from the Spinning Enhanced Visible and InfraRed Imager (SEVIRI) and meteorological boundary layer height (BLH) to estimate spatially homogeneous maps of PM10 over the study region in 2008-2009. AOD from MODIS is used to compare the results with those of similar studies. The validation of the satellite maps reveals higher accuracy in flat areas (r ∼ 0.6, RMSE ∼ 10 μg m-3) than in alpine valleys and elevated sites. In contrast, the inverse distance interpolation of in-situ measurements is able to produce more accurate (r > 0.8, RMSE < 6 μg m-3) PM10 maps. An assimilation schema was developed considering the interpolation of ground measurements as a background field, updating it with satellite observations wherever they are available. The accuracy of the assimilated maps is assessed and compared to the background fields. It is found that satellite data is of limited benefit in the considered region due to the good spatial coverage of the ground networks and the difficulties inherent to the satellite PM retrieval over rugged topography. The results of the assimilation are positive (∼1 μg m-3 improvement in RMSE) when a number of ground sites (80%) are excluded. It is concluded that satellite data are of higher interest for regions with a sparser distribution of measurement sites (e.g., distance > 100 km between sites). © 2011 Elsevier Ltd.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
