Validation of MODIS Aerosol Products with AERONET Measurements of Different Land Cover Types in Areas over Eastern Europe and China

Abstract

The purpose of this study was to assess the performance of moderate resolution imaging spectroradiometer (MODIS) Collection 6 and 6.1 Dark Target, Deep Blue and Dark Target, and Deep Blue combined aerosol products with a spatial resolution of 10 km over Eastern Europe and China within the period 2001–2018. The data obtained for aerosol optical depth (AOD) was verified with 10 a stations located in various regions. Uncertainty and sample precision was assessed with the use of expected error (EE) over land (±(0.05 + 15% × AODAERONET) and ±(0.05 + 20% × AODAERONET)), root-mean-square error (RMSE), mean absolute error (MAE), and relative mean bias (RMB). The results showed that Collection 6.1 Dark Target retrieval algorithm surpassed Collection 6, where 58.5–75% retrievals fell within the EE as compared with Collection 6 (36.9–69.8%), with values of RMSE = 0.098–0.297 and MAE = 0.087–0.24. In Collection 6.1, Deep Blue and Dark Target and Deep Blue combined retrieval algorithms operate better; the retrieval results have been improved as compared with Collection 6: the retrieval results improved from 49.2–70.8% and 53.4–73.6% to 52.3–76.3 and 70.6–87.7% with RMSE = 0.101–0.17 and 0.101–0.258, and MAE = 0.087–0.127 and 0.078–0.146, respectively. The best results were obtained over the stations in Eastern Europe, and the worst ones over the coastal regions. The results of assessment vary wildly depending on the changes of different types of surface, surface reflectivity, and aerosol models at all areas.