Comparison of Misr Aerosol Optical Thickness with Aeronet Measurements in Beijing Metropolitan Area

Xin Jiang,Yang Liu,Bin Yu,Ming Jiang

DOI: https://doi.org/10.1016/j.rse.2006.06.022

IF: 13.5

2007-01-01

Remote Sensing of Environment

Abstract:Aerosol optical thickness (AOT) data retrieved by the Multi-angle Imaging SpectroRadiometer (MISR) from 2002 to 2004 were compared with AOT measurements from an Aerosol Robotic Network (AERONET) site located in Beijing urban area. AERONET AOT data were first averaged within a two-hour time window and MISR data were collected using AOTs of the central regions covering the AERONET site and the average AOTs of the surrounding 3 x 3 regions. MISR and AERONET AOTs are highly correlated, with an overall linear correlation coefficient of 0.93 at 558 nm wavelength. On average, MISR AOT at 558 nm is 29% lower than the AERONET AOT at 558 nm interpolated from 440 rim and 675 nm. A linear regression analysis using MISR AOT as the response yields a slope of 0.58 0.03 and an intercept of 0.07 +/- 0.02 in the green band with similar results in the other three bands, indicating that MISR may underestimate AERONETAOT. After applying a narrower averaging time window to AERONET data and controlling their temporal variabilities, the agreement between MISR and AERONET AOTs is significantly improved with the correlation coefficient of 0.97 and a slope of 0.71 0.03 in an ordinary linear least squares fit. Aweighted linear least squares, which reduces the impact of spatial averaging, yields a better result with the slope going up to 0.73 +/- 0.03. When only the central region MISR AOTs are included in the regression analysis, the best agreement is achieved with a slope of 0.91 +/- 0.03 and an intercept of 0.00 +/- 0.01, and MISR AOT is only 9% lower than AERONET AOT on average. By investigating PM10 spatial distribution of Beijing, we found substantial spatial variations of aerosol loading with higher PM10 concentrations inthe MISR central region, which introduced low bias invalidating spatial average MISRAOT. When MISR andAERONETAOTare better spatiotemporally matched, their agreement can be much better than the previously reported results. Our findings also suggest that MISR aerosol retrieval algorithm might need to be adjusted for the extremely high aerosol loadings and substantial spatial variations that it will probably encounter in heavily polluted metropolitan areas. (c) 2006 Elsevier Inc. All rights reserved.