Data Merging Of MISR And MODIS Aerosol Optical Thickness And Application In Estimating Ground-level Fine Particulate Matter Concentrations

Atmospheric aerosols play a significant role in the global climate and radiationbalance. The aerosol optical thickness is a measure of light extinction by aerosol inthe atmospheric column above the earth’s surface. It is also used to characterizeatmospheric turbidity, aerosol content and air quality. Due to limited ground-basedaerosol stations and high temporal variability of aerosol, it is difficult to achievecontinuous and large coverage data, whereas satellite remote sensing technologyprovides a new approach to solve these problems.The launch of the MISR (Multi-angle Imaging Spectroradiometer) and theMODIS (Moderate-resolution Imaging Spectrometer) instruments onboard NASA’sfirst earth observation satellite—Terra in1999has provided global measurements ofaerosol optical thickness (AOT) products. MODIS has a high spatial resolution and awide swath; while MISR’s multi-angular observation capability can retrieve moreeffectively over bright surface areas like snow-covered region and desert. Merging ofthe two sensor’s data can overcome the disadvantages of the single sensor.In this paper, methods of weighted averaging, statistically optimal estimation andwavelet decomposition and reconstruction are applied to merge aerosol opticalthickness at550nm from MISR and MODIS, the merged data are evaluated andanalyzed to study the capability in monitoring the ground-level fine particulate matterconcentrations (PM2.5).In the study region of10°N-40°N、100°E-130°E, both MISR and MODIS aerosoloptical thickness products at550nm are matched and correlated with AERONET dataduring2011to2012, respectively. The corrected MISR and MODIS products will beused for data merging.Methods of weighted averaging, statistically optimal estimation, pixel-based and local variance-based wavelet analysis were applied to merge aerosol optical thicknessof MISR and MODIS in2011-2012. For the daily merged results, the accuracies ofobjective analysis methods (weighted averaging and statistically optimal estimation)are better than that of wavelet analysis method, while for larger time scales and spatialcoverage, wavelet analysis merging reflects richer detailed information. Comparedwith single sensor, the data coverage of the merged products increase by6.7%~13.2%(than MODIS) or32.8%~39.4%(than MISR).By comparing the merged aerosol optical thickness with fine particulate matter(PM2.5) data provided by ground air quality monitoring stations, the time seriesvariations agree well between the two parameters. Linear regressions between PM2.5index and merged aerosol optical thickness in five cities are analyzed, and thecorrelation coefficients vary among0.4722~0.7358, with an average value at0.6186,showing a significant correlation between the two parameters. This indicates thataerosol optical thickness derived from satellite sensors can provide supplementalinformation for determining ground-level fine particulate matter concentrations. Thefindings illustrate the strong potential of satellite remote sensing in regional air qualitymonitoring as an extension to ground networks.