Atmospheric correction of satellite data using multi-wavelength lidar data with MODTRAN3 code Original Research Article

Standard atmospheric models are employed in radiative transfer codes such as MODTRAN3. The actual atmospheric conditions, especially those due to aerosol particles, change both spatially and temporally. For the atmospheric correction of satellite data it is highly desirable to use the data measured simultaneously with the satellite overflight. Lidars are quite useful to obtain the aerosol information. The Atmospheric Data Correction Lidar, a multi-wavelength lidar system at CEReS, Chiba University, has been applied for that purpose. Lidar signals are collected on June 27, 1997 and December 15, 1997, and analyzed to obtain the aerosol extinction profile. The extinction at 550 nm is interpolated from lidarderived data and substituted into MODTRAN3 as user-defined 26 layers. To compare this with the result using standard atmospheric model (urban spring/summer and urban fall/winter), the meteorological ground visibility for the latter is adjusted to give the optical thickness the same as the lidar data. The atmospheric correction to NOAA AVHRR data (channel 1 and 2) is performed with two step correction, the first step removing the atmospheric scattering contributions. The second correction, on the other hand, deals with the adjacency effect. This algorithm gives rise to appropriate albedo values in the range of 0.1 to 0.3 for land surfaces and 0.01 to 0.03 for sea surfaces. Although the effect of atmospheric correction is significant, the difference between the results based on lidar data and the standard model is found to be relatively small.