Publications

Kumar, A; Mehta, M (2023). Investigating the applicability of a simple iterative approach for aerosol optical depth (AOD) retrieval over diverse land surface types from Landsat 8 and Sentinel 2 using visible and near-infrared (VNIR) spectral bands. ATMOSPHERIC ENVIRONMENT, 314, 120082.

Abstract
Previous attempts at aerosol optical depth (AOD) retrieval from top-of-atmosphere satellite signals using a simplified approach have been limited in their assessment over selective surface types, calibration sites and given the low surface reflectance components at lower wavelengths, have mostly relied, fully or partially, on near-UV or blue spectral bands for AOD retrieval. The present study attempts to investigate the applicability of a simple iterative approach for retrieval of AOD over diverse land surface types using any of the visible and near-infrared (VNIR) spectral bands. The results have been validated for 18 AERONET sites from 7 countries in South Asia and we found a strong correspondence between retrieved AOD and the AERONET AOD with good correlation (r > 0.80), both at 550 nm as well as at individual spectral bands. For image-based comparison, the results have been compared with those from MODIS products. While the retrievals show strong spatial similarity with MODIS, in absolute terms, they are mostly underestimated. Nevertheless, the coincident retrievals from both sensors do appear to be very similar, with intercomparison also showing good correlation. Overall, the iterative approach for AOD retrieval performs reasonably well for different surface types and can be easily adopted for most sensors, especially those which do not have deep blue or blue spectral bands.

DOI:
10.1016/j.atmosenv.2023.120082

ISSN:
1873-2844