ASSESSMENT OF UNCERTAINTIES IN THE COMPUTATION OF ATMOSPHERIC CORRECTION PARAMETERS FOR LANDSAT 5 TM AND LANDSAT 7 ETM+ THERMAL BAND

Abstract

This research examines the uncertainties present when computing atmospheric correction parameters (upwelling (L_i ) and downwelling (L_i ) radiances, and transmittance ( )) for the 9 flaring sites in Rivers State, Nigeria, and to estimate the total uncertainty introduced into the land surface temperature (LST) when they are applied. Eight (8) Landsat 5 Thematic Mapper (TM) and Twenty-Four (24) Landsat 7 Enhanced Thematic Mapper Plus (ETM ) from 18 August 2000 to 08 March 2013, with less than 3 percent (%) cloud contamination, were considered to evaluate a trend. Option B of the Atmospheric Correction Parameter (ATMCORR) Calculator was adopted to obtain L_i , L_i , and  for the thermal band (band 6) of the Landsat scenes analysed. The L_i , L_i , and  obtained were applied to the calibrated at-sensor radiance band 6 (high gain) data to compute the surface-leaving radiance L_i(T_i) with the emissivity ( ) of each station estimated by using standard values for determined land surface cover. The Planck equation was inverted using the Landsat calibration constants to derive the LST. To determine the uncertainties introduced into the retrieved LST, the difference between the computed mean surface temperature (ST) measured at each flaring site during ground truthing and the combined average LST retrieved from the Landsat datasets is applied. The results show that the larger the  of water body at the site, the higher the uncertainty in LST, and vice versa. Bonny LNG has the highest error of 24.712 K, Eleme II has the lowest value of 0.752 K. Based on the results obtained, it can be concluded that the ATMCORR Calculator can provide an automated method to derive L_i , L_i , and  for Landsat 5 and Landsat 7 thermal band for generating LST in the Niger Delta, Nigeria.