Mapping daily snow/ice shortwave broadband albedo from Moderate Resolution Imaging Spectroradiometer (MODIS): The improved direct retrieval algorithm and validation with Greenland in situ measurement
Mapping daily snow/ice shortwave broadband albedo from Moderate Resolution Imaging Spectroradiometer (MODIS): The improved direct retrieval algorithm and validation with Greenland in situ measurement
Loading...
Files
Publication or External Link
Date
2005-05-26
Authors
Liang, Shunlin
Stroeve, Julienne
Box, Jason E.
Advisor
Citation
Liang, S., J. Stroeve, and J. E. Box (2005), Mapping daily snow/ice shortwave broadband albedo from Moderate Resolution Imaging Spectroradiometer (MODIS): The improved direct retrieval algorithm and validation with Greenland in situ measurement, J. Geophys. Res., 110, D10109, doi:10.1029/2004JD005493.
DRUM DOI
Abstract
Snow/ice albedo is a critical variable in surface energy balance calculations. The
Moderate Resolution Imaging Spectroradiometer (MODIS) data have been used routinely
to provide global land surface albedo. The MODIS algorithm includes atmospheric
correction, surface reflectance angular modeling, and narrowband to broadband albedo
conversion. In an earlier study, a "direct retrieval" methodology was proposed to calculate
instantaneous albedo over snow and ice-covered surfaces directly from top-of-atmosphere
(TOA) MODIS reflectance data. The method consists of extensive radiative transfer
simulations for a variety of atmospheric and surface snow conditions and links the TOA
reflectance with surface broadband albedo through regression analysis. Therefore the
direct retrieval algorithm implicitly incorporates in a single step all three procedures used
in the standard MODIS surface albedo algorithm. This study presents improvements to the
retrieval algorithm including validation with in situ measurements distributed over the
Greenland ice sheet. Comparison with surface observations demonstrates that the direct
retrieval algorithm can produce very accurate daily snow/ice albedo with mean bias of less
than 0.02 and residual standard error of 0.04.