Algorithm to retrieve the melt pond fraction and the spectral albedo of Arctic summer ice from satellite optical data

Abstract

A new algorithm to retrieve characteristics (albedo and melt pond fraction) of summer ice in the Arctic from optical satellite data is described. In contrast to other algorithms this algorithm does not use a priori values of the spectral albedo of the sea-ice constituents (such as melt ponds, white ice etc.). Instead, it is based on an analytical solution for the reflection from sea ice surface. The algorithm includes the correction of the sought-for ice and ponds characteristics with the iterative procedure based on the Newton–Raphson method. Also, it accounts for the bi-directional reflection from the ice/snow surface, which is particularly important for Arctic regions where the sun is low. The algorithm includes an original procedure for the atmospheric correction, as well. This algorithm is implemented as computer code called Melt Pond Detector (MPD). The input to the current version of the MPD algorithm is the MERIS Level 1B data, including the radiance coefficients at ten wavelengths and the solar and observation angles (zenith and azimuth). Also, specific parameters describing surface and atmospheric state can be set in a configuration input file. The software output is the map of the melt ponds area fraction and the spectral albedo of sea-ice in HDF5 format. The numerical verification shows that the MPD algorithm provides more accurate results for the light ponds than for the dark ones. The spectral albedo is retrieved with high accuracy for any type of ice and ponds.