A multi-channel approach to Greenland snowmelt estimation

Abstract

Snowmelt monitoring on the Greenland ice sheet is important as an indicator of global climate change. A previous study employed a cross-polarized gradient ratio (XPGR) method to determine melt extent using two channels of SSM/I data. In this work, principal component analysis is presented as a method of characterizing such snowmelt using all seven SSM/I channels. A matrix is given for the projection of the seven channel data onto a two dimensional subspace wherein the melt/refreeze cycle is clear. Evidence is presented for the use of all seven channels to quantify melt intensity as well as duration.