Evaluation of surface and near-surface melt characteristics on the Greenland ice sheet using MODIS and QuikSCAT data

The Greenland ice sheet has been the focus of much attention recently because of increasing melt in response to regional climate warming and can be studied using Moderate Resolution Imaging Spectroradiometer (MODIS) and Quick Scatterometer (QuikSCAT) data. To improve our ability to measure surface melt, we use remote sensing data products to study surface and near-surface melt characteristics of the Greenland ice sheet for the 2007 melt season when record melt extent and runoff occurred. MODIS daily land surface temperature (LST), MODIS daily snow albedo, and a special diurnal melt product derived from QuikSCAT (QS) scatterometer data, are all effective in measuring the evolution of melt on the ice sheet. These daily products, produced from different parts of the electromagnetic spectrum, are sensitive to different geophysical features, though QS- and MODIS-derived melt generally show excellent correspondence when surface melt is present. Values derived from the MODIS daily snow albedo product drop in response to melt and change with apparent grain size changes. For the 2007 melt season, the MODIS LST and QS products detect 766,184 km(2) +/- 8% and 862,769 km(2) +/- 3% of melt, respectively. The QS product detects about 11% greater melt extent than is detected by the MODIS LST product probably because QS is more sensitive to surface melt and can also see subsurface melt. The consistency of the response of the different products demonstrates unequivocally that physically meaningful melt/freeze boundaries are detected. We have demonstrated that when these products are used together we can improve the precision in mapping surface and near-surface melt extent on the Greenland ice sheet.