Estimation of melt pond fraction over high-concentration Arctic sea ice using AMSR-E passive microwave data

Melt pond fraction (MPF) on sea ice is an important factor for ice-albedo feedback throughout the Arctic Ocean. We propose an algorithm to estimate MPF using satellite passive microwave data in this study. The brightness temperature (T-B) data obtained from the Advanced Microwave Scanning Radiometer-Earth observing system (AMSR-E) were compared to the ship-based MPF in the Beaufort Sea and Canadian Arctic Archipelago. The difference between the T-B at horizontal and vertical polarizations of 6.9 and 89.0 GHz (MP06H-89V), respectively, depends on the MPF. The correlation between MP06H-89V and ship-based MPF was higher than that between ship-based MPF and two individual channels (6.9 and 89.0 GHz of horizontal and vertical polarizations, respectively). The MPF determined with the highest resolution channel, 89.0 GHz (5 km x 5 km), provides spatial information with more detail than the 6.9 GHz channel. The algorithm estimates the relative fraction of ice covered by water (1) over areas where sea ice concentration is higher than 95%, (2) during late summer, and (3) in areas with low atmospheric humidity. The MPF estimated from AMSR-E data (AMSR-E MPF) in early summer was underestimated at lower latitudes and overestimated at higher latitudes, compared to the MPF obtained from the Moderate Resolution Image Spectrometer (MODIS MPF). The differences between AMSR-E MPF and MODIS MPF were less than 5% in most the regions and the periods. Our results suggest that the proposal algorithm serves as a basis for building time series of MPF in regions of consolidated ice pack.