On the Estimation of Melt Pond Fraction on the Arctic Sea Ice With ENVISAT WSM Images

Abstract

The accuracy of microwave radiometer ice concentration (IC) retrievals in the Arctic is degraded by melt ponds on sea ice during the melting season. For the development of IC retrieval algorithms and for the quantification of their uncertainties, data sets on the area fraction of melt ponds $(f_{mp})$ are needed. $f_{mp}$ retrieval with optical satellite data is limited by clouds. Thus, we have studied $f_{mp}$ retrieval with ENVISAT wide swath mode (WSM) synthetic aperture radar (SAR) images which have large daily coverage over the Arctic Sea ice in 2007–2012. The WSM images used here were acquired north of the Fram Strait in June–August 2009. Data on $f_{mp}$ were available from the Integrated Climate Data Center's daily Moderate Resolution Imaging Spectroradiometer (MODIS) $f_{mp}$ product in a 12.5-km grid. Relationships between SAR $\sigma^{\circ}$ and MODIS $f_{mp}$ were studied visually by comparing daily SAR mosaics and $f_{mp}$ charts and by analyzing $f_{mp}$ and $\sigma^{\circ}$ time series and spatially and temporally coincident $f_{mp}$ and $\sigma^{\circ}$ data. The correspondence between the changes of $f_{mp}$ and the $\sigma^{\circ}$ statistics is too low to suggest $f_{mp}$ estimation from the WSM images. In some cases, there was a 2–3-dB $\sigma^{\circ}$ increase during the ponding period. It is assumed that the variation of snow and sea ice characteristics diminishes $\sigma^{\circ}$ changes due to the melt ponding and drainage. Good correlation between $\sigma^{\circ}$ and $f_{mp}$ has only been observed for smooth landfast first-year ice in previous studies. A very interesting observation was the large temporal $\sigma^{\circ}$ variations during the late melting season, which are likely linked to the atmospherically forced freezing–melting events. These events may also influence radiometer IC retrievals.