Retrieval of Sea Ice Surface Melt Using OLCI Data Onboard Sentinel-3

Abstract:

During the Arctic summer, the appearance of melt ponds on top of sea ice causes a change of sea ice albedo which in turn affects the energy balance of the entire region. Melt ponds facilitate top, bottom and lateral sea ice melt, affect the sea ice transmittance and underice ecosystems. It has been suggested that the melt pond fraction (MPF) during the melt onset defines the magnitude of the sea ice minimum in September. While in situ measurements of melt ponds are long available, a continued remote sensing MPF dataset is yet to be delivered. Such a dataset would help assess the role of melt ponds in the recent sea ice trend and aid climate modeling as currently no published climate model includes melt ponds on top of sea ice.

In the current work we present such a dataset obtained from OLCI data onboard Sentinel-3. This MPF dataset has been produced with the Melt Pond Detector (MPD) algorithm which has been applied to the historic MERIS data earlier. The new OLCI MPF dataset (2017 till present) can be directly compared to the MERIS MPF dataset (2002-2011) due to the similarity of both sensors. Daily and weekly MPF averages of both MPF datasets are presented. It can be seen that e.g. during the melt onset in 2018 and 2019 the maximum MPF has been observed in the Laptev Sea/East Siberian Sea, which is a rather unusual development compared to the earlier years of MERIS dataset where the melt onset mostly started in the Beaufort Sea.

Another application of the MPD algorithm is the MPF retrieval on top of the Antarctic sea ice. Although quite sparse in comparison to the Arctic, some observations of surface sea ice melt in the Antarctica are still reported (e.g. January 2017 and 2018 near Syowa Station, January 2016 and 2018 near McMurdo station) and can be for the first time detected using a remote sensing retrieval. Case studies of melt ponds on top of the Antarctic sea ice retrieved from OLCI data are presented in this study as well.