Multiyear sea ice concentration from combined microwave radiometer and scatterometer observations

Multiyear sea ice

Sea ice is an important component of the global climate system and at the same time a sensitive climate indicator. Different sea-ice types can be separated based on their age and stage of development: the two main classes are first-year ice (FYI) and multiyear ice (MYI). MYI is older, thicker ice that has survived at least one summer melt and FYI represents the seasonal ice cover. The sea ice area in the Arctic is rapidly decreasing with a rate of about –4% per decade. Especially pronounced is the decrease in thick, MYI. Large parts of the thick MYI have been replaced by increasingly thin FYI, which impacts the weather and climate through different radiation and dynamic feedbacks, and has consequences not only for the Arctic but also for mid-latitude regions. For example, MYI has more ridges, i.e. stronger surface topography, which results in a smaller surface fraction to be covered by melt ponds in summer and thereby a higher albedo and less absorption of shortwave radiation. The stronger and thicker MYI is more resistant against wind forcing, i.e. less leads can open up and less energy and gases are transferred between the ocean and atmosphere. The larger ice thickness makes it less likely to completely melt in summer, which in turn again keeps the albedo higher and avoids the ice-albedo-feedback to kick in.

Multiyear sea ice retrieval from microwave satellite observations

Microwave satellite observations have been widely used for monitoring sea ice in polar regions because they are independent of sunlight and can penetrate clouds. Total sea ice concentration can be derived reliably from microwave remote sensing data because of the distinct difference of microwave signatures between open water and sea ice. However, the discrimination of ice types (including MYI) can be difficult because different ice types may have similar typical microwave signatures, i.e., brightness temperatures at different polarizations, backscatter values or combination of both. Such ambiguous radiometric signatures, which can make FYI radiometrically look like MYI, can be caused by, e.g., thaw-freeze cycles in the snow or increased surface roughness caused by sea ice deformation due to wind and waves, especially along the ice margins.

Improved MYI concentration retrieval

The improved Arctic MYI concentration retrieval developed at the University of Bremen are based on the Environment Canada's Ice Concentration Extractor (ECICE, Shokr et al., 2008) using brightness temperatures from the microwave radiometer AMSR-E and radar backscatter from the Ku-band scatterometer QuikSCAT. They are corrected by two correction schemes: one using temperature records from atmospheric reanalysis to identify MYI anomalies and replace them with interpolated MYI concentrations (Ye et al., 2016a), and the other utilizing mainly ice drift records from satellite observations to constrain the MYI changes within a plausible contour (Ye et al., 2016b). After the two corrections are applied, the MYI concentration estimates in the Arctic Basin, particularly along the ice margins and in the peripheral seas, are much improved and more realistic (see Figure 1).

Currently, the MYI dataset at University of Bremen are available for the years 2002-2009 on a 4.45 km polar stereographic grid. From 2003 to 2009, the MYI area has decreased by about 1.5 million km2 on average (see Figure 2). The distribution of MYI for all Januaries from 2003 to 2009 is shown in Figure 3. In the earlier years the center of the Arctic Basin was completely covered by MYI, while in recent years only a small area of MYI survives along the northern coasts of Greenland and the Canadian Arctic Archipelago. In order to better understand the mechanism of MYI changes and ascertain trends of the Climate change, a long and physically consistent MYI time series is needed. The current MYI dataset will be extended to present using data from the radiometer AMSR2 and the C-band scatterometer ASCAT.

Figure 1: Multiyear ice concentration from the ECICE algorithm (top) and those after the two correction schemes applied on selected days in January 2007 (bottom). Many regions with false MYI areas, especially along the ice margins, are corrected in the bottom row.
Figure 2: Monthly averaged Arctic multiyear ice area in winter months (October–April) from 2002 to 2009.
Figure 3: Monthly average multiyear ice concentration in the Arctic for January of the years 2003–2009 demonstrating the drastic decrease in MYI area during that time period.


Shokr, M., Lambe, A., and Agnew, T. (2008). A New Algorithm (ECICE) to Estimate Ice Concentration From Remote Sensing Observations: An Application to 85-GHz Passive Microwave Data. IEEE Transactions on Geoscience and Remote Sensing, 46(12), 4104–4121, doi:10.1109/TGRS.2008.2000624.

Ye, Y., Heygster, G., and Shokr, M. (2016a). Improving multiyear ice concentration estimates with air temperatures. IEEE Transactions on Geoscience and Remote Sensing, 54(5), 2602- 2614, doi:10.1109/TGRS.2015.2503884.

Ye, Y., Shokr, M., Heygster, G., and Spreen, G. (2016b). Improving multiyear sea ice concentration estimates with sea ice drift. Remote Sensing, 8(5), 397, doi:10.3390/rs8050397.


Questions regarding this topic can be addressed to Dr. Yufang Ye via yufang(at)