Using Remote Sensing for Arctic and Antarctic Sea Ice Change Detection

Using 25km daily remote sensing grid cell data from a satellite mounted microwave sensor, National Snow and Ice Data Centre (NSIDC) scientists have shown that the minimum extent of Arctic summer sea ice is almost half that detected in the late 1970s.

As with any remote sensing data, the NSIDC tempered their interpretation with both an understanding of the sensor’s limitations, and the nature of the Sea Ice itself. In this case, they understood that error could be introduced by motion, weather and surface effects, so a cell was only flagged as having sea ice if it was detected there on multiple days.

The scientists then processed their grid cell maps (derived from the remote sensing data) to allow for moving 5 year average measurements. The results showed a 1% increase in ice coverage at the end of the summer melt since satellite image analysis began in 1979. For the same period, ice coverage at the end of the Arctic summer melt had decreased from just over 50% of the Arctic Ocean to just over 32%. In 2012 the coverage was 24%.

We should not raise our hopes about this increase in Antarctic ice coverage because Arctic and Antarctic geographies are so different. The Arctic is a confined ocean basin surrounded by land while the Antarctic is surrounded by sea meaning that the sea ice there tends to accumulate year-by-year because it’s not so affected by warm currents. In contrast, the Antarctic is unconfined so sea ice can drift into warmer waters and melts. Antarctic ice tends to be 1-2m thick while most Arctic sea ice is 2-3m thick, and up to 5m thick.

There are three major concerns that arise from these results:

  1. Sea ice reflects heat: The fact that Arctic ice survives winter is extremely important because being white, it reflects heat. When it melts, the water that’s exposed absorbs heat and acts to melt even more ice. Antarctic ice does not hang around to reflect summer sunlight, and in recent geological times that has always been the case.
  2. The baseline is recent: The baseline for this comparison is 1979 because there were no quantitative measurements prior to that. So it’s possible that the extent of the melting might be greater than these results indicate.
  3. The satellite cannot measure ice thickness: The remote sensing satellite sensor only measures surface reflection, and not ice thickness. However, submarine mounted upwards looking sonar data suggest that between the 70s and 90s, Arctic sea ice average thickness decreased by more than a metre.

Hope on the horizon?

My fingers are crossed that there will eventually be parallels between the role of Chlorofluorocarbon (CFC) in ozone depletion, and CO2 in Climate Change. In the mid 1970s it was noted that CFCs breakdown ozone, and in the mid 1980s the ozone hole over Antarctica was discovered. However, it was not until a satellite image map of the ozone hole was produced that the non-science community grasped the full magnitude of the problem. Action soon followed. Its my hope that as the evidence for global warming (including GIS and remote sending maps) becomes more convincing, public opinion and regulation will follow as swiftly as it did for the CFC issue. Maybe then the path to a low CO2 economy will be, as was the case with CFCs, swift, but hopefully not too late.


Farman JC, Gardiner BG, and Shanklin JD, 1985, Large losses of total ozone in Antarctica reveal ClOx/NOx interaction. Nature, no. 315.

GILLIS J, 2012, Running the Numbers on Antarctic Sea Ice, New York Times,

Molina M and Rowland F, 1974, Stratospheric Sink for Chlorofluoromethanes: Chlorine-Atom Catalyzed Destruction of Ozone.” Nature, no. 249.

NSIDC, 2004, “Observed Changes in Arctic Sea Ice Cover and Projections for the Future” Testimony to the United States Senate Committee on Commerce, Science and transportation

NSIDC, 2012, Animation of the Antarctic melt season

NSIDC, 2012, Animation of the Arctic melt season

NSIDC, 2012, Sea Ice datasets download:

NSIDC, 2012, Sea Ice Index Processing Steps:

Leave A Response

* Denotes Required Field