Up to 20% of collared polar bears located on ice that officially does not exist, says the PBSG

Here are two more priceless quotes from the minutes of the last meeting of the IUCN Polar Bear Specialist Group (PBSG) – on issues with sea ice percentages used to define “ice-free” and the problem of bears with collars showing up on sea ice that, according to ice data, does not exist. These quotes are in addition to the ones I posted earlier this week (here and here).

See the original document for the context here.

Bold and red highlights in the quotes below are mine:

“Defining sea-ice thresholds are important because there can be a big difference in the measured length of “ice-free” periods depending on whether 50% coverage or 30% coverage is the threshold used.” Pg. 22 [I’ve written about this before (here and here), and yes, it makes a big difference]

“There was further discussion on sea ice thresholds and that telemetry and sea ice do not always line up. For example, recent analysis shows that up to 20% of locations of collared females are in areas where the ice data indicates there is no ice.” Pg. 23 [Which “recent” analysis is this? And why is this not something the public have been told?]

Summary and comments:
1) Even though a recent study demonstrated that using 30% coverage as the cut-off for determining ice breakup in Western Hudson Bay best describes how polar bears respond to ice melt in spring (and 10% coverage in fall during freeze-up), most polar bear studies (e.g. Derocher et al. 2013; Robbins et al. 2012) are still using the 50% concentration metric – for both spring and fall – to determine the length of ice-free periods.

Why? Undoubtedly, because it’s easier: 50% concentration is what the sea ice experts use, so the polar bear biologists don’t have to calculate their own data. But also, perhaps, because it better fits their narrative that polar bears are rapidly losing their sea ice hunting platform?

2) Sea ice data, as determined by satellite images, clearly have some issues if 20% of polar bears being tracked with satellite radio collars are indicating bears on the ice when sea ice charts show no ice available. In summer, when the image analysis software can’t tell the difference between melt-water on top of the ice and open-water, it simply declares the area to be open-water (Scott and Marshall 2010:156 – see quote below).

The bottom line is this: it is becoming increasingly obvious that the sea ice data now used by polar bear biologists (generated by sea ice experts) are not producing data on sea ice conditions relevant for polar bears.

And yet, it is sea ice coverage projected into the future (by those same sea ice experts), that is being used to predict that polar bears will be threatened with extinction within three generations.

Add the problems described above to the ones discussed elsewhere in the June 2014 meeting minutes (see quotes and discussion here) and it’s clear there are some serious problems. Surely, you might be thinking, these polar bear specialists are beginning to understand that any prediction of future polar bear population decline that’s based on sea ice data will not be accurate enough to be credible?

Apparently not, as the plans outlined by the PBSG for the upcoming IUCN status assessment (2015) suggest. Go figure.

Scott and Marshall (2010:156) on systematic underestimation of microwave data:

“The accuracy of the concentration data is stated to be at its worst, at ±15% of the actual sea-ice concentration (Cavalieri et al., 1992), during summer in the Arctic, which unfortunately corresponds to the time of breakup we are interested in. The low accuracy is largely due to the effects of surface melt ponds on the sea ice, which can lead to underestimation of the sea-ice concentration. In addition, at the ice margins and areas of ice breakup, the many different concentrations that can exist within one pixel will be smoothed to an average figure. Several studies have shown that because of the surface melt, in particular, the passive microwave-derived data tend to systematically underestimate ice concentrations (e.g. Agnew and Howell, 2003; Shokr and Markus, 2006). When compared to sea-ice charts, the passive microwave sea-ice concentrations derived from the NASA Team algorithm were found to underestimate concentration during summer melt by 20.4% to 33.5%. The improved NASA Team 2 algorithm has been shown to underestimate concentration by 18.35% on average, with a standard deviation of 16.8% (Shokr and Markus, 2006). These studies suggest that although a random error of around 15% is still reasonable, there will be a systematic underestimation of ice concentration in the passive microwave concentration data set.” [my bold]

Scott and Marshall (2010:157) also suggested that Canadian Ice Service (CIS) daily and weekly charts may be “more accurate than passive microwave data for estimates of ice concentration, particularly in the presence of surface melt,” although more satellite data are now more likely to be incorporated by CIS into their analyses than previously (i.e., before 1987).

References
Derocher, A.E., Aars, J., Amstrup, S.C., Cutting, A., Lunn, N.J., Molnár, P.K., Obbard, M.E., Stirling, I., Thiemann, G.W., Vongraven, D., Wiig, Ø., and York, G. 2013. Rapid ecosystem change and polar bear conservation. Conservation Letters 6:368-375. http://onlinelibrary.wiley.com/doi/10.1111/conl.12009/abstract

Robbins, C.T., Lopez-Alfaro, C., Rode, K.D., Tøien, Ø., and Nelson, O.L. 2012. Hibernation and seasonal fasting in bears: the energetic costs and consequences for polar bears. Journal of Mammalogy 93(6):1493-1503. http://www.asmjournals.org/doi/abs/10.1644/11-MAMM-A-406.1

Scott, J.B.T. and Marshall, G.J. 2010. A step-change in the date of sea-ice breakup in western Hudson Bay. Arctic 63:155-164. Open access http://arctic.synergiesprairies.ca/arctic/index.php/arctic/issue/view/55