Recent paper on W. Hudson Bay polar bears includes new official sea ice freeze-up data

From Polar Bear Science

Even though it’s in graph form only, we finally have an update on sea ice freeze-up dates for Western Hudson Bay for 2016-2020 (but not breakup dates).

This graphed data published by Miller et al. 2022 extends by five years that published in 2017 by Castro de la Guardia and colleagues, which contained graphed data for breakup and freeze-up dates from 1979-2015 (with exact dates for 2005-2008 only).

It confirms a statement I made last month, that between 2016 and 2021 “there has been only one ‘late’ freeze-up year (2016)–but five very early ones.” Of course, 2021 was not included in this new dataset, so that would be “four very early ones” up to 2020.

Here it is, as part of Figure 2. The freeze-up data is upper left (a):

Here is that tiny panel (a) on its own, enlarged (a screencap from the pdf), showing the trend line from1991-2020:

Extracting numbers off graphs is time-consuming and often inaccurate, but apparently polar bear specialists don’t care: that is all the public and their colleagues are being given.

In the lists below, the first number is the year, the second number is the Julian day of the year (e.g. 313), and the third is the calendar date of the corresponding Julian day from the Miller paper vs. the de al Guardia paper. For the last five years, I’ve added my own assessment made at the time, and because this is a screencap, here are the links for my estimates: for 2020 (when bears were killing seals on offshore sea ice by 31 Oct), 2019201820172016. I didn’t transcribe all of the pre-2005 dates, only a select few.

Most of entries in the period of overlap between the two datasets are either identical or vary by only 1-3 days. However, those in bold vary by almost a week or more (e.g. 2008, 2011), even though there is no explanation in the Miller paper regarding why that would be so:


As you can see, the earliest freeze-up year was 1993 (although 1991, 1986, and 1978 were almost that early) and the latest apparently in 2016 (although 2009, 1998, and 1981 were almost as late depending on which dataset you look at).

Given the potential error rate of 2-3 days either way, freeze-up was as early in 2020 as it had been in 1978, 1979, 1986, 1991, and 1993 (the earliest on record); freeze-up dates in 2017, 2018, and 2019 were the same as the average in the 1980s (de la Guardia et al. 2017).

Overall, Miller and colleagues found no temporal trend in sea ice or departure dates of polar bears from shore between 1991 and 2020, and perhaps counter-intuitively, that bears departed for the ice earlier in years when freeze-up was earlier.

This means WH sea ice coverage in the fall has not been ‘steadily declining’ over the last 30 years and polar bear have not been departing for the ice later and later in the season over that period, as many imply.

However, although the authors collected long-overdue data on body condition of females with cubs and independent juveniles, these values are reported only as a ‘vulnerability index’ which is impossible to compare with raw data collected in the 1970s and 1980s.

While these indices indicate that the body condition of females with cubs-of-the-year (but not those with yearling cubs) “declined over the last 30 years“, it is impossible to say by how much compared to detailed studies done prior to the 1990s (eg. Derocher and Stirling 1992, 1995; Ramsay and Stirling 1988) or to those used to justify classifying polar bears as ‘threatened’ on the US Endangered Species List (Regehr et al. 2007).

In other words, body weight data–so critical to the argument that the health of WH polar bears is declining due to sea ice loss–is still being withheld.


Castro de la Guardia, L., Myers, P.G., Derocher, A.E., Lunn, N.J., Terwisscha van Scheltinga, A.D. 2017. Sea ice cycle in western Hudson Bay, Canada, from a polar bear perspective. Marine Ecology Progress Series 564: 225–233.

Derocher, A.E. and Stirling, I. 1992. The population dynamics of polar bears in western Hudson Bay. pg. 1150-1159 in D. R. McCullough and R. H. Barrett, eds. Wildlife 2001: Populations. Elsevier Sci. Publ., London, U.K.

Derocher, A.E. and Stirling, I. 1995. Temporal variation in reproduction and body mass of polar bears in western Hudson Bay. Canadian Journal of Zoology 73:1657-1665.

Miller, E.N., Lunn, N.J., McGeachy, D., and Derocher, A.E. 2022. Autumn migration phenology of polar bears (Ursus maritimus) in Hudson Bay, Canada. Polar Biology 45:1023-1034.

Ramsay, M.A. and Stirling, I. 1988. Reproductive biology and ecology of female polar bears (Ursus maritimus). Journal of Zoology London 214:601-624.

Regehr, E.V., Lunn, N.J., Amstrup, S.C. & Stirling, I. 2007. Effects of earlier sea ice breakup on survival and population size of polar bears in Western Hudson Bay. Journal of Wildlife Management 71: 2673-2683. Paywalled, subscription required.

4.9 10 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
Ron Long
January 15, 2023 2:08 pm

I hope they don’t run out of ringed seals, I hate the thought of the polar bears looking for an alternative meal. Thanks to Dr. Susan for the update.

Tom Halla
January 15, 2023 2:18 pm

I am the cynical type who assumes that withheld data disagrees with the thesis of the group doing the coverup.

real bob boder
Reply to  Tom Halla
January 15, 2023 2:25 pm

Most certainly you are correct

Jeff Alberts
Reply to  Tom Halla
January 15, 2023 5:59 pm

Oh come now! That sort of thing never ever ever ever happens. Probably.

Coeur de Lion
January 15, 2023 2:32 pm

Are the Churchill bears fat or thin or is that a stupid question?

Jeff Alberts
Reply to  Coeur de Lion
January 15, 2023 6:00 pm

There you go body-shaming the bears.

Reply to  Jeff Alberts
January 16, 2023 1:39 am

Not to mention the terrible use of the C word!

Ben Vorlich
January 15, 2023 2:56 pm

Why the Julian Calendar and not Gregorian? What happens when they get another day out of step? As I understand it they differ by a day every 128 years.

Reply to  Ben Vorlich
January 15, 2023 5:27 pm

Julian dates, not the Julian calendar.
Julian dates just gives the number of days since the first of the year.

Reply to  Ben Vorlich
January 15, 2023 5:31 pm

It’s basically just the number of the day (from Jan 1 = day 1), which can be more convenient to plot.

Reply to  Ben Vorlich
January 16, 2023 1:40 am

Not a problem, the world will have burnt up well before then.

January 15, 2023 3:45 pm

Bear counts….so incredibly inaccurate that the people doing it don’t want to blab for fear of losing their jobs and annual trip of a lifetime. Could be 70% or 250% of what you counted on any given day of flying.

Last edited 22 days ago by DMacKenzie
Reply to  DMacKenzie
January 16, 2023 2:24 pm

I know of some beaches where “bare counts” come in pairs and are very accurate…..

January 15, 2023 4:13 pm

If you want to keep losing the debate keep counting polar bears

Reply to  Eben
January 15, 2023 5:33 pm

That’s why there have been decades without an official count – too inconvenient to support the claims they’re endangered.

Stuart Baeriswyl
January 15, 2023 5:53 pm

This was a nice article blog to read. Yet I don’t understand why ‘breakup’ and ‘freeze-up’ dates take five years to publish. Isn’t this something that could be measured via satellite imagery or by airplane or drone weekly (Spring time then Fall) and public every year?

Jeff Alberts
Reply to  Stuart Baeriswyl
January 15, 2023 6:11 pm

The data needs to be put in a slow-cooker, to get the desired consistency.

January 15, 2023 9:06 pm

“indices indicate that the body condition of females with cubs-of-the-year (but not those with yearling cubs) “declined over the last 30 years

A statement that comes across as opinion.

“In other words, body weight data–so critical to the argument that the health of WH polar bears is declining due to sea ice loss–is still being withheld.”

Making their claim about “body condition” more likely from opinion as they fear releasing hard data.
Given how many pictures of fat mothers with 2 or 3 fat cubs and how rare pictures of starving mother polar bears with starving cubs are, and they’ve been scouring polar bear territory searching for sick or starving polar bears.

January 16, 2023 12:51 pm

Churchill, Manitoba. Polar bears. Sea ice.

I was there in 2009. Their museum was all about climate change. They had the temperature graph for Churchill … ended at 1998.

Look at the graphs. For a start, why did they stop at 1998? To 98 showed a steep warming. To 2009, not so much. Since 2009? Hanging around. No alarm.

Maximum warmth is not increasing. Maximum cold has decreased, yes, but not ground/sea ice changing temperatures. -42 to -38!!!! Something else is going on.

This business of nighttime and winter increased warmth without daytime summer warmth is confusing. We have globally seen a decrease in cloud cover. I wonder if more sunlight on the water is keeping the water warmer. Not CO2.

Cloud cover changes are not receiving the media coverage they deserve. Even when noted, the obvious implication for temperatures is not discussed. Because any additional warmth from less clouds is temperature rise not caused by the “physics” of more CO2.

Like rises in global specific humidity while absolute humidity decreases. What is the implication of this? Because of temperature rise? The math shows this? Or does absolute humidity rise …. as cloud cover decreases? And I more humidity lead to more cloud cover?

Willis: how do you reconcile changes in cloud cover, relative and absolute humidity with temperatures (at altitude)?

Loren Wilson
January 16, 2023 1:21 pm

In scientific journals, all graphs support a table of data. If I submitted a graph without the data in a table, the editor or reviewers would have rejected the paper as incomplete and required the data in tabular form before the paper was considered ready to be published. For plots where a large amount of data are represented (Willis’s 1° by 1° analyses of satellite data, for example), it is appropriate to put the data in an appendix or on-line supplementary information.

January 17, 2023 1:27 pm

The data on the freeze-up date since 1980 seems to show a slight upward trend (toward later in the year), but if only the post-2010 data are considered, the trend would be sharply downward (toward earlier in the year).

Did the thawing of the Arctic stop and reverse circa 2010?

%d bloggers like this:
Verified by MonsterInsights