Tag Archives: truncated data

Stirling and Derocher’s sea ice trick – omitting facts to make polar bears appear endangered

Polar bear biologists Ian Stirling and Andrew Derocher continue to insist that western Hudson Bay (WHB) polar bears are already showing negative effects of reduced sea ice due to global warming. In their 2012 summary paper (Stirling and Derocher 2012), they updated someone else’s graph of global sea ice (to 2011) but used a graph for Hudson Bay (HB) ice breakup dates that ended in 2007. However, we know from other evidence that at least one of those years (2009) would have required extending the scale of the breakup date graph upwards and flattened the slope of the trend line. Updating the HB breakup date graph would not have supported Stirling and Derocher’s premise that polar bears in WHB are starving due to increasingly earlier sea ice breakup, so they simply left the data out (see Fig. 1).

In other fields, this is called fraud.

Is it fraud here? You decide.

I’ve expressed my outrage about this before (here and here), because we know from news reports that in 2009, breakup of Hudson Bay sea ice was unusually late: the Port of Churchill (in WHB) did not open for ship traffic until Aug. 12, a full three weeks later than average (July 21) – and the latest opening of the Port since records began in 1974.

I try not to keep thinking of Stirling and Derocher’s unscientific behaviour but was reminded of it again on Monday (March 4) when I attended a lecture at the University of Victoria given by paleoclimatologist Michael Mann. To my disbelief, Mann tried to argue that global temperatures predicted by NASA scientist James Hansen in 1988 have “closely resembled” actual temperatures since then – by presenting a graph of actual temperatures (observations) that ended in 2005, despite the fact that recent temperatures have not risen at the rate depicted in his graph (see previous post, #8). He did say, as an aside, that “you could argue that if the data were extended out to the present, the line might more closely resemble scenario C [a flat line]” but then continued with his story that observations were matching the ever-rising-temperatures of Hansen’s scenario B (see Figure 2 below).

For both parties – Stirling/Derocher and Mann – the recent data points left off their graphs did not fit their narrative: sea ice in Hudson Bay is not on a steady, precipitous decline and global temperatures have not continued to rise as predicted by Hansen in 1988. The graphs look like science, but they are not.

Continue reading