See Comments down arrow

Models versus observations: Greenland edition

13 Feb 2019 | Science Notes

We should worry about Greenland, they say, because if it warms up a lot all that land-based ice will slide into the ocean and raise ocean levels in ways melting sea ice does not. OK, but is Greenland warming the way models predict? Over at Watts Up with That Bob Tisdale has compared the climate models used by the IPCC to a special Greenland surface temperature data set over the period 1925 to 2012 prepared at Berkeley university. The models on average say it should have warmed by +0.161 degrees warming per decade. The observations show… cooling. But who needs data when we have models.

The Berkeley data set shows a slight cooling of -0.006 degrees per decade since the 1920s. Moreover, the temperature record actually goes back to 1900, but it reveal a strong cyclical pattern. Tisdale notes that it is misleading, to put it mildly, to compute the trend over one-and-a-half cycles, since starting at the bottom of one cycle and ending at the top of the next biases the trend upwards. (These tricks are the sort taught in Darrell Huff’s classic How to Lie with Statistics not in university geology courses or in physics labs… we hope.) Amazingly, even using this statistical legerdebrain only yields a trend of +0.099 degrees per decade. Even fudging the data yields a result well below the model average.

Tisdale also notes that in addition to getting the trend wrong, the models are all over the place at estimating the current average temperature in Greenland. Across the various models used the difference among estimates span 7 to 8 degrees.

So they can’t reproduce the current temperature of Greenland to within 7 degrees, and they get the warming rate completely wrong over the past nine decades. I joke in my history classes that I became a historian because it’s easier to predict the past than the future; I get the War of 1812 right every time. But for the models the reverse is true. They muff the known history of the 20th century, the 19th or any time period you care to name. But they’re even better than your horoscope looking forward, so good we can base the future of our economy on their predictions about the impacts of climate change.

I gaze into my crystal ball and see … gullibility.

Leave a Reply

Your email address will not be published.