Stories
Slash Boxes
Comments

SoylentNews is people

posted by martyb on Thursday March 07 2019, @03:22PM   Printer-friendly
from the global-warming-vs-local-cooling dept.

The largest of the great lakes in the United States, Lake Superior

Lake Superior’s ice coverage has greatly surpassed expectations this year.

Earlier in the season, forecasters predicted the lake would reach a little more than 50 percent ice coverage this winter. But as of Friday, Lake Superior was over 85 percent covered, far exceeding the prediction and the lake’s long-term average of 55 percent, according to the National Oceanic and Atmospheric Administration’s Great Lakes Environmental Research Laboratory, or GLERL.

This year’s frigid conditions triggered the rapid expansion of the ice that exceeded predictions, said Jia Wang, a research ice climatologist and physical oceanographer at GLERL.

[...] Earlier this week, ice coverage increased about 10 percent within 12 hours, rising from around 75 percent at 2 p.m. Wednesday to nearly 85 percent by 2 a.m. Thursday.

[...] The last time the lake ice reached 100 percent coverage was 1996, which is the only time 100 percent coverage on Lake Superior has been noted since records started in 1973, according to GLERL data.

http://www.ironmountaindailynews.com/news/local-news/2019/03/lake-superior-ice-coverage-nears-90-percent-exceeding-predictions/

[Updated to fix title; changed "Exceeds 90 Percent" to be "Exceeds 85 Percent". --martyb]


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 0) by Anonymous Coward on Friday March 08 2019, @02:12AM (8 children)

    by Anonymous Coward on Friday March 08 2019, @02:12AM (#811435)

    No, in Simpson paradox every subgroup can be cooling while still having an overall warming.

  • (Score: 1) by khallow on Friday March 08 2019, @04:24AM (7 children)

    by khallow (3766) Subscriber Badge on Friday March 08 2019, @04:24AM (#811455) Journal

    No, in Simpson paradox every subgroup can be cooling while still having an overall warming.

    No, that is not Simpson's paradox. For arithmetic means over such sets, like the temperature of regions of the Earth's surface, Simpson's paradox is that a proper subset, even a nearly complete one can mathematically have a very different mean than the whole due to far different values in the parts left out.

    But once you have everything counted, you have the whole. And Simpson's paradox no longer applies. For a math proof, suppose we have N components with fractional weight x_k (fraction of surface area in the case of calculating mean temperature of the surface of the Earth) and mean temperatures for each part is T_k, thus, the formula for the mean temperature of the whole is the sum of x_k*T_k, k=1 ... N (with sum of the x_k equal to 1). If each T_k is less than a temperature T, then the corresponding term of the average, x_k*T_k x_k*T. Thus, the average, sum of x_k*T_k (sum x_k)*T = 1*T = T. That is, if the temperature of every part is less than T, then so is the average temperature.

    • (Score: 0) by Anonymous Coward on Friday March 08 2019, @04:51AM (6 children)

      by Anonymous Coward on Friday March 08 2019, @04:51AM (#811462)

      That isn't what the data is like. You need to have a variable number of stations in each region for each year and calculate a trend for each, then the overall trend using all the stations at once.

      • (Score: 1) by khallow on Friday March 08 2019, @06:13AM (5 children)

        by khallow (3766) Subscriber Badge on Friday March 08 2019, @06:13AM (#811469) Journal

        You need to have a variable number of stations in each region for each year and calculate a trend for each, then the overall trend using all the stations at once.

        Which has nothing to do with Simpson's paradox.

        • (Score: 0) by Anonymous Coward on Friday March 08 2019, @08:40AM (4 children)

          by Anonymous Coward on Friday March 08 2019, @08:40AM (#811485)

          Sure it does, the final average isn't just the average of the regional averages.

          • (Score: 1) by khallow on Friday March 08 2019, @03:14PM (3 children)

            by khallow (3766) Subscriber Badge on Friday March 08 2019, @03:14PM (#811539) Journal
            Then it is not a average, final or otherwise. Words have meaning.
            • (Score: 0) by Anonymous Coward on Friday March 08 2019, @03:44PM (2 children)

              by Anonymous Coward on Friday March 08 2019, @03:44PM (#811558)

              I don't think you even looked at the simpson's paragraph page, it is all due to varying sample sizes.

              • (Score: 0) by Anonymous Coward on Friday March 08 2019, @04:24PM

                by Anonymous Coward on Friday March 08 2019, @04:24PM (#811580)

                Blame autocorrect: paragraph -> paradox

              • (Score: 1) by khallow on Friday March 08 2019, @05:03PM

                by khallow (3766) Subscriber Badge on Friday March 08 2019, @05:03PM (#811591) Journal

                I don't think you even looked at the simpson's paragraph page, it is all due to varying sample sizes.

                The original assertion that all regions of the world can have cooling trends while the global temperature is warming is still wrong. The example given in the article was of grouping observations along the same parameter that the trends were determined along. That introduces observation bias. That bias doesn't exist when one groups temperature readings by spatial region and looks at temperatures in time. One no longer has the conditions leading to the spurious trend creation.