The ice extent in Lake Superior dropped from 92% to 85% in the middle of a sub-zero night, without a storm to rip apart the ice, which seems very odd. Perhaps it involves some sort of correction or adjustment being made to the data. However when such corrections are made they should be explained, especially when news items have come out about the 92% setting a new record.
Chris Beal, (AKA “N.J. Snow Fan”) immediately noticed and documented the drop in ice extent.
I have no problem with making corrections, as I am a person prone to making mistakes. However it is important to document your corrections. If you don’t, it can look like you are trying to hide a mistake or, worse, attempting to fudge data and create false impressions.
We live in a world where “spin” is unfortunately a given, rather than being seen as a dishonesty people should feel ashamed for. Consequently the public is especially on guard. People understand that, “The price of Freedom is eternal vigilance.” Therefore any “correction” or “adjustment” you make is likely to be noticed, especially as, “The web never forgets.”
If some correction was made to the Lake Superior data in the dead of night, perhaps to hide a blunder which created a “record-setting” 92% ice-extent in the first place, it was a big mistake. The whole web is watching. Such corrections should be made humbly in the light of day.
I don’t imagine the boss will be happy, when he comes into work this Monday morning.
SELF-CORRECTION —AM I SIMPLY TOO SUSPICIOUS?—
Upon thinking of the 5% change in listed ice-cover more deeply, I commented as follows over on the WUWT site.
Another thing that might have led to the “decrease” is the simple fact the wind slacked off, which gave various cracks and openings in the ice time to warm the air closest to the ice. That in turn could effect the model’s readings and interpretations of whether the area viewed is ice-covered or not. In other words, it isn’t a visual thing, looking at actual ice versus actual water. Rather it is a measurement of temperature, which is plugged into some best-guess formula, which decides if it is water or ice. (I don’t know this; it is merely my benefit-of-the-doubt surmising.)
For 5% of the ice to vanish might simply involve the 2 meter air temperatures rising from zero to thirty, which can occur over incompletely ice-covered lakes when the wind slacks off. It has nothing to do with the ice actually forming or melting. The model is simply programmed to think the ice is there when the temperature is zero, and think this ice cannot be there when the temperature is thirty.
In such a case no data-tampering need be involved. Instead we are just witnessing the imperfections of a model.
(Please notice I adjust my views and correct myself in the broad light of day.)