Sunday, September 09, 2012

Guess: September is Going to Be Very Warm

I know the data only goes through the 7th of the month, but my dumb little model is saying that the UAH LT value for September is going to be very warm: at the moment, it says +0.59°C above the baseline, which would easily be the warmest September in their 30+ year records, and the 3rd warmest month of all.

I've found in other recent months that the final UAH LT value changes only by at most ±0.1°C from this point to the end of the month. (Partly because I rely on a month-long average; and I simply assume that the cooling rate for each future day will be the average daily rate since 2001: -0.015°C/day for September.) That gives 0.59°C.

My guess last month was pretty close: I guessed +0.31°C, and the final value was +0.34°C. My method isn't much more than looking for a linear relationship between the monthly average of the raw temperatures and their published values, with a correction factor added based on how well (or poorly) I did in previous months.

My suspicious is that my method will work OK until it doesn't, which will come when (I suspect) UAH makes a periodic (~12-18 months?) correction for satellite factors like a declining altitude.

And there does seem to be an El Niño forming (regardless of what Roger Tallbloke says); NOAA said last week "El Niño conditions are likely to develop during August or September 2012." (Note: this link will be out of date in a week.)

Take this for what it's worth, but you heard it here first (unless I'm wrong).

PS: Notice I put the "~" above the 2nd n in "El Niño." When I was an undergraduate, my best professor, from whom I learned an enormous amount of good physics, was a big fan of using this symbol for variables, akin to the way some people use a prime symbol, as in "a'". He called it "twiddle." The classroom had a chalkboard on all four walls, and desks that rotated. He would start up in one corner, and work his way around the room, usually having to erase before he was done.

PPS: That reminds me of a story (that I've probably told before) about Robert Oppenheimer, who, during a lecture, referred to the equation "beneath" an equation he was pointing to. Someone says, excuse me Dr. Oppenheimer, but there is no equation underneath the equation you're pointing to. Oppenheimer said, "No, not underneath, beneath -- I've erased and written over it."

No comments: