The Amazon is Cool

Well, it's summer in the North, and the birds are singing, the bunnies, well, you know what, and people all over are sweating.  It has been particularly bad in Europe, the Western US, and disastrous in South Asia, both India and Pakistan.

As Eli could have told you the search is on for excuses, especially in light of new papers which attribute and increase in heat waves to climate change.  Some, not Eli to be sure, recommend very large air conditioners to handle the problem.   ARPA-E has been working on that, well along the line of solar and thermal driven units, not extra jumbos, and there is progress.

Others, actually the same some playing the others card, are going the ostrich route, nothing happening here, move on.  This has lead to a cheerful back and forth between La Curry and El Tamino.  Curry posted some slides from a talk by NOAA's Prashant Sardeshmukh describing shifts in mean temperature and the standard deviation of the temperature distribution and coming to the conclusion that there has been no increase in heat waves.

Now to be honest there are some issues with this.  First, the slides only deal with changes in December-January-February, which is the northern hemisphere winter (Yes, Eli knows about Australia and the newly popular concern of some of those who block the Bunny's tweets with Africa) but most land and most people find themselves in the northern hemisphere, which, also is where they get hit by heat waves during the summer.  Oh yeah, it's real TLT out there, the temperatures are from the 850 mbar level where no one is hot because of the lapse rate, and from reanalyses, not measurements.  Of course Eli expects that such manipulation would not be allowed in the blogs of denial.  Eli is often disappointed

Curry shows a slide of temperature distributions by Sardeshmukh defining a heat wave as when the temperature exceeds a fixed limit.  Curry then shows a slide by Sardeshmukh showing the global distribution of temperature changes,

and the global distribution of the change in the width of the temperature distributions

and, then what Sardeshmukh purports to be the probability of a heat wave

Red is increase, blue decrease.  Apologies to the color blind and Doug McNeall will be here in a moment.  Somewhat seriously, Uncle Rabett was so color blind that you could not let him out of the house without checking his dress, which often followed an early Rowan and Martin theme if not examined by Aunt Rabett.  There has to be a color shifting app to handle that.

Eli will let Tamino, who noticed this explain in detail, why this is sausage.  If you look at the little box on the right, it is an area where both the temperature and the width of the distribution go up, but the probability of a heat wave, according to Sardeshnukh go down.

How is this done, well, according to Curry, lots of air conditioners are installed as the world warms and this means that for a "real" heat wave the temperature at which one is declared, goes up too.

Go read the tweets and especially Tamino's two posts (one, two)  Sardeshnukh has weighed in at Curry's claiming that he did define heat waves as being past the post at a fixed temperature, and Tamino is asking (politely, as Mozart would) for the data.

Curry, however, was at least this morning, of Tamino's opinion about what Sardeshnukh had done

Tamino’s argument is essentially a quibble about how heat waves are defined, there are various definitions

For the heat wave forecasts that my company provides to the energy sector, so they can anticipate high energy demand, we define heat wave in terms of the standard deviation above the climatological mean for that location (we use 1.5 standard deviations for the energy demand applications, whereas Sardeshmukh used 2 standard deviations). For our heat wave forecasts in Ahmedabad India, we use specified temperature thresholds. Other definitions are tied to a specific temperature increment, e.g. 5C above daily average.

Sardeshmukh’s analysis uses two different baseline temps: one prior to 1950 and the other post 1950, and then calculates deviations from those means. His whole point is that the standard deviation and skewness changes can dominate, resulting in fewer large excursions from the mean.

As Tamino said in his first post

Which makes me wonder, what the hell is going on? If he was trying to emphasize that we can’t use the change in mean and standard deviation to understand the 2-sigma exceedance, well duh. When the mean goes up and the standard deviation goes up and you change the limit to define “extreme” temperature, you should expect nothing less.

All this will play out relatively quickly, but it did inspire Eli to play a bit with the temperature and standard deviation figures above using (shudder Powerpoint, which has some interesting features.  By sharpening up the colors and overlaying the two figures One can make out areas where both temperature and standard deviations changed in the same directions and in opposite directions

If they both increase, the color is red, if one increases and the other decreases purple.  There is ONLY one spot where both the temperature and the standard deviation decreased, right about at the mouth of the Amazon.