The West Burns and the East Drowns … so it averages out, right?

Willis Eschenbach is wrong on both counts when he announces at the WUWT blog that “… according to NOAA, there’s been no increase in either droughts or wet periods in the US since 1895 …”

He treats us to this graph of PDSI (the Palmer Drought Severity Index) averaged over the 48 states of the continental U.S.A. since 1895:

He also informs us that the trend line shown does not pass “statistical significance,” which means it doesn’t give enough evidence to conclude that there’s any trend at all. Maybe some other test would — but this, the most common one (fitting a straight line), doesn’t.

Here’s my graph of the same data, showing yearly averages:

I also put the negative numbers at the top, positive at the bottom, since negative numbers mean dry and positive mean wet. I also made those more negative than -2 red, those more positive than +2 blue, and put in a trend line (which is not “statistically significant”).

When PDSI in your area dips low (below -2 or so), it’s a good indicator that your region is in drought. When it gets above +2, you’re definitely not in drought and you might be flirting with flooding (but for flooding rather than drought, there are better indicators than PDSI). The problem is, that when we average over the entire continental US, the negatives and postives cancel each other in the average. But that does not mean conditions are “average”! If one region shows a trend toward more drought, while another trends toward flood, the fact that the average trend is nothing at all will be little comfort to the one or the other.

NOAA (who provide the data Eschenbach used) also has regional data. Here’s PDSI for the Western U.S. climate division:

I’ve added a red line showing an estimate of the trend (which is not a straight line) and pink shading to reveal its uncertainty range. The relevant point is that it has shown a distinct increase, and one way to show that is the simple “fit a straight line” test. The result is “statistically significant” (although the trend isn’t really a straight line). Any way you look at it, increase in droughts, statistically significant.

For the U.S. Central climate division the picture is quite different:

Once again I’ve added a (nonlinear) trend line and its uncertainty. PDSI hit its highest value (lowest on the graph) this year, conditions were so wet. Precipitation did too. And there’s that trend line too. PDSI and a lot of evidence tell us increase in wet periods, statistically significant.

If you were looking for clues about how climate change will affect your area now and in your future, bear in mind that Willis Eschenbach is wrong on both counts.

Happy holiday!

This blog is made possible by readers like you; join others by donating at My Wee Dragon.

17 responses to “The West Burns and the East Drowns … so it averages out, right?

  1. Willis’ work is done in bad faith from a scientific point of view. His work advances an ideological agenda. To present this kind of work as scientific is dishonest. Thanks for calling BS on this stuff. It is unfortunate, but a lot of voters are drawn to the underlying ideology and don’t recognize that the work that WE produces is a sham.

  2. Reminds me of the joke or sometimes cartoon which pokes fun at statisticians who often champion the mean as a summary statistic without enough justification, shouting about Chebyshev’s inequality or Markov’s inequality. These remain true, but are not helpful in the case presented.

    In particular, the problem at hand is usually showing two bullseye archery targets, and the Statistician decides to aim at the mean of their locations. Of course, that’s empty space and the arrow zooms passed them.

  3. Can always count on Willis Eschenbach to be sloppily wrong. His only skill is that he writes fast and can pound the keyboard hard. A good example of his sloppy work was from a year ago in the midst of the raging Paradise/Camp fire where he claimed that warming in California was exaggerated by an order of magnitude according to his chart of the data. It turned out he messed up on converting the numbers among other goofs and waited several days before correcting his charts. By that time all his blog readers had moved on and the damage was done.



    his blog post here:

  4. Hmmm. This amounts to an error of the sort “a recent statistical survey has definitively shown that the average American has one breast and one testicle”.

  5. It’s probably a cheap shot, but my honest reaction to the lede was “Willis Eschenbach wrong? Now *there’s* a shocker!”

  6. Southeast of Australia…

    That’s one megafire…

  7. Just posting this update on Aussie bush fires, though off-topic. The extent is extraordinary and our volunteer services are stretched (thanks to international firies lending a hand).

    According to wiki the area burned for NSW/QLD is the broadest in a single year. At 52 years, I’ve never seen anything like this fire season, and I’d be curious to see some numbers put on it.

    Here’s to a better New Year.

    • Best of luck and wishes to everyone, Barry!
      Jan, from fortunate Massachusetts.

    • This is the kind of national event that might wake up an electorate and help them choose appropriate climate action. Australia is much like the US in its climate politics, but maybe that will shift based on this fire. Environmental activists may not need to get the messaging just right to persuade Aussies who have seen this fire season to choose differently when they go to vote.

      My sense is that the bad faith actors like WE will always punch above their weight and IQ because their message is well-funded by entrenched interests and the bad faith messaging has been difficult to defeat no matter how persuasive and reasonable the message might be from the hippies and enviro crazies.

      It has been my sense now for several years that change will only come through widespread, first hand experience of the bad outcomes that arrive as the planet warms.

      To my Aussie mates, I say: leave the coal in the ground now. It’s a disaster to get it out of the ground. It’s a stranded asset. We can’t burn the coal without getting scorched in the process.

      Also, as always with fires, stay safe, don’t lose your life trying to save your property.

    • I spent Christmas in the Adelaide Hills with my Dad and checked the sprinklers poking out of the exterior walls of his house, a faint buttress against bush fire that would consume his property if it came through. The Cudlee Creek fires of last week razed vineyards half an hour’s drive from his home. We saw the remains when we drove out to lunch at a nearby winery.

      Came home to Sydney yesterday and saw what I saw 5 weeks ago when I left to work in Adelaide – haze across the little valley in Bondi where the land meets the sea. I’ve missed the worst of it – 11 times the air pollution safe level. And fires have sparked up near the edges of the outer city while I’ve been away. It’s not the first time Sydney sky has been darkened with smoke (or beautifully reddened with dust), but never in Sydeny’s history has this haze gone on for weeks.

      Of all the temperature records broken in Australia over the last few years, the droughts and other indicators of anthropogenic climate change, this bush firre season is by far the most powerful signal. We have every state in this wide land affected by these fires, even our nation’s capital, parts of which which have recently been inundated with smoke 13 times the safe margin.

      For those directly in the line of fire, over a thousand homes destroyed across the counry, thousands of livestock killed, unknown numbers of native and introduced species dead, nearly 6 million acres burned since September, and the end is not yet in sight.

      This is not normal. This isn’t a state emergency, as it has always been in the past, it’s national. It’s all over the news. It’s in our air. This season is like nothing we’ve ever seen, That’s not hyperbole. That’s what the fire services say. That is what everyone knows. A fire commissioner on TV described the freak events that led to an 8 tonne truck being flipped by the by the wonds created by a firestorm, a fatal event for some firies trying to save property and life. He said even service veterans could not believe what they witnessed with their own eyes. We have reports of 70 meter (230 foot) high fire fronts, and we have images of 14 kilometer high smoke columns making their own weather systems.

      I am very interested indeed to see how this bush fire season affects slimate policy in Australia, because nothing else before comes close to tipping the argument.

      This is a litmus test. If you are interested in how reality on the ground affects policy WRT anthropogenic climate change, then you should watch Australia’s response to this astonishing and appalling bush fire season. This is not normal, not by a long shot.