Take a look at this picture:
This year, my little brothers and I had to scrap plans for spending a week at Lake Nacimiento in the Coastal Range west of the Salinas Valley, because water levels were so low it would be a hike from many lakeshore cottages down to the lake. You want pretty on a vacation, and acres of cracked mud isn’t all that pretty. (Also, tends to be in the 90s at the lake, which makes being on or in it more important.
Two things: First, Lake Oroville and Lake Nacimiento are both reservoirs, meaning that we human beings decided a lake needed to be in a place where nature had not put one. Second, yes, hasn’t been much rainfall the last couple years. When you put the two together – reservoirs collecting water where water didn’t naturally collect, and a lack of the rain and snow that supply the water those reservoirs are filled with, you get dry reservoirs.
A friend was visiting from Vermont, and her kids thought everything looked so dead out here – by late August, the Golden State is just the Dead Brown State. But that happens every year regardless of how much it rained back in February. She pointed out that the trees looked stressed. Only the transplants, I pointed out back – the native oak and walnut looked about how they always look this time of year, while the fancy trees favored by landscapers did look a bit ragged and brown.
I checked out the rainfall totals for Los Angeles over the last 135 years (Concord’s figures were not readily available on the Web – LA’s were. Go figure.) I did a quick eyeball test, and it looked like about +/- 7″ of rainfall from the average was “normal”. Out of 135 years covered in the records, exactly 100 times the annual rainfall totals were between 8″ and 22″ – 7″ either side of the 15″ average. So, one might conclude that “normal” annual rainfall is anything between 8″ and 22″. Seems reasonable.
Now, a real statistician who, unlike me, would not need to crack some books and look some stuff up to do the math (I think, since I don’t have a grant, I may be legally prohibited from doing so in any event) could fine tune this up a bunch, figure the standard deviation, plot a curve, and make it all scientifilicious – and reach about the same conclusion as my eyeballs. The point here is that calling rainfall totals a drought when they fall well within the “normal” range that one would expect is curious. Pointing out dried up reservoirs and ratty looking non-native liquid amber trees as evidence of a drought is also suspect. When the natural lakes start drying up, and the big native trees start suffering, I’ll be a lot more worried.
So, we’ve had 3 years of below average rainfall – but well within the expected range of “normal”, It was not enough to keep the reservoirs filled. I’m hoping this year we get 20″ of rain and a hundred feet of snow in the Sierra, because then we can go back to worrying about something else. But it’s a roll of the dice – it’s no more likely that 2014 will be a wet year rather than a dry year versus the average. Sometimes, you’ll get 4 in a row. I’m not sure we’d be in a drought even then – we’re in the middle of a water management challenge. Happens when 35 million+ people move to a semi-desert area.
Back to the picture above. The caption to that picture reads:
A stream of water trickles through a section of Lake Oroville that used to be under water on Aug. 19, 2014, in Oroville, California. As the severe drought in California continues for a third straight year, water levels in the state’s lakes and reservoirs is reaching historic lows. (Photo by Justin Sullivan/Getty Images) | Justin Sullivan via Getty Images
Yep. But the article that this picture accompanies isn’t about the historical natural variations in rainfall. It’s about:
Why, yes! Yes, it might!
“This will be worse than anything seen during the last 2,000 years and would pose unprecedented challenges to water resources in the region,” Toby Ault, lead author of the study, said in a press release Wednesday.
Why is it ‘might’ in the headline, but ‘will be’ in the money quote? Strolling through the article, we find the bugaboos:
Thanks to the expected effects of climate change, there’s at least an 80 percent chance of a decade-long drought occurring in the Southwest over the next century, a new study has found.
You mean, those effects expected but not actual over the last 17 years? And, most important and scientifilicious of all:
Ault is part of a group of researchers from Cornell University, the University of Arizona and the U.S. Geological Survey who used computer models to estimate the likelihood that a drought lasting 10 years, 35 years or 50 years could occur in the Southwest in the next century.
They used a computer model, so it must be true! Notify investors in Long Term Capital Management!
And, you know, Cornell, which is reputed to be a fine institution, seems to have its fingers in a surprisingly large percentage of junk science. Starting with Sagan and going downhill from there. Someone should look into that.