Predicting drought in the American West just got more difficult


People hoping to get a handle on future droughts in the American West are in for a disappointment, as new USC-led research spanning centuries shows El Niño cycles are an unreliable predictor.

Instead, they found that Earth’s dynamic atmosphere is a wild card that plays a much bigger role than sea surface temperatures, yet defies predictability, in the wet and dry cycles that whipsaw the western states. The study, published Monday in Science Advances, is a detailed assessment of long-term drought variability.

The findings are significant for water management, agriculture, urban planning and natural resources protection. Recent droughts have claimed many lives and caused damaging crop losses, making drought forecasting a high priority. Meanwhile, the West faces rapid population growth at the same time that forecasts show dry times ahead due to global climate change.

“The main finding is not terribly hopeful for short-term drought prediction,” said Julien Emile-Geay, a study author and associate professor of Earth sciences at the USC Dornsife College of Letters, Arts and Sciences. “We found that, historically speaking, year-to-year droughts in the western United States were less predictable than previous studies have claimed.”

New study examines 1,000 years of droughts in the West and beyond

Emile-Geay and postdoctoral scholar Michael Erb, who is lead author from USC and now at Northern Arizona University, joined with other scientists at the University of Washington and Columbia University to produce the study.

The researchers set out to answer the question: What determines droughts in the West?

They examined North American droughts and global conditions spanning more than 1,000 years. Megadroughts, which lasted decades, and dry spells predate the Industrial Revolution, American expansion on the continent or European colonialism. For example, a megadrought in the late 13th century likely contributed to the dispersal of the Anasazi people.

The prevailing explanation is that the El Niño-Southern Oscillation plays a key role in these drought episodes. The oscillation is a two-sided coin based on water conditions in the eastern equatorial Pacific Ocean. El Niño conditions occur when sea temperatures in the region are warmer than normal and are associated with wet years in the American Southwest; La Niña conditions occur when water is cooler than normal and are associated with dry years in the Southwest.

But the scientists found that rule of thumb didn’t jibe well with all drought cycles of the past. While it’s true there’s a correlation between La Niña and drought, these ocean water conditions accounted for only about 13% of the variability, the study says.

“La Niña proved to not be the only game in town,” Emile-Geay said. “La Niña is part of the game, but not the biggest part.”

A notable example of this phenomenon occurred in 2015-16, an El Niño year when Southern California did not receive the increased precipitation that was predicted. Instead, the relief came unexpectedly the following year, a La Niña year that should have been drier than normal.

What other variables can lead to drought?

The scientists also examined other drought influencers, including water temperatures in the Atlantic Ocean and volcanic activity. While those phenomena can influence drought conditions, they are too weak or episodic to explain many droughts.

Instead, the study says droughts can originate in the atmosphere. The air around Earth is highly dynamic and influenced by more variables than the ocean alone.

“The atmosphere creates a lot of variations in moisture supply on its own, and it can cook up droughts all by itself, without being told what to do by the ocean,” Emile-Geay said.

But while the past is key to the future, it does not hold all the keys. In the future, research says carbon emissions will continue to trap heat and warm the surface, and the West will experience increasingly dry conditions as a result.

“Our study suggests that the atmosphere will continue to add a strongly unpredictable element to moisture conditions in the southwestern United States, on top of drying induced by global warming,” Emile-Geay said. “That is, the Southwest is headed for a drier future overall but with the atmosphere adding a wildcard that may, at times, make things better or worse for the people and ecosystems that depend on that water.”

The study integrates numerous sources of information spanning centuries to support the findings. The dataset, called the Last Millennium Reanalysis, aggregates climate models, modern temperature measurements and nearly 3,000 climate proxies, such as tree rings, corals, and ice cores. The reanalysis was developed by scientists at USC, the University of Washington and the University of Colorado, with the support of the National Oceanic and Atmospheric Administration. According to the NOAA, the reanalysis aimed to “transform the ways in which the climate community investigates low-frequency climate.”


The study authors are Erb of NAU, Emile-Geay of USC, Gregory Hakim and Eric Steig at the University of Washington and Nathan Steiger at Columbia University.

This work was supported by the NOAA (grants NA14OAR4310175 and NA14OAR4310176), the National Science Foundation (grants NSF AGS-1702423 and NSF AGS-1805490), and by USC and NAU.

From EurekAlert!

0 0 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
August 8, 2020 2:12 am

Here yea, here yea, I have fresh bottled water from the finest hoses from France (the hoses are from France, not the water) and am letting it go for bit coins and carbon credits. Please, step in line and garner this fine commodity while it lasts. You will be the envy of your friends and can even rub elbows with cardboard cutouts of Algore and Dr. Mann before he goes bankrupt. Supplies are limited!

August 8, 2020 3:01 am

“Instead, they found that Earth’s dynamic atmosphere is a wild card that plays a much bigger role than sea surface temperatures, yet defies predictability, in the wet and dry cycles”

The internal variability of climate?

Reply to  Chaamjamal
August 8, 2020 12:24 pm

Actually, my thought was — the climate as an end-result is defined as atmospheric conditions (temp/humidity/precip/wind/storms/etc. How can the actual definition of the climate be the “wild card”?

1. If you can’t predict the atmosphere, you can’t predict the climate.
2. If you can’t predict the climate, you can’t predict that “the Southwest is headed for a drier future”.

Reply to  w
August 11, 2020 11:19 am


They had me right up to the end. They show that ENSO is not a complete predictor of drought. Then they consider other possible influences but say they are inadequate explanations. So, they admit that they still can’t predict droughts in the US Southwest.

But then, they say, “Oh well, Climate Change is the culprit, so we can expect worsening droughts in AZ & CA.”

Don’t these charlatans realize how stupid that makes them sound?

Reply to  Chaamjamal
August 8, 2020 12:37 pm

A couple of sentences later, they declare that despite their inability to predict wet and dry cycles, they know that droughts are going to get worse, because that’s what the models predict.

Ian W
August 8, 2020 3:13 am

But while the past is key to the future, it does not hold all the keys. In the future, research says carbon emissions will continue to trap heat and warm the surface, and the West will experience increasingly dry conditions as a result.

Research does not say anything about ‘carbon emissions continuing to trap heat’ , modelling assumes that carbon dioxide absorbs then re-emits infrared radiation from the surface some of which ‘down wells’ back to the surface.

Modelling then assumes that the more than 70% of the surface that is water or plants is warmed by this downwelling infrared. When infrared ‘heat’ would be absorbed by the first water molecule it hits and the added energy to a surface molecule will increase evaporation cooling the surface of the water as latent heat of evaporation is removed. The lighter humid air at the surface then convects upward carrying the latent heat with it and may dependent on the lapse rates form clouds as molecules aggregate and release latent heat some will return to the surface to enhance more evaporation the remainder will warm the ambient air increasing convection. The clouds formed by the aggregated molecules will increase albedo which during the day cools the surface further as short wave light energy that would have penetrated deep into water warming it is now reflected back to space.

Some of the modelling assumptions that are essential for ‘global warming’ need to be tested observationally. But even the simplistic diagrams showing energy ‘balance’ show far too much land area. They also make the further invalid assumption that a surface covered in trees and plants behaves just like rock when radiated with infrared and sunlight, when it is known observationally to be false.

It seems strange that entire edifices of mathematics, ‘climate projections’ have been built and government policies decided, when the base assumptions of the hypothesis are unsafe and have not even been tested observationally.

Tim Gorman
Reply to  Ian W
August 8, 2020 4:42 am


Mark Luhman
Reply to  Ian W
August 8, 2020 1:13 pm


Ron Long
August 8, 2020 3:14 am

OK, so a little scientific investigation says that the atmospheric variation that produces droughts, expressed short-term as weather and long-term as climate, is so chaotic that known controls, like ENSO, don’t produce predictable correlations, but changing the atmospheric composition by CO2 rising from 280 parts per million to 410 parts per million is an absolute Catastrophic Anthropogenic Climate Change? Trust me on this one? Next?

a happy little debunker
August 8, 2020 3:25 am

Science unsettled – same as it ever was…

Reply to  a happy little debunker
August 8, 2020 6:15 am

Predicting is easy, not hard.
The coming climate crisis is nothing
more than an annual scary prediction.
It’s getting the prediction right
that is hard to do

Stephen Skinner
August 8, 2020 3:38 am

When the Sahara became wet and green about 8k to 5k years ago it was due to extra heat coming in raising temperatures and causing an inflow of moisture. Water vapor is the most prevelant greenhouse gas as a proportion of the atmosphere and yet it was the increase in temperature first that increased water vapor. Once all the main drivers for temperature increase had shifted water vapor was unable to keep temperatures from falling and the Sahara quickly returned to an Arid climate. If one uses the mathematical ‘order of operations’ it appears that the GHG feedback mechanism may only work after all the other mechanisms have completed whatever they are doing. The tail does not wag the dog.

Brooks Hurd
Reply to  Stephen Skinner
August 8, 2020 5:00 am

This may be the researchers method of keeping their paymasters placated. The funding stream is controlled by people who are convinced that CO2 is the planet’s thermostat. The researchers appear to have discovered a very chaotic past with no easily discernable explanation of the causes of historical droughts. However, they do not want to jeopardize their future employment, thus they needed to include a reference to climate change.

August 8, 2020 4:05 am

“Instead, they found that Earth’s dynamic atmosphere is a wild card that plays a much bigger role than sea surface temperatures, yet defies predictability, in the wet and dry cycles that whipsaw the western states.”

But never fear!! We didn’t let that prevent us from making a prediction!

Reply to  Merrick
August 8, 2020 9:30 am

how does it get “more difficult” from impossible?

Stephen Skinner
Reply to  Latitude
August 9, 2020 5:59 am

Moral relativism. Such individuals can string any old words together to make something sound plausible or good enough for a childrens Sci-Fi.

Wolf at the door
August 8, 2020 4:09 am

A lot of money spent on a research that says WTF?

Reply to  Wolf at the door
August 8, 2020 10:22 am

You don’t even know, Wolf. Since 2014, J E-G has received over THREE MILLION DOLLARS to churn out this kind of charming stuff …


Loren C. Wilson
August 8, 2020 5:26 am

As a former Californian, I can predict drought every year and be correct more times than not. Their idea that increased temperature will lead to more drought is on pretty shaky ground. Overall, there will be more evaporation and precipitation world-wide. As there are several climate zones in California, at least some of these will get more rain.

Bruce Cobb
August 8, 2020 5:48 am

The problem is that they are trying to reconcile real, as in natural, climate change, with fake, as in manmade climate change. It can’t be done. But, they need more money to do it. Because hey, even fake scientists gotta eat.

Richard (the cynical one)
August 8, 2020 6:10 am

But it is refreshing to hear “We have no idea what is going on.” Now if they could just follow where the facts take them without referring to the obsolete AGW roadmap, who knows what they might find out.

August 8, 2020 6:55 am

The climate is globally connected. When you have rising air in one part of the tropics you have to have subsiding air in another. The Madden-Julien Oscillation is a great indicator of this. The pattern drifts west to east and typically, though not always, cycles. Sometimes it stalls or slows down and you get heavy rains in one area and no rain in others for extended periods of time. The point is drought in the western US might very well be due to what is going on in Indonesia or the Indian Ocean.

Another point worth investigating is the influence the Solar wind and the magnetic influence of the Sun on Earth’s weather. We all know it cycles and CME’s throw a tremendous amount of energy our way ionizing all kinds of things in the atmosphere. If you have a very wet tropical air mass that is suddenly infused with a lot of charged particles that can cause water to condense around them, you can get all kinds of thunderstorms firing off. A lot of typhoons and hurricanes form immediately after a CME hits the Earth.

Sweet Old Bob
August 8, 2020 7:00 am

“Meanwhile, the West faces rapid population growth at the same time that forecasts show dry times ahead due to global climate change.”

Define “rapid” population growth .

Scouser in AZ
Reply to  Sweet Old Bob
August 8, 2020 9:29 am

Define “…dry times ahead due to global climate change.”

Tim Gorman
Reply to  Sweet Old Bob
August 9, 2020 5:35 am

baby boom as a result of covid? Democrats finally winning on totally open borders?

Kevin kilty
August 8, 2020 7:13 am

Lives lost to drought? Are we speaking metaphorically or are these real deaths? Are they like climate refugees? References and body counts, please.

Two decades ago I looked carefully at ENSO cycles since the 1940s and their correlation to weather in the interior West, and I could find none.

One will note that an absolute reliance on beliefs continues, though, because…” warmth will cause drying.”

Joel O’Bryan
Reply to  Kevin kilty
August 8, 2020 7:29 am

With the decline in science integrity and politicization of science in recent years, it is very predictable that academic researchers will feel “compelled” by pressures from university administrators to make apologies for their “white privilege” in their manuscripts and press releases, and how it likely “colors” their conclusions. It’s coming.

Joel O’Bryan
August 8, 2020 7:21 am

Recent droughts have claimed many lives

Really? I guess they need to define “recent”. Is 1275 AD “recent”? A geologist would say, “yes.” A meteorologist would say, “No.”
Remember, we’re discussing the US Southwest, not Ethiopia.

Their other point about “carbon emissions” also triggered my BS alarms, but Ian W. discussed that specious claim (above comment) already. Since they were studying the past back a millenium, that was outside the bounds of their data and expertise, Total genuflection to the climate scam on the researchers’ part. There is nothing in their data that informs on how future climate and droughts will play out regarding an emissions-related pCO2 rise.

August 8, 2020 7:52 am

“Recent droughts have claimed many lives and caused damaging crop losses …”
What a sweeping, vague assertion. In a state that is largely dependent on water capture, storage (surface and aquifer) and basin transfer to provide drinking water and crop irrigation, who in California has died of thirst due to drought? Urban water gets priority over farmers, so nobody should have died of thirst in a drought. As for crop losses, that is largely a function of planning and building sufficient storage and distribution systems, developing more unified/integrated water management programs, and improving water use efficiency (e.g., advanced irrigation technologies; crop varieties; farming practices). Marginal farming areas may suffer in an extended drought, but look at THE DATA. Despite the authors’ assertions, California’s statewide crop yields have continued to rise on average >1% per year over the last 4 decades. Of course, eventually, population growth, particularly in the southern, dry part of the state, will begin to bump up against the limits of reliable (drought resistant) water supply. Maybe the state should look into limiting immigrant population growth, since there is already a net out-migration of 100-200 thousand residents per year.

Since this paper basically concludes, “we can’t predict drought (shrug),” just keep on working toward better, more efficient allocation and use of water. Well, that’s what they have already been doing for decades. Climate change? That has had and will have little to do with the state’s ongoing water challenges and is here admitted to be unpredictable.

Jeff Alberts
August 8, 2020 8:20 am

“They examined North American droughts and global conditions spanning more than 1,000 years.”

Or did they examine proxies? Which may or may not respond to drought conditions all the time.

Reply to  Jeff Alberts
August 8, 2020 9:44 am

It says they examined 3000 proxies….of course never stated that the proxies often disagreed with each other….but it was good of them to actually state a negative result, something most dubious claim climstudies avoid, in order to continue the grant gravy train.

August 8, 2020 8:40 am

Grant droughts are closely linked to atmospheric CO2 levels.

Steve Oregon
August 8, 2020 9:59 am

So it appears there is just too much natural, random chaos with crap shoot roll of the dice variations of every influence, (Known and unknown) to model.
But when unknowns abound unscrupulous scientists get creative.
A poster child for this kind of pretense of science is the Lubchenco, Barth & Chan sham at OSU that concocted a conclusion that global warming was causing a “novel rise in ocean hypoxia”.
These shameless university professors crafted from “no records available” a tall tale of AGW ocean dead zones which inexplicably remains unchallenged by WUWT or any other scrutinizing entity.
Here is a bit of their invented science. Careful how you comprehend the claim of “Five decades of available records”. There are no real records of hypoxia. Just useless, random ocean water samples going back to 1950, assumptions of significance and pieces of meaningless anecdotal observations.

” The rise of anoxia has occurred against a backdrop of recent increases in the frequency and severity of shelf hypoxic events in this system (Fig. 1B). Five decades of available records show little evidence of shelf hypoxia and no evidence of severe inner-shelf hypoxia before 2000 (Fig. 1A). Recent studies indicate that the onset of shelf hypoxia can reflect basin-scale fluctuations in atmosphere-ocean processes that alter the oxygen content of upwelled water, the intensity of upwelling wind stress, and productivity-driven increases in coastal respiration (5, 6). Strongly coupled atmospheric and oceanic circulation underpins ecosystem dynamics in wind-driven upwelling shelves and ecosystem susceptibility to modulations of upwelling wind stress from climate warming (7, 8). The present-day global distribution of shelf anoxia reflects broad cross-system differences in vertical proximity to OMZs, shelf productivity, and circulation. The novel rise of shelf anoxia in the CCLME highlights the potential for rapid reorganization in the distribution of anoxia and the sensitivity of productive upwelling shelves to discontinuous ecosystem change.”

August 8, 2020 10:45 am

They are using something called the Last Millennium Reanalysis (LMR) as their gold standard for what yesterday’s climate looked like. This is not surprising since the LMR is the brainchild of the same authors of this POS. Julien Emile-Geay has received over three million dollars in grants since 2014 to pursue this blind alley.

Here is the authors description of the LMR:

An “offline” approach to DA [“data assimilation” – w.] is used, where static ensemble samples are drawn from existing CMIP climate‐model simulations to serve as the prior estimate of climate variables. We use linear, univariate forward models (“proxy system models (PSMs)”) that map climate variables to proxy measurements by fitting proxy data to 2 m air temperature from gridded instrumental temperature data; the linear PSMs are then used to predict proxy values from the prior estimate.

Let’s unpack that a bit. They draw “static ensemble samples” from a bunch of the usual crappy climate models to give us a “prior estimate of climate variables”. In other words, they start with a TOTALLY IMAGINARY CLIMATE as reconstructed by averaging bad models.

Then they use other unverified and unvalidated models, called “proxy system models (PSMs)” to fit proxy data to modern temperature data … a procedure that is well-known to “mine” for hockeystick-shaped data.

Finally, ignoring the GIGO rule, the original garbage from the crappy climate models is used as input to their specially tuned proxy system models … and the output from that is claimed to represent reality.

We are solemnly assured that none of the dozens of tunable parameters involved in that process were harmed in any way, and that the outputs of that chain of errors and misunderstndings are solid science …


Reply to  Willis Eschenbach
August 8, 2020 10:49 am

Oh, yeah. Steve McIntyre points out that in addition to cranking out bogus science, J E-G is a real ugly piece of work …


Reply to  Willis Eschenbach
August 8, 2020 6:28 pm

Many thanks to both Willis and Steve McIntyre!

Reply to  Willis Eschenbach
August 8, 2020 3:34 pm

Wonderful description Willis, thank you.

August 8, 2020 12:34 pm

“Meanwhile, the West faces rapid population growth”

Another myth, most western countries are facing population declines. Very few western countries have birth rates at or above replacement levels. Most are below, to well below.

August 8, 2020 12:36 pm

Did I read this right? First they declare that they don’t know what causes droughts, but then they declare that droughts are going to get worse because the models predict it.

Ulric Lyons
August 8, 2020 3:43 pm

“A notable example of this phenomenon occurred in 2015-16, an El Niño year when Southern California did not receive the increased precipitation that was predicted. Instead, the relief came unexpectedly the following year, a La Niña year that should have been drier than normal.”

The rain in California did begin to return from late 2015 and increased slightly through 2016, but it could not increase strongly until the Northeast Pacific Warm Blob had dissipated from after mid 2016. The deluge in early 2017 follows the same pattern as the early 1879 California floods after the 1877-78 Super El Nino.

Jude Richardsn, forester
August 8, 2020 4:12 pm

Used to be I needed I to know weather weekend predictions so that I could plan defensive firefighter operations. I relied on looking out the window at cloud patterns. I often suspected that forecasters could be more accurate in their predictions if they just looked out the window.

August 8, 2020 4:42 pm

Try looking at the 18.6 year lunar declination cycle. The very wet years can also be predicted.

Mike Dubrasich
August 8, 2020 6:12 pm

“Our study suggests that the atmosphere will continue to add a strongly unpredictable element to moisture conditions in the southwestern United States, on top of drying induced by global warming.”

A warmer atmosphere contains more water, not less. During glacial stadials precipitation declines. At the LGM rainforest cover diminished and was replaced by grasslands or desert. During the early Holocene, when global temps were higher than today, rainforests expanded and deserts bloomed.

PV=nRT. Increasing T increases P. High humidity occurs when the partial pressure of water vapor approaches the equilibrium vapor pressure. Saturation increases with T. Cold air is dry air. Prove me wrong.

August 8, 2020 6:33 pm

“Instead, they found that Earth’s dynamic atmosphere is a wild card that plays a much bigger role than sea surface temperatures, yet defies predictability”

Over three decades of climate alarmists claiming to ‘know all, see all, predict all’ this is a blunt admission of ignorance. Even when they are unable to admit their personal fallibility in all things atmospheric, oceanic and land surface.

A real piece of work… Not!

Mark Luhman
August 8, 2020 7:01 pm

“Predicting drought in the American West just got more difficult” Not true at all only some useful idiots finally figure out you can’t do it. To said idiots, it must be a shock you, to the rest of us it is a yawn. To the grant people are you really that stupid to give money to stupid people. Of course they all are the same crowd so they don’t know they are stupid.

%d bloggers like this: