By Andy May
The featured image is a photograph of the Big Bend area in Texas in October 2015 when the drought index was moderately moist. The photo was taken by the author.
Stories of some sort of current catastrophic drought in the western U.S. are greatly exaggerated (apologies to Mark Twain). While the western U.S. is dry due to its location under a zone of perpetual high atmospheric pressure and little rain, the present drought conditions in the region are moderate by historical standards. This post is mostly an update of two earlier posts by Anthony Watts here and here. Watts’ posts have stood the test of time quite well. It is amazing how often the “consensus” recycles old, debunked myths.
We should distinguish between meteorological droughts which are a long-term reduction in precipitation (IPCC AR6 WGI, pages 1157 & 1760) caused by persistent high atmospheric pressure that reduces cloud formation and agricultural drought which is a reduction in soil moisture and reduced flow in waterways (also called hydrological drought). Meteorological drought is a climatic phenomenon, and agricultural drought is only partly meteorological since it can be alleviated by irrigation, dams, and other man-made changes to drainage systems designed to reduce the flow of valuable fresh water into the oceans. This post is about meteorological drought and long-term drought cycles.
Deserts occur in two bands around the Earth. In the Northern Hemisphere they occur between roughly 20° and 40° North. The band of Northern Hemisphere deserts includes a small portion of the North American Southwest as shown in figure 1, modified from (Cherlet, et al., 2018). The curvature of the band is due to the northern position and curvature of the ITCZ or the Intertropical Convergence Zone.
The ITCZ is the climatological equator, which constantly moves from its southern position during the Northern Hemisphere winter to its summer position in the far north, and then back again. Since the Sun is always directly overhead at noon locally in the ITCZ, evaporation is always maximal. The evaporated water vapor, which is less dense than dry air, rises creating an updraft. When the water vapor rises high enough, it condenses into rain and clouds and the remaining dry air, which is denser than humid air, falls. This creates high-pressure zones that result in deserts.
The timing of drought conditions in the southwestern U.S. is mostly controlled by ENSO (El Niño & La Niña), and influenced by the ~22-year Hale Solar Cycle, the ~100-year Feynman Solar Cycle, the ~67-year AMO (the Atlantic Multidecadal Oscillation), and the Pacific Decadal Oscillation (Mitchell, Stockton, & Meko). Also see Jiang, et al. 2019 (Jiang, Yu, & Acharya, 2019) here. Figure 2 shows the area evaluated, and figure 3 is an assessment of climate proxies in the western continental United States.
The percent of the western United States in drought is shown in figure 3. The measure of drought used is the Palmer Drought Severity Index of PDSI (Palmer, 1965). PDSI values less than -1 are considered to be drought conditions in the figure. Wayne Palmer devised the index that bears his name. It is a measure of how unusually dry or wet an area is according to its long-term average climate. For a description of the PDSI calculation and a review see here (Heddinghaus & Sabol, 1991).
The PDSI is objective and avoids arguments over drought that are based on the effects, such as reduced flow in streams (hydrological drought), reduced soil moisture that affects crops (agricultural drought) or the effect on the economy. The economic effects of drought can be caused by meteorological drought or can be due to development (or overdevelopment) that is based on a need for more water than is available in the area.
The Palmer Index is based on precipitation and temperature records and a simple supply-and-demand model of soil moisture. It is not very effective over short periods of time, but can describe long-term (many months or years) conditions quite well. The index varies from -10 to 10, with negative numbers indicating declining soil moisture (drought) and positive numbers indicating wet conditions. Zero is neutral or “normal” as determined from historical records. One of the advantages of PDSI is it is widely used and values (either from instrumental readings or climate proxies) exist over long periods and in many areas. Figure 3 plots the area in drought, in the western U.S. since 800AD. Drought severity is not just a function of the PDSI, but also the size of the area affected and the duration of the precipitation deficit. The data used to create the index in figure 3, before the instrumental era, is mostly from tree ring proxies (Cook, Seager, Cane, & Stahle, 2007).
Currently drought conditions in the western U.S. are modest, relative to the beginning of the Little Ice Age around 1300AD. It can get much worse than the conditions that exist today. Matthew Bekker and colleagues (Bekker, DeRose, Buckley, Kjelgren, & Gill, 2014) point out in their study of tree rings around the Weber River in Utah:
“While the 20th Century instrumental period includes several extreme individual dry years, it was the century with the fewest such years of the entire [576-year] reconstruction. Extended droughts were more severe in duration, magnitude, and intensity prior to the instrumental record, including the most protracted drought of the record, which spanned 16 years from 1703 to 1718.”
The majority of dry years around the Weber River occurred in the 1400s and the 1500s, well before human emissions of CO2 and other human activities were significant. There were several periods of drought in the 20th century, and the worst period of drought was around 1934 as shown in figure 4.
Using the same criteria and color scheme, figure 5 shows the Palmer Drought Severity Index for the United States in July 2019.
In contrast to the drought years mapped in figure 4, 2019 is relatively normal to moist over most of the country as shown in figure 5. The southern New Mexico area in red and the far northwest are the only areas in severe drought.
Figure 6 shows the conditions in December 2024. The Big Bend area in south Texas is the only large area in exceptional drought (deep red) or severe drought (red), the other areas are small. Today the country is much better off than in 1934.
Drought in the southwest has always come and gone in a semi-periodic fashion, usually in cycles of about 22 years and a longer cycle of about 100 years. When the Pike expedition of 1806-1807 went through the southwest it was extremely dry and helped create the “Great American Desert” myth. Later, in the 1870s, when the early transcontinental railroads were built, it was wetter, and people called it the “Garden of the Great Plains.” (Cook, Seager, Cane, & Stahle, 2007). Historical perspective is important.
Droughts can start and end quickly, especially in drought-prone areas like the American Southwest. It is hard to characterize drought in an area with static maps. For an animation of the PDSI over the U.S. over a selected period of time, see here. Especially when interpreting news reports of droughts, this site can help keep some perspective.
Parts of the Southwest always seem to be in drought, but the dry areas move around a lot. 2009 to 2010 were good years followed by drought in 2011-2014. 2015 through 2017 were pretty good, 2018 was dry and 2019 through the middle of 2020 was good. Then in mid-2020 it turned dry and now is showing signs of getting better, except in the Big Bend area of Texas and southern New Mexico. The takeaway is that, while the American Southwest is a drought prone area, the drought comes and goes, and is much less severe than in the past, especially from 900 to 1300AD.
Download the bibliography here.
The evaporated water vapor, which is denser than dry air, rises creating an updraft.
___________________________________________________________________
H20^ GFW of 18 is lighter less dense than O2^ GFW of 32 and N2^ GFW of 28
GFW= Gram Formula Weight
Thanks for catching the typo, I fixed it.
A good illustration of cyclic drought in the Western US states is the “Dust Bowl” of the 1930s. In the late 19th century and early 20th century millions of acres of former short grass prairie (moderately dry climate plant dominance) in the western half of the Great Plains (i.e., Dakotas, Nebraska, Kansas, Eastern Montana, eastern Colorado, and Oklahoma), driven by economic development after the end of the Indian Wars and expanded migration to the US, were planted in dry land wheat. During the early years of the 20th century the rainfall throughout much of this region was sufficient for raising unirrigated (“dry”) wheat. As the grasslands were plowed under, there was next to no soil conservation measures as are practiced today, and of course western Great Plains are always quite windy. Breaking up the turf to plant wheat destabilized the topsoils, but as long as there was sufficient moisture, everything still worked great.
But then in the 1930s an extended drought took hold, and with the dry soils, high winds, and lack of soil conservation measures vast and lengthy dust storms took hold throughout the region, devastating both farms and the communities that supported the farms. The Great Depression also had an impact with bank failures and foreclosures on farms. The end result was the infamous “Dust Bowl” and a massive migration out of the Great Plains to California and Oregon where conditions and jobs were more favorable.
Today, there are dry years and wet years, but with proper soil conservation measures and crop rotations, even the dry years don’t result in another Dust Bowl. But it does illustrate how building an economy around a given set of environmental conditions can be very precarious if those conditions change.
Much the same happened to the Ancestral Puebloans (formerly called “Anasazi”) who erected large groups of massive stone buildings, irrigation ditches, and left other artifacts of their life in the 8th through 11th centuries … but when the climate changed in the 12th-13th centuries and became much dryer, those people abandoned their settlements and moved east to the Rio Grande valley to establish today’s Pueblo culture that the Spanish encountered during their colonial days. In the Rio Grande valley the Puebloans established irrigation canals that still operate today in order to feed their people and sustain their communities.
When I was in high school (many years ago) a friend and I climbed a flat top mesa in the Jemez NM (probably reservation) where the had been an Anasazi village which was fairly large. The mesa was defended on the approach routes by fairly large cutouts.
On top there was considerable evidence of numerous adobe homes. And other structures both with evidence of wood rooves probably sourced near the foot of the mesa.
There was also a dry stream bed which looked to be quite large at one time but was dry although it may have had seasonal rains.
As I understand it was probably abandoned about 1100 to 1200 AD.
The Worst Hard Time: The Untold Story of Those Who Survived the Great American Dust Bowl byTimothy Egan; 2006.
Nice summary.
Fun factoid. The Colorado Compact allocating its water among 7 states was signed in 1922, a period of relative western water abundance. There is no way that works now. Perhaps Phoenix becomes the next Chaco Canyon?
Is that because of reduced rainfall or because of the major population increases in the served areas and in the water source areas?
I believe the full name is the Colorado River Compact:
https://www.watereducation.org/aquapedia-background/colorado-river-compact
“I just wrote a model where seven parties can each receive 20% of a pie!”
“However, it is a proprietary model, so I am not going to share my code.”
— Typical global warming “scientists”
If my eyes are not mistaken, the major drought years seem to correspond with the Medieval Warm Period and the wet years seem to correspond with the Little Ice Age.
True, seemed odd to me also. But we are just talking about the western U.S.
Understood with respect to precipitation. One way to look at is weather patterns associated with warmer weather correlate with western US drought and paterns asociated with cooler weather correlate with wet conditions. One possibility is warmer weather leads to the ITCZ moving further away from the equator and pushing precip further north and vice versa for cooler weather.
It was one of thise “that’s odd” moments for me.
The major drought periods also seem to correspond with the movement of early Andean civilisations as presented by Lonnie Thompson at about 12 mins into this presentation. The ice core analysis from the Andean glaciers hints at extended (~400-450 year) periods of either mostly La Nina or El Nino conditions before the Little Ice Age.
Well, there you have it! Andean civilizations caused climate change!
I read a paper (perhaps on WUWT?) years ago, that while most of the major deserts in the world received more rain during the MWP and less rain during the LIA, some of the smaller deserts did show the opposite pattern.
In my recollection, the Sonoran Desert was one of the “opposite” deserts, and the Great Basin Desert may have been one too.
So the climate is cyclical, vaguely predictable and doing exactly what it has always done. Good to know.
The Anasazi will attest to the droughts in the southwest. Probably others too.
The Anasazi are also testament to a recent period of far more moderate climate.
Lots of people would like to live in the arid west, while expecting to have abundant cheap water like most of the east.
Very nice Andy, clear and understandable.
Thanks Bob.
I have long objected to how we talk about “drought”.
Intensity and Impacts
Not “wet”, not “moist”, not “normal”, not even “dry”. Instead we go straight to “abnormally dry”. There is an agenda built straight into the classifications.
I totally agree. The U.S. “Drought Monitor” with maps only shows one tail of a two-tail distribution. This visually creates the illusion that things are bad, really bad, and sometimes seemingly catastrophic. It makes you thirsty just to look at it. Hmm, I wonder why they set it up this way …
California has a self imposed shortage of water with the agricultural business using 80% of the water then shipping most of the crops out of state. Right now the reservoirs are at 114% of average storage and the immediate future looks good. Especially with the projected “atmospheric river”.
Very interesting. We had a camping group that used to hike the Big Bend. One year, recall early 1988, there was a botanist working on a fire-scarred area. Told us it was a good sign since it had been wet not long before. Not long later fire scars were not so evident. Also told us about all the seeds accidentally brought in, not many make it although some can be a problem.
I am very familiar with the worse than 30s, 50s Texas drought which shows on the comparable July 1934 and 1956. Knew some from Oklahoma who remembered the trees dying in the 30s. Historically there were probably worse ones in the 1800s, lets hope we go back to a millennium climate ago. The Rio Grande is a textbook case of development drought as we have crossed into Mexico at its mouth without getting wet. It now has a fossil delta, some flooded off the coast. Others may be on the way.
Yes, the 1950’s were almost as warm as the 1930’s, and almost as dry, too, as the maps show.
I think we see similar patterns today.
Compare these two temperature charts, the UAH Satellite chart and the U.S. Regional Surface Temperature chart (Hansen 1999):
Hansen 1999:
UAH Satellite chart:
The U.S. chart shows the very hot 1930’s, with the hottest year being 1934, where Hansen said it was 0.5C warmer than 1998. This would also make 1934, warmer than 2016, and of equivalent warming to the high point in 2024. So even though we are talking about warm temperatures, we are not talking about unprecedented temperatures.
Looking at the U.S. chart you see that the temperatures cooled after 1934, and cooled down though the 1940’s, but then the temperatures warmed back up, and temperatures in the 1950’s were almost as warm as in the 1930’s, and the drought index in this article shows the 1950’s were just as dry as the 1930’s.
After the temporary warming in the 1950’s, you can see that the temperatures then cooled quite a bit down through the 1970’s, at which point climate scientists were raising alarms about a possible new Ice Age coming. The cooling from the high point of 1934, to the low point of the late 1970’s, is approximately 2.0C
I think the UAH Satellite chart is showing us a similar pattern to the U.S. chart since 1998.
After the “Ice Age” cooling of the 1970’s, the temperatures started to warm again, and they warmed at the same magnitude as the warming in the 1930’s, and almost reached as high a temperature as in the 1930’s, coming within a few tenths of a degree of the 1934 high point.
Here’s the similarity I think I see: After 1998, the temperatures started to cool, just like after 1934. And the temperatures cooled until around 2012, where the temperatures started a steady climb up to the high point of 2016. So now the temperatures are back to the level of 1998, just like the temperatures in the 1950’s, were almost as warm as the 1930’s.
So it looks like to me that after the 1930’s high point the temperatures cooled a little but essentially went sideways through the 1950’s, and that’s what it looks like the temperatures of today are doing: They hit a high point in 1998, and have essentially gone sideways ever since.
Now the $64,000.00 question: What happened to the temperatures after the 1950’s warming spell?
Answer; The temperatures cooled so much climate scientists were wringing their hands over another ice age being just around the corner.
What’s going to happen to today’s temperatures after the current sideways movement? History says there will be cooling.
“the ~100-year Feynman Solar Cycle, the ~67-year AMO”
The AMO envelope varies between roughly 40 to 70 years long, the long term average has to be 54-55 years, because every other warm AMO phase is during each centennial solar minimum. The Dalton Minimum began in SC5, the Gleissberg Minimum began only seven solar cycles later from SC12 (80 years later), so the AMO envelopes were much shorter in the 1800’s. The current centennial minimum began twelve solar cycles after the start of the Gleissberg Minimum (130 years later), so the last two AMO envelopes were longer than the average, at 60 and then 70 years.
A northeast Pacific ‘warm blob’ causes drought in California, as in 2014-15. That also happened around 1876-77, millions of sheep died in the drought, then it got very wet again with the 1878 Super El Nino rains, and then huge wildfires in 1879.
Andy ==> Very well done. Think of the Anasazi — the West is now and was dry — and it has been a lot drier.
Andy,
Very nice.
It always slays me how climate science can talk about future temperature in a “global” fashion from data in models but never discuss how models explain past historical information on a local and regional basis.