Guest Post by Willis Eschenbach [Updated, see end of article]
One of the claimed dangers of a few degrees warming of the Earth is increasing drought. Drought is a very difficult thing to fight, because it is hard to manufacture water. So this is a frightening possibility.
I have long claimed that “a warmer world is a wetter world”. I have said this without any actual data, based solely on the following logic.:
Increased temperature —> increased evaporation —> increased precipitation.
Today I graphed the numbers for the US precipitation. I used the USHCN state-by-state precipitation database, which also includes area-averaged values for regions of the US, and for the US itself.
First, here is the change in precipitation in the US since 1895:
Figure 1. Annual precipitation in the US. PHOTO SOURCE
Since the both the US and the globe have warmed since 1895 it seems that a warmer US is a wetter US. However, precipitation is spotty and unevenly distributed. One area can be very wet while a nearby area is dry, so what about the precipitation in each of the states?
The USHCN database contains state data. Since there are drier states and wetter states, I looked at the percentage increase in precipitation rather than the absolute change in precipitation. Here are the state-by-state results:
Figure 2. State by state changes in precipitation, 1895-2009. Values are change per century divided by average annual rainfall.
One of the things that AGW supporters have been saying would result from warming is that the desert belts would move poleward. These are the great belts that circle the earth at about 30° North and 30° South latitude. The North American belt encompasses the Southwestern US (Southern California, Arizona, New Mexico, and Texas) and Northern Mexico. If these belts were actually moving poleward as the globe warmed over the last century, we should see decreased precipitation in the Southwestern US.
Instead, all of the southwestern states have increased rainfall. The main area with decreased rainfall encompasses the Rocky Mountain states in the central Northwestern US.
My conclusions? Precipitation is indeed spotty. A warmer US is indeed a wetter US. And there is no decrease in the Southwestern US data which would show that the great northern desert belt is moving polewards. So either the desert belt is not moving poleward, or the movement is offset by the overall increase in precipitation.
[UPDATE] Some commenters have correctly pointed out that I have only shown the precipitation, which doesn’t show the change in droughts. This is because droughts are a combination of soil moisture, temperature, rain, and other factors. This is measured by the Palmer Drought Severity Index (PDSI). The PDSI index values have the following meaning:
-4.0 to less (Extreme Drought)
-3.0 to -3.9 (Severe Drought)
-2.0 to -2.9 (Moderate Drought)
-1.9 to +1.9 (Near Normal)
+2.0 to +2.9 (Unusual Moist Spell)
+3.0 to +3.9 (Very Moist Spell)
+4.0 to above (Extremely Moist)
I used the USHCN database cited above to look at the state-by-state trends per century in the PDSI. Note that these are not the average PDSI values by state, which are without exception in the range -1.9 to +1.9 (near normal). Figure 3 is a histogram of the trends per century. A “histogram” shows the number of states (left scale) that have a certain trend range (bottom scale).
Figure 3. Histogram of state trends per century of the PDSI
The trend in most of the states (39 out of 48) is toward less drought (increasing PDSI). However, most of the trends (32 of 48) are between 0 and +2.0, which is not a large change. As a result, most of the trends are not statistically significant. Figure 4 shows the significant state trends:
Figure 4. Significant trends in the PDSI in the US states.
As you can see, despite the warming of the last 115 years shown in the USHCN dataset, while some of the PDSI trends have decreased, almost all of the statistically significant changes in the PDSI are positive (less drought). And few of the changes are statistically significant.
The IPCC models say that increasing warmth will lead to increasing drought, particularly in the mid-latitudes:
In a warmer future climate, most Atmosphere-Ocean General Circulation Models project increased summer dryness and winter wetness in most parts of the northern middle and high latitudes. Summer dryness indicates a greater risk of drought.
Despite these model prediction, we have seen no such increase in drought in the US. For most of the US, there has been so statistically significant change in the PDSI index showing the number and strength of droughts in most US states. And where there has been a statistically significant change, it is in the direction of reduced drought.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Well, here in Australia, we have had recent “100 year wet events”. Lake Eyre for instance, is now flooded.
Lots of rain in New South Wales, Queesnland and Victoria. While all the time, since last year’s Victorian bush fires, predictions we’d have more dry, more dought and more bush fires.
Nah! Nature is not cooperating with the computer models.
Would have thought the – minus temp story was important enough to leave number 1 for at least a full 48 hours or related stories anyway so the press can have a look at it/digest etc… When big stories like that are quickly “demoted” or put number 2 etc.. or “lower priority” it detracts from them.. sorry my take on this one…
The AGW folks are not going to like you using their own data to prove them wrong.
I think you will find that the warmistas have already changed horses on this one when they had to explain how their prediction of a world where was snow was an increasing rarity was confounded by the reality of the winter of 2009-10. Yes, now CAGW is going to bury us under winter snow and maybe drown us with summer deluges.
Sometimes simple logic works. “Increased temperature —> increased evaporation —> increased precipitation.”
The real savings of water however is in the growth rate of crops. Cuurently it would take from ten to twenty percent more water to grow the amount of food we currently grow if C02 was still at 280 PPM!
The monetary cost of increasing our world water supply by this amount would be trillions. The political stress of such a need could lead to wars. Any warming efffect of increased C02 decrease rapidly, the benefits continue on at least a lineal scale, if not greater, to at least 1000 PPM.
The positioning of deserts has little to do with temperature, but mostly with geography: they form around 30deg N & S latitude due to Hadley Cell Circulation, and usually on the leeward side of mountain ranges.
the unstated corollary is that if in fact it is cooling over the oceans where the bulk of the evaporation takes place then we will see dryer conditions overlall.
The ARGO bouys are showing a slight cooling worldwide over the oceans, but this is an average, some areas have cooled signifigantly, so if the prevailing winds come from a cooler ocean area we might expect dryer conditions over land in the winds path.
Willis,
This is useful data to counteract overwrought “it’s worse than we thought” predictions. But what if it turns out that once WUWT’s documentation of UHI, poorly sited weather stations, weather station drop-out, and now the Mysterious Missing Em’s means it has actually cooled since 1895?
Oh, never mind.
It is curious that one of the states that had the largest decrease in rainfall (Wyoming) abuts one that had one of the largest increases (South Dakota). It makes no sense that the rain stops at the border. I wonder if the data is corrupted. Did the measuring techniques change over time (and space?)
Excellent post Willis!
There is a great many factors that effect precipation movement to be unpredictable. Evaporation and water is the regulator to keeping oceans from boiling and temperatures in check. In the last 100 years, we have changed that landscape and developed mass industrial technology that drinks up great amounts of water.
I have noticed a trend that the mountains play an important role in directing evaporation formations southward along the west coast then move northeast ward once past the mountain formations.
Winds are interesting when they are layered and hitting objects like hills or mountains.
Fascinating graph since here in NW Wisconsin in the US we hear how we are below ‘average’ in rainfall and the lakes are well below the ‘ordinary’ high water mark but the graph makes things look a little above ‘normal’. Is there something new (?) showing up here like a several year weather pattern versus climate instead of just today’s weather versus climate?
The fly in the oinment will be Australia. The expert opinion in the AR4 was that the US would generally benefit, but the mountainous western states would have les rain. Between US and Australia looks like their regional projections are a bit off. I think David Stockwell has pointed this out, as did the likelihood of errors by Dr.s Browning and Gravel.
Patrick Davis (03:38:14) :
Since 1967, the ocean salinity has changed on the surface and become more salty in the equatorial area. This would definately effect evaporation and cause massive draught as it did. Now the salinity is flipping the other way in the northern regions and is less salty.
I’m ignorant. We had some very large hailstones here a few weeks ago, up to a couple of inches in diam. Does a conventional rainfall weather station capture these comprehensively, or is it hit and miss? Do they have any effect on pan evaporimeters, like splashing?
Sometimes simple logic is overly simplistic.
Increased temperature —> increased evaporation —> increased precipitation
But the water doesn’t evaporate and then precipitate in the same place, or in a dramatically changed climate in the same places as it used to. In addition, more of that added precipitation will quickly be lost again by evaporation.
The changes that you see now are not necessarily the changes you will see as the planet continues to heat. The fact is, at the moment the amount of warming that has been seen is minimal, because the planet doesn’t warm instantly, but the CO2 in the atmosphere now has already committed us to at least 1.5C warming, where we’ve only seen 0.5C to date.
But when the planet warms more, yes, some areas will see much more precipitation, and while you may oversimplify it and say “more water is good for crops,” it can also cause flooding, and continued, previously uncommon flooding can destroy habitats. Ecosystems will change.
Meanwhile other regions (like the southwest U.S.) will see more evaporation, with that water precipitating elsewhere, like in the northeast U.S. For regions that are already under stress for water, this is very bad news.
“Joe (04:30:08) :
Since 1967, the ocean salinity has changed on the surface and become more salty in the equatorial area. This would definately effect evaporation and cause massive draught as it did. Now the salinity is flipping the other way in the northern regions and is less salty.”
So what you are suggesting is there is a “~100 year” cycle which brings about dry, ~100 years later wet, ~100 later dry, ~100 years later wet etc, conditions. And in well mixed sea water around the equater, I didn’t know the surface was saltier than the bottom.
Drought is not simply a change in rainfall. Indeed, drought is a complex attribute to define — and not surprising there are many definitions of drought depending on what sector is being considered. At the least one needs to consider drought as a function of moisture availability for an activity. Hence this means drought is a function of: total rain, average temperature, dry spell duration, rainfall intensity, maximum temperatures, hot spell versus cold spell duration, seasonality, inter-annual variability, etc. One can actually have drought with no change or even counter-intuitive change in rainfall totals! Hypothetically, for example, imagine a shorter rain season of more intense rain events with longer dry spell durations and higher maximum temperatures. The response of, say, agriculture which has significant sensitivity to crossing environmental thresholds and dependencies on timing, would be marked, yet the rainfall totals may actually not change much from climatological baselines. i.e. it’s HOW the rain falls in conjunction other relevant parameters (such as Tmin and Tmax etc) that is important if you want to assess impacts. As usual, the impact of climate change can’t simply be swept under the rug with simple generalized statements on rainfall totals. (Likewise, nearly all of climate change needs far more nuanced assessment that can ever be properly addressed in a blog).
Living in a state that’s surrounded by water, Florida, I know for a fact that if you have a lingering HIGH PRESSURE WEATHER SYSTEM, no clouds form and you get no rain. It has nothing to do with climate.
The key to the exercise is to realize that climate and it’s variation are far more mysterious (in the sense of not being easy to model) than we thought. We must stop making multi-trillion $ policy decisions based upon models that are tweaked according to a political (economic) agenda and that can’t predict next month let alone the next century.
Great post Willis.
I think an unbiased examination of drought cycles would find correlations to the ENSO, PDO and even the ~1500-yr (Bond/Dansgaard-Oeschger) cycles.
Lake sediments indicate that drought- and wet-prone conditions of West and East Africa alternate with the 1500-yr cycle. During the Medieval Warm Period (~800 AD to ~ 1300 AD) and the Modern Warming (~1880 to present), East Africa has been more drought-prone than West Africa… During the intervening Little Ice Age (~1300 AD to ~1880 AD), West Africa was more drought-prone than East Africa. (Verschuren, 2001 and Russel et al., 2007)
Drought conditions of the American Southwest might also have a similar “mega-trend.” On a more local scale, the conditions correlate fairly well with the PDO. The “poster child” of global warming droughts, Lake Powell, has actually been gaining water volume since 2004. The entire Colorado River lake system has grown by 7% since 2004. Only Lake Mead has shown a decline in water volume over that time period…
As already pointed out by David Becker, Ph.D. (04:18:08):
Adjoining WY and SD are at opposite ends of the scale.
And not only that:
”High end” SD is surrounded by states that are all at the lower end of the scale.
Even though there can indeed be significant divergences from expected standard deviations in the real world, since this data goes back at the way to 1895 it seems more than a little odd that SD should be such a significant ”local outlier”.
Wow awesome 30 second disprove of what thousands of actual scientists are researching! Lets close the CSIRO her in Australia and just pay Mr Eschenbach a cool million for 5 minutes work a week and he can have the rest of the time off.
Hmm I just realised that comment may not strike many of you as sarcastic. It is.
Interesting.
Typo correction: “The MAIN area with decreased rainfall encompasses the Rocky Mountain states in the central Northwestern US.”
Does the precipitation trend have a unit root?
🙂
Re fredb (04:51:05) :
If your arms wave more perhaps you can fly. What observational point are you making?