Websites of interest:

“North American Drought – A Paleo Perspective”:http://www.ngdc.noaa.gov/paleo/drought/drght_history.html (National Oceanic and Atmospheric Administration)

Interview with Greg McCabe:

My name’s Greg McCabe, I’m a physical scientist with the U.S. Geological Survey at Denver, Colorado.

[57]
The work that we’ve been doing is that we’ve been looking at droughts on a decadal timescale – in other words, low frequency occurrences. So that there’s times when we get into what you might call a regime, where droughts are more common than they would be at other times. So we’re looking at things at kind of a decadal basis. And what we found when we looked at this, is that you do have the influence of the Pacific Ocean, because of the El Nino and La Nina processes, which those processes are kind of year-to-year fluctuations. But there’s also a low-frequency reflection of these El Ninos and La Ninas in the Pacific. And that has an impact on drought occurrence as well as, what we found in the Atlantic, the North Atlantic ocean has a significant impact on drought occurrence in the United States. And that was kind of the first idea that came out about a few years ago, around 2001. Dave Enfield of NOAA published a paper showing relationships between the North Atlantic and precipitation in the United States, and also with stream flow in the Mississippi River basin. And then just the past year or so we took that a little further. And what we did is that through some statistics we were able to identify the primary modes of drought variability in the United States. And it turned out that the two most common modes of drought variability were related to variance in the Pacific Ocean and variance in the North Atlantic Ocean.

[82]

Well, what I mean by the primary modes is in a statistical sense, we can break apart, if for example, you have observations of drought across a big space and through time, you can break that apart into different patterns of drought and when they occur, their strength, and so forth. And when we broke that apart we came up with a couple of patterns that could explain a lot of the variability in drought. And those primary patterns were highly correlated with not only Pacific sea surface temperatures but also N. Atlantic sea surface temperatures.

What we found is actually that the Atlantic, it wasn’t so much that it was a stronger effect than the Pacific, but in effects it was as strong as the Pacific, which surprised us, because North America is located in the westerlies, where most storm systems come from west to east, a lot of emphasis has been placed on the Pacific ocean. But, the Atlantic has this influence on the circulation in the Northern Hemisphere and it shows up that it has an influence on drought. And so that was a real interesting finding.

Well, from what we’ve looked at, the statistics tell us that when the North Atlantic is warm, there is a good chance of drought over a large area of the United States. Now whether that is primarily focused in a northern part of the country or the southern part depends on the state of the Pacific Ocean. And what we’ve seen, at least over the past few years, the drought that we’ve been experiencing looks very similar to the drought that occurred during the 1950s, and the sea surface temperatures in the Pacific and in the Atlantic are very similar to the sea surface temperatures during the 1950s. So we’re seeing sort of a 1950s type drought occurring right now. As far as what will happen in the future, that’s difficult to say. One of the things that we know from droughts, and looking at reconstructions of droughts from tree rings and other proxy data, is that droughts are not very cyclical, and they’re very difficult to predict. So, it’s hard to say how long this drought will last, but the pattern that we see now is very similar to what occurred in the 1950s.

The 1950s drought was the worst drought to occur in the southwestern U.S. in the 20th century. There were parts of the southwest that essentially received no precipitation for five to six years. There’s lots of vegetation that died off, and there were certain months during that drought where almost 100% of the area in the southwest was in severe to extreme drought. So, making that comparison to the 1950s is important because it was the most significant drought in the 20th century in the southwestern U.S. And now it looks like we’re in a similar pattern, but now in this period of time we have a greater population, more agriculture, and greater demand on the water. So there seems to be more stresses.

After the mid 1970s, there was a shift in the Pacific Ocean where we had more frequent El Ninos and La Ninas, which, when you have an El Nino, you have a wetter southwestern U.S. on average. And so after the mid-1970s we had several years of wetter El Nino conditions, so generally wetter than normal conditions in the southwest. But these types of droughts are really not surprising, in that we’ve had them in the past. And actually, if you’ll look at some reconstructions of past climate, say for the past three or four centuries for parts of the western U.S., the 20th century was much wetter than some previous centuries. And there were droughts in the past that lasted as long as thirty to fifty years. So, it’s not unusual to see a drought like this. The likelihood of one of those fifty-year droughts, right now, I couldn’t give you a probability or a forecast or anything like that. But, they have occurred in the past and I don’t know why they wouldn’t occur in the future.

One of the things that we have seemed to notice is that the longer droughts seemed to cover larger areas. And so when you have a long drought, it usually impacts a larger area. Mainly what I was looking at was the actual frequency or occurrence of droughts within a decadal timescale.

What we did was, for the 48 United States, there’s 344 climate divisions. And for each one of those climate divisions, there are time series of monthly temperature and precipitation since 1895 up to the present. And what we did is that we simply defined a drought as, we used annual precipitation data, we defined a drought year as a year with precipitation in the lowest 25% percent of all annual precipitation measurements for a climate division. So we had a different drought threshold for each climate division. And so what we could do then, for each climate division, we counted up the number of years with precipitation that was in the lowest 25% of all the values for that climate division. We counted those up for 20-year climate windows, and we moved through the time series. So in a sense we created a time series of 20-year moving drought counts, or 20-year moving drought frequencies for each climate division. And then once we had those time series we applied some statistical analysis to identify these primary modes of drought variability. And so we used a simple measure of drought and some simple methods, and it worked out to where I think we pulled out some important information.

When we started, we had no preconceived idea what would be the primary drought modes, and what climate factors would be related to it. We simply let the data tell us what was going on, and it was surprising that the Atlantic ocean came out so important, and that it was as equally important as the Pacific on these decadal timescales. Now when you go to annual timescales, year-to-year, the Atlantic seems to have very little impact whereas the Pacific has a significant impact because of the El Ninos and La Ninas that occur. But on decadal timescales, the Atlantic came out to be very important, and it’s even highly correlated with things like, if you look at the stream flow in the upper Colorado River basin, if you for example do a 10-year moving average through the time series of stream flow, there’s strong relationships between that time series and the North Atlantic sea surface temperatures. So, it’s surprising that this signal was in there.

Well, it’s pretty well accepted that the pacific is driving them. This work on the Atlantic is relatively new. Some people accept it and are excited about it. And there are others that are still questioning it. However, there’s more and more evidence that’s coming out that the Atlantic is important. There are two recent papers, one came out in Science by Zigfried Schubert and his colleagues where they ran a model to try and reproduce the 1930s dustbowl. And in their climate model, when they did this they saw that the dustbowl, part of the cause of the dustbowl was a warm North Atlantic. And then another paper was recently published that talks about the influence of Arctic sea surface temperatures on drought in the U.S. And obviously the Arctic and the North Atlantic are connected. And one of their conclusions is that a warm North Arctic was related to drought in the United States. But they were focusing on the Arctic region and not looking at the North Atlantic in that study. So, there’s more and more work that’s being done to support this idea.

Well, one of the things that has to be done that’s just starting is that the North Atlantic has to be monitored more. A lot of emphasis has been placed on the Pacific because of the El Nino and La Nina, and there’s a lot of measurements out there. But there aren’t as many in the North Atlantic. And so, right now we’re not sure what switches the North Atlantic from being warm to being cool. And until that can be determined, it won’t be very easy to do any kind of forecasting. So this is just kind of the beginning of the work that needs to be done. It’s just kind of one piece of a puzzle. And until we know what makes the Atlantic warm and then cold, we really can’t do much with forecasting.

It’s generally just, you know, a degree or two or something, I don’t have any data in front of me, but I guess for just a human to perceive the change, they’re not going to. But, the ocean is huge, so just a one or two degree change in ocean temperatures represents a lot of energy, because it’s such a large volume of water and has such a large heat capacity. And as far as why things are changing in the Atlantic, we don’t really know that yet. The Pacific is better studied and there are all kinds of ideas about why an El Nino starts or why a La Nina starts. But by having an El Nino or La Nina, they know the result, where warm water moves to the eastern tropical Pacific and so forth. But for the North Atlantic, it’s just starting to be studied.

Another thing I’ll be working on this summer is trying to look at what are the changes in the atmosphere that occur when the North Atlantic gets warm, and when it’s cooler. So we want to start looking at some atmospheric pressure patterns and see if we can figure out some of the mechanisms that behind all of these correlations that we’re finding.

I don’t think that global warming is what’s driving these oscillations, if you want to call them that, and these decadal variabilities in the North Atlantic and Pacific. Although, the signal is probably in there, because we know that there are trends in the ocean temperatures that look like trends in air temperatures. But these oscillations seem to be operating on top of that trend in temperature. But a lot of these things, for example, the AMO that you might get if you got it from Dave Enfield, he de-trends all of his data, so that what you’re looking at is simply oscillations without any long-term trending in it. And that’s the index that we used in our study, the de-trended data. So I don’t think that what we’re seeing here is a global warming signal, I think it’s something else. But the variability is probably on top of that trend in temperature.

Right now in Denver, we’re on our third or fourth year of water restriction, and I think that it’s one of the things that we need to think about, what if one of these droughts from like the late 16th century mega-drought, occurred today? How would we react, or how could we react or respond to it? And I don’t know. The demand for water has increased quite a bit. And, if the 20th century was really wetter and more benign than previous centuries, then we might be in trouble down the road. No one really knows yet, but it’s kind of a scary thought to think that we could get into a 20 or 30 year drought in the southwest. What would we do?

It’s hard to say what policy makers will do. I know that I talked to some people from the Forest Service, and they’re interested in trying to use some of this information for their planning. But, as far as water management and policy makers for people to decide how much development should occur in an area, I’m not sure how much of this they’ll use. They probably should use it, but their decisions are often driven by other factors other than water availability. So I’m not sure what the impact will be. I mean, we’ve had information on El Nino and La Nina for a number of years, and very little of that has been used for water management. And, so, stuff like what we’ve shown that is fairly new is going to take a long time for people to start using it I think.

The following person was interviewed for today’s program. Our thanks to:

Gregory J. McCabe, Ph.D.
U.S. Geological Survey
Denver Federal Center
Denver, Colorado