Earth & Sky

Websites of interest:

Research reveals use of tree rings and ocean temperature shifts in anticipating mega-droughts (EurekaAlert.org, February 14, 2003)

International Tree Ring Data Bank (National Oceanic and Atmospheric Administration)

Interview with Julio Betancourt:

My name is Julio Betancourt, I’m a research scientist with the U.S. Geological Survey’s Water Resources Division. And I’m based at the Desert Laboratory, which is a University of Arizona institution on Tucson’s west side. And so, along with some of the other hats that I wear, I’m also and adjunct professor in the department of Geosciences in the University of Arizona.

It’s more like the 1950s. If you look at something called the drought area index, which is the percent of area across the U.S. that experiences a certain level of drought. And that level is defined by the Palmer drought severity index, which sort of looks at temperature and precipitation back over the previous six months. The Palmer drought severity index of -3 is pretty severe drought. And if you look at area that experiences that level of drought over the last hundred years, the 1930s and the 1950s stand out as the two big vents of the century. And those are the kinds of droughts that I refer to as mega-droughts. And the term mega-drought, you’ll hear it a lot, but it’s not a well-defined term. It basically refers to these long duration droughts that affect large parts of the country. I quantify it a little bit by saying that it’s any drought that equals or exceeds the statistics of the 1930s or the 1950s droughts. And these are droughts that are experiencing severe drought over 30% of the country, in a sustained fashion over several years.

You know what we’re dealing with in places like Arizona, but throughout the west, are longer and hotter summers because of increasing temperatures. And on top of that, you have natural precipitation variability – you have these periodicities in high or low precipitation. And so it’s not a coin toss, it’s not a fifty-fifty chance that a dry year is going to be followed by a wet year and vice-versa. Wet years tend to be clustered and dry years tend to be clustered. And it’s sort of that decadal, or multi-decadal scale variability that’s worrisome. And it’s worrisome mainly, put into the longer-term context, instead of just talking about the last 100 years, if you talk about the last 1000 years, the 20th century may not be representative of the natural variability of the climate system. So, we have some blind spots. And now, of course, the issue is that since the summers are longer and hotter, when you do get a drought, the fire season can be longer. The plants use up the soil water quicker, and so you can get a lot of mortality of trees, for example. In the last two or three years, we’ve had millions of acres of pine forests and pine woodlands impacted by bark beetle kills. For those of us that know and drive the west a lot, it’s a pretty obvious happening on the landscape, that you have huge numbers of trees that are conifers, they’re evergreens, with completely dead canopies. And so this fire season coming up is very worrisome, and there are already indications that it’s going to be a bad fire season. There were fires over by Fort Collins in northern Colorado, that were fairly big fires that just happened in the beginning of April. And that’s pretty early for the beginning of their fire season. Their fire season usually is later. So we’re seeing unusual things, like the moving of the fire season towards an earlier onset.

Well, it’s not just climatic, it’s also the fact that fuels have been built up over the last century for a number of reasons. One is that grazing, over the last 120 years or so, has reduced the fine fuels that support low intensity episodic surface fires that kill the young trees and leave the mature ones unscathed. And now what’s happening is that you’re getting a lot of young trees that are sort of coming in through the lower part of the canopy and laddering up the fuels. So the fire season is becoming much more spectacular, not just because of climate variability, but also because of the fact that these woodlands, that haven’t been suffering these low intensity fires, are now beginning to suffer stand-replacing, catastrophic fires during these major droughts.

Imagine that if the way that you characterize drought is basically to quantify the area that’s impacted by drought in any given year. And you plot that out over time. And you look at times when large areas of large areas of the United States were experiencing drought, and then times when not very much area was actually experiencing drought in the U.S. For example, in the last century, you have three prolonged wet spells, 1905 to about 1930-1940s, and then 1976 to 1995. As a preview to things that we will come to in the conversation, that period from 1976 to 1995 is actually one of the wettest periods in the last thousand years in the tree ring record of the western U.S. So imagine these three wet spells, and then punctuated by three dry spells, which are the 1930s, the 1950s, and then this drought, that really began during the La Nina even of 1996 and then persisted into 2004. And one of the questions that we have, when you see large areas experiencing either dry spells or wet spells over the U.S., is there are a lot of very different climatologies that produce precipitation over the U.S. So when you get large areas that are either experiencing wet or dry, normally that has to involve something fairly big. And that thing is actually changes in the low frequency behavior of the oceans, and in particular the Pacific and Atlantic oceans. And so underlying this curve of area experiencing drought over the 20th century, is the slow behavior of the oceans, the decadal and multi-decadal scale behavior of the oceans. The oceans have a lot of memory. And there’s a lot of inter-annual variation in the oceans, but actually the ocean’s characterized by much more low-frequency behavior like decadal to multi-decadal behavior. It just takes a long time for these large bodies of water. The tropical pacific, for example, covers 1/3 of the circumference of the Earth. So, it’s a big heat capacitor. And it stores up heat and then releases it. When something happens in the ocean, it’s going to be felt in the land. So the clues to what cause this variability in prolonged wet and dry spells over the continent, the key is actually the oceans and their low frequency behavior. The mechanisms though are not really well understood. So our research group as well as colleagues working with us are studying how the ocean might have impacted precipitation variability over the U.S. over the century. And then there’s another group that’s actually working on how the oceans might have also affected precipitation variability as reconstructed from tree rings. And what we’ve found thus far is actually fairly interesting. The focus over the last few decades in terms of climate for western North America in particular has been squarely on the Pacific ocean, particularly related to things like El Ni?o and La Ni?a. And now people have come to recognize that the Pacific, particularly the North Pacific, has decadal scale variations that are dubbed the Pacific Decadal Oscillations. This is an index of variations in the North Pacific ocean that are now sort of linked with long-term precipitation anomalies in western North America. So the research until recently had sort of progressed to focus on the impact on both interannual and interdecadal variability in the Pacific Ocean. But in 1999 and 2000, Dave Enfield and some of his colleagues with NOAA in Miami, started to feature the low frequency variability of the North Atlantic ocean, and how it might affect precipitation over the United States. And if you think about that, for the most part the U.S. is upstream of the North Atlantic. So how could the North Atlantic affect precipitation over the U.S.? There’s some difficult things to get into, but statistically it’s pretty clear that the North Atlantic also has an influence on what I’ll term drought frequency in the U.S. In particular, these three periods of drought in the 1930s, 1950s, and then the drought of 1999-2004, are all linked with very noticeable and prolonged warming in the North Atlantic Ocean. And the statistical relationships are actually fairly strong. So now we know that it’s not just the Pacific ocean that influences long-term precipitation variations over western North America and actually the whole U.S., but the North Atlantic also plays a role in both summer and winter. And so we’ve done this work in the instrumental record, by instrumental record I mean that we look at precipitation from weather records collected over the last hundred years, and then compare them to sea surface temperature measurements that have been made over the last hundred years as well. And the relationship is actually fairly convincing. And you could say, well that’s true for the last hundred years, but if you had records that span over the last millennium at annual resolution, then maybe they wouldn’t show the same thing. In other words these relationships are not stable over time. And one of the things that we’ve been doing is using tree ring data, some of it collected by us and some of it collected by others that are now published by websites where you can download the data and analyze it. Now in those tree ring data, we find that the same relationships that we identified for the instrumental period for the 20th and 21st century actually exist over the long-term, over the last 1000 years. And in fact the pattern of drought is as follows. The western U.S., in particular the area that was affected by the 1950s drought, meaning the central and the southern Rockies and the adjacent areas of the southwestern U.S. and the Plains – that area tends to experience drought when the North Atlantic is warm, and when the tropical Pacific is cold. Whereas, most of the U.S. tends to experience drought conditions when the North Atlantic is warm, regardless of what the Pacific is doing. So it almost looks like the North Atlantic warming controls how much area in the U.S. is actually experiencing drought. While the Pacific may control sort of the timing of the drought. And there have been some really interesting developments out of the research that we’ve been doing, it’s actually fairly exciting. In March 2003, we published a paper using 750 years of tree ring chronologies from the central and southern Rocky Mountains, basically covering parts of Wyoming, Montana, Colorado, Utah, New Mexico, and Arizona. And, one of the interesting things that we found is that at these longer timescales, like decadal timescales, there’s quite a bit of coherence in these wet and dry periods in those 750-year chronologies. When there’s not tends to be a time when the North Atlantic is neither warm nor cold and just kind of average. Probably the most interesting time of this coherence is what everybody refers to as the Late 16th Century Mega-drought. And it was a drought that in the U.S. occurred from 1575-1595. It was from coast to coast and from Mexico all the way to Canada. It was actually the drought that was responsible for troubles in the Plymouth colony, and it was also the drought that was responsible for the drying of Mono Lake in California. And, it’s a very interesting drought because it was so widespread that we figured that in fact the North Atlantic has to control the extant of these droughts that sometimes last one or two decades – that the North Atlantic would have had to have been involved. And we didn’t know how until recently. We had a paper accepted in Geophysical research letter where we reconstructed North Atlantic sea surface temperatures from tree ring chronologies that span the whole North Atlantic basin – so tree ring chronologies from Europe, the Middle East, and the United States. And when you look at the reconstruction of North Atlantic sea surface temperatures from these tree ring data they show in fact this very abrupt warming in the North Atlantic at the same time as the Late 16th Century Mega-drought. So there’s some satisfaction in the fact that we’ve been able to retrodict, sort of predict backwards in time what the North Atlantic might have been doing in time based on the sort of the geography of large scale droughts in the United States that happened in the preinstrumental period as recorded by tree rings. And of course now what happens is that we can do the same kinds of statistical analysis that we’ve been doing on the instrumental record in the 20th and 21st century, we can actually do it over the last millennium as well, because we can actually reconstruct not just North Atlantic sea surface temperatures, but also tropical Pacific sea surface temperatures.

JB: Tree rings, depending on the area and the species and the response of the tree growth itself to climate can record not just precipitation variations but also temperature. And temperature on land, particularly in places like the eastern seaboard of the U.S. or in Europe are actually reflecting the air temperature. And the air temperature is very much affected by sea surface temperatures in the North Atlantic. So when you hear people talk about, for example the Medieval warm period in Europe between 900 and 1300. Or you hear people talking about the little ice age from 1400 to 1850. Those were periods of time when Europe was very warm. And tree rings would show that. Those were also times when the North Atlantic would have also been very warm. So the North Atlantic sea surface temperature in large part controls the air temperatures over land in places like Europe and the Middle East and the eastern seaboard. So we use tree rings from key areas where we know that the temperatures vary according to what’s happening in the North Atlantic Ocean. We use tree rings in those areas in order to reconstruct how the North Atlantic sea surface temperatures would vary over a longer period of time for what we have temperatures for.

JB: There’s lots of different ways of sampling tree rings from trees. One, the most common one is that you actually use what’s called a Swedish bore – you bore a tree and extract a core from a tree that goes from the outside to the center of the tree. And the pattern of narrow and wide rings within that core, or within the cross section – you can also cross section the tree. The tree will survive if you core it. It won’t survive of course if you cut it in half with a chainsaw. So, the pattern of narrow and wide rings is actually a record, depending on the species and it’s sensitivity to climate, it can record variations in temperature and precipitation. And it does so in a very faithful fashion. And in fact, in many cases there’s a closer relationship between let’s say the last hundred years of weather station records for precipitation and the tree ring chronology from twenty kilometer away than there would be between two weather stations that are only 40 km apart. And tree rings tend to record not just the year to year variability but also the decade to decade variability in precipitation and temperature. So they’re very faithful recorders of climate.

JB: Well, sure. Take the North Atlantic for example. The general public is aware of something called the Gulf Stream. And this is a surface current in the North Atlantic Ocean that carries warm water from a tropical latitude, carries warm water north into the North Atlantic towards Europe. And, the reason why Europe, by the way, places like England, have relatively warm climates for their latitude is the gulf stream – the transport of warm water from the tropic to high latitudes. What the public probably doesn’t know very well is that the Gulf Stream doesn’t just end there. As the Gulf Stream is transporting these tropical waters, the waters get colder, and they also evaporate and get saltier. And as they get colder and saltier, they get denser. And they sink at depths in what’s called the Norwegian Seas. So they go to the North Atlantic and then there’s a big kind of sill in the ocean between Greenland and Scotland. Those waters from the Gulf Stream cross over into what is called the Norwegian Seas, and then they sink at depths. And then there’s something called deep-water formation, there’s a cold current that returns back towards the tropics at depths. So there’s this big convey belt of water in the North Atlantic. And the movement of those waters takes a long time. This conveyer belt is actually linked with something called the thermal-haline circulation that covers a large part of the world’s ocean. And for a packet of water to make it all the way around takes a thousand years. So when I say that the climate system has a memory, I mean just that. It takes more than a few months and more than a year for these masses of water to change. So, if you take North Atlantic sea surface temperatures, they’re really an expression of what’s happening at the surface, with all of these currents that have different temperatures and different salinities, and therefore different densities. So the variations in sea surface temperatures implicate something that’s happening at depths, and something that has a low frequency behavior that exceeds much more than a year. So these are intrinsic variations in the oceans that have a persistence of more than a year. So, for example when the North Atlantic warmed in the middle part of the century, associated with the 1930s and the 1950s drought, it was warm basically from about 1925 to about 1965. And then following that, the North Atlantic was actually cold from the 1960s all the way to 1995. And in 1995 the North Atlantic started to warm again. In fact there’s an interesting anecdote from a meeting that we were attending in 1998. It was a meeting focused on the 1950s drought where there were a bunch of climatoglogists, and we were all sitting around talking about the 1950s drought. And I remember that people were very interested in the fact that in 1995 the North Atlantic had turned warm. And this was in May or June of 1998. And we were all wondering what would happen if the tropical Pacific actually cooled. And we thought that if that happened, we’d end up with something like the 1950s drought. And in July of 1998,, the tropical Pacific started to cool, and it stayed that way until July of 2002. And of course that was the driest year in this particular drought. So a pattern very much like the 1950s emerged after the warming in 1995. And that’s actually one of the concerns that we’ve had, is that it looks like the oceans are reconfiguring themselves to look more like what the oceans looked like during the 1950s drought, and possibly the 1930s drought.

Our experience with the instrumental record of the 20th century, as well as the tree ring record of the last millennium, is that there is in fact an association between North Atlantic warming and drought over the United States. And that association is fairly strong. And what worries us is that the North Atlantic Sea surface temperature anomalies tend to be very persistent. So once you’re in one of these warming patterns, you’re going to be there for a while – maybe as much as 20 to 30 years.

I think that the key question here is how do complimentary modes of the Pacific Ocean and the Atlantic Ocean govern the timing, duration, and the extant of drought over the United States. And I think that a climatologist would be very curious to hear what the mechanisms would be for the North Atlantic to influence U.S. precipitation. I think they can understand a little bit more how the Pacific, the Pacific is basically upstream of the western U.S. So I don’t think they’re surprised to know that the Pacific Ocean variability would affect variability of precipitation in the western U.S. But the other way around from the Atlantic is a little bit trickier question. And by way of answering, I go back to a study that again was done by Dave Enfield of NOAA where he suggested that two things might be happening. One is that the North Atlantic actually does control the abundance and variability of summertime precipitation over the U.S. Because basically in summertime places like the southeastern United States in the Midwest are downstream of the Atlantic. In other words most of our summertime moisture comes from the Atlantic and the Gulf of Mexico in summertime. And the mechanism there is much easier to explain. Because if the tropical Atlantic and the sub-tropical Atlantic are warm, one of the things that happens is that the sub-tropical jet that’s responsible for bringing precipitation in summer across the southeastern United States and into the Midwest. When you have North Atlantic warming, the subtropical jet is displaced to the south, and it’s aimed at Mexico, instead of at the Midwest. And when it’s cold, it’s fairly steady and it’s displaced northward and wetting the southeastern U.S. in the Midwest. So it’s not hard to imagine how North Atlantic warming and cooling could affect summertime moisture over the continent. It’s much more difficult to explain how it might affect wintertime moisture. And, most of the wintertime moisture over North America actually is embedded in the upper westerlies or the Polar jet stream. And the polar jet stream has a lot of sinuosity, it goes from west to east. The more direct it goes from west to East, the drier western North America becomes, particularly further south of western North America like the southwest. But when the jetstream is fairly sinuous, it’s fairly wavy, you have a lot of capability to integrate temperate and tropical weather systems. So there are these two patterns. You can call the more direct west to east one the zonal pattern and the one that has a more north-south component meridianal. And, one possibility might be that the North Atlantic, and in particular the Arctic ocean, which has some links to the North Atlantic, controls the position and sinuosity of the winter polar jet stream which is basically the winter storm track over North America. And, in fact there was a paper published in Geophysical Research Letters by a couple of colleagues from a California university that’s modeling study that shows when you reduce the sea ice extant over the Arctic Ocean, the tendency is to make the westerlies in the winter storm track more directly west to east. And, this is kind of complicated to explain but what it does is it produces very very high pressure over the West Coast of the U.S. And in fact this is exactly what we see for the period between 1930 and 1960. And that is this relatively big high pressure dome, kind of stationary dome parked over the West Coast of the U.S. with much lower pressure over the Caribbean. And, this is kind of a characteristic that we’ve all seen in association with the Atlantic Multidecadal Oscillation, or this variation in North Atlantic sea surface temperature that happens over the long term.

When the Atlantic temperatures are very warm, the jet stream tends to be displaced to the north. When it’s cold, it tends to be displaced to the south. And so what you would expect in wintertime is, when the North Atlantic is warm, you would expect drier conditions over the western U.S. in winter. And when it’s cold, you would expect wetter conditions in winter. So in sort of rapping that up, it may not just be the influence of the North Atlantic on one thing or the other, it may be the influence on a couple of different things. And it may be that the North Atlantic sea surface temperatures are a slave to something else that’s happening even at higher latitudes, like at over the Arctic, at polar latitudes. And the tendency in fact, when you go through and look at Arctic sea surface temperatures, they look very much like North Atlantic sea surface temperatures, except they tend to lead, Arctic sea surface temperatures tend to lead North Atlantic sea surface temperatures by a few years. But the same decadal to multidecadal scale patterns appear in both.

JB: Again, let’s say from 1900 to now, we basically use weather station records all over the United States. And those records are compiled in different ways. One way to compile them is in what are called climate divisions. There are 344 climate divisions over the U.S. One of the things that we looked at was the lowest 25% of the precipitation distribution at each of those 344 climate divisions over the last century, and that we defined as drought. Whenever the precipitation was in the lowest 25% of the distribution in the last century we called that drought. And then we applied a statistical method called principal components analysis that identified the leading modes of the variability in time and space. So basically we mapped the behavior of drought frequencies over the U.S. over these 344 climate divisions. We mapped it in space for the U.S., and then we mapped it in time. And then we compared them to the two low frequency indices of the Atlantic and the Pacific, the Atlantic Multidecadal Oscillation and the Pacific Multidecadal Oscillation, and the patterns were exactly the same in time and space. So it’s pretty evident that the Atlantic Multidecadal Oscillation and the Pacific Decadal Oscillation represent the leading modes of drought frequency variability over the century. And then we’ve gone back and actually done a very similar analysis using the tree ring record. There are tree ring records all over the United States that have been collected basically almost since the turn of the century. And these have now been compiled and they’re available in an atlas that’s on a website. So you can actually map out the occurance of doubt with high precision. In fact, if you compare the tree ring and the instrumental portrayals of which areas are experiencing drought in any given year, they’re almost identical. The tree ring record is doing a very adequate job of characterizing the geography of drought in any given year that you see in the instrumental record. So, we actually have done these studies in several steps. And in one study what we ended up doing was looking at a number of chronologies over this area that stretches from Yellowstone all the way down to New Mexico and Arizona. At that point we realized that there was a lot of coherence in decadal scale precipitation variability across the area. And we started to ask what actually produces that coherence. i mean these are all sites that have very different peaks in precipitation throughout the year. In one area the peak in precipitation may be in July and August. And the climatology, what produces that precipitation, is unique. And then in other areas that precipitation peak is in May and June, and again the climatology that produces that precipitation is fairly unique. And then in other areas the peak in precipitation happens in winter, or in early spring, and they have very unique mechanisms for producing that precipitation. So it’s hard to imagine what was causing the coherence at these longer timescales, these decadal timescales, of sites with very different seasonalities of precipitation and mechanisms of producing precipitation.

JB: By coherence I mean synchrony. That’s right, the orchestra is all playing together. So imagine if the trombone was playing a tune that was a little bit off from the violinist. The racket that that would make. So what happens is that the orchestra is sometimes playing in unison, and in phase. And sometimes it’s not. And one of the questions that we had was what produces the synchrony of drought across these large areas and of wet periods across these large areas, as opposed to the asynchrony when they’re not in phase. And we’re now working on some new analysis that suggests that in fact the North Atlantic may be in large part controlling whether all these areas with different precipitation seasonalities and mechanisms for precipitation are in sync or they’re out of sync. And in fact you can say it a lot simpler than that. The North Atlantic Ocean warming and cooling are controlling the extent of drought or wet conditions across U.S., primarily because when the North Atlantic is warm, large areas of the U.S. are experiencing drought, and when it’s cold, large areas of the U.S. are actually wet.

JB: It’s kind of mixed. If you can just think back about 25 years ago, when people were talking about the El Ni?o Southern Oscillation phenomenon. At that point, it was basically a statistical mode, and people knew that there were these relationships between what happened in the tropical Pacific and precipitation remote regions like the U.S. And people knew that there was a statistical relation, but they didn’t know the extent to which all of this represented a physical mode that could be described in a physical model rather than just a statistical model. And now of course, we’ve gone way beyond that. And now for example we have physical models of the tropical Pacific in the El Ni?o Southern Oscillation Phenomenon that are very robust and can predict El Ni?o and La Ni?a as much as 6-9 months in advance. And in fact recently there’s been some evidence that the models are having some success even on the two-year timeframe. And this is remarkable. Now you have physical models that will actually tell you whether an El Ni?o or La Ni?a is developing, and you may be able to see it in May of one year, and the impact on the wintertime precipitation in the western U.S., for example, is in December-January-February-March. And then imagine what you’re trying to predict is not just the precipitation, but the severity of the fire season that comes in the ensuing spring and summer. And so, this recognition of El Ni?o as a physical mode that can be physically modeled is allowing us to anticipate the severity of the fire season in the southwestern U.S. as much as a year in advance. Well, the status of the low frequency, the decadal scale variability, these indices that I’ve been talking about, the Pacific Decadal Oscillation, and the Atlantic Multidecadal Oscillation – we’re kind of at the same place that we were 25 years ago, in the sense that those are now statistical modes that are recognized, but people don’t know exactly what to think of them yet because – there’s been some attempts to model, but they’re not yet bombproof physically recognized modes of climate variability that are physically modeled and everybody knows exactly which hairs are out of place. So we’re kind of in the statistical phase of development. But imagine what the impact of actually being able to physically model the El Ni?o Southern Oscillation Phenomenon and the predictive capability that gives you, where you can actually say whether or not it’s going to be dry in the southwestern U.S. 6-9 months or even a year in advance. Imagine that the same kind of capability could in fact develop from these low frequency behaviors in the ocean, these decadal scale behaviors in the ocean. So for example, once you started to see some precursor conditions in the ocean, in the North Atlantic for example, that you could tell, like we did sort of informally in 1998 when we were sitting around in a meeting and we were wondering, you know, the North Atlantic has warmed, what’s going to happen when the Pacific cools? Imagine if you had a fairly good lead time of that, then maybe what you could do is anticipate whether you were going to be in a prolonged wet period or a prolonged dry period across large areas of the United States. So, there’s a lot at stake in these indices of ocean variability becoming recognized and modeled as physical modes of ocean variability. Because they may actually provide a lot of long term predictability of climate. I think that these climate studies are progressing very, very quickly and in a very sophisticated way. From one year to the next, there are incredible advances. Unfortunately what that means is that the practitioners, the people who manage water and other resources are always about a decade behind. And so I think that, for example, the El Ni?o Southern Oscillation Phenomenon, even today is still not being used to develop fire readiness in the southwest or to anticipate that you’re going to have a relatively wet year in which you can accomplish a lot of prescribed burning to reduce fuels in the urban/wildland interface. Those kinds of use of this long term climate predictability outside of the five day forecast, those kinds of products take a long time before they’re being used by practitioners and by water managers and resource managers and economists and government agencies and the like. And so I think it would behoove the public, and in particular the practitioners, it would behoove them to pay attention to these climate advances. Because there may be a lot in store in terms of long-term predictability of climate.

JB: I think that most of us who’ve been doing at least these statistical studies are very concerned about how long the North Atlantic might stay warm. We’ve worked on a reconstruction of basin-wide North Atlantic sea surface temperatures. By the Atlantic Multidecadal Oscillation, what that refers to is basically a 10 year running mean of sea surface temperatures, temperatures in the upper meter of sea water, between 0 – averaged over the whole North Atlantic – and 70 degrees. And the tendency is for that index to actually stay warm for 20-30 years, and then for droughts to be embedded in that period, depending on exactly what it is that the Pacific is doing. So our concern is that the current drought, which lasted 5-7 years depending on where you are, could in fact continue. And, one of the problems in society, particularly relative to climate variability, is that you tend to get into these dry periods, and then it rains. And then the resolve that you built up during the dry period sort of goes out the window. In fact a friend of mine at the Nebraska Drought Mitigation Center, Don Wilhite, refers to this as the hydro-illogical cycle. It’s the tendency, for example, during a La Ni?a year, when it’s really dry, like 1996, to start a bunch of drought plans across the afflicted states. And then during the 1997-98 El Ni?o when it was wet, drop those plans, and then have to restart them again once the drought resumed. And so there’s a tendency to rely too much on what happens this year, and not take into account the fact that dry years tend to be clustered together in dry decades, and wet years tend to be clustered together in wet decades. And you may have a dry year or two in a relatively wet period, just like you can have a wet year or two in a relatively dry period. And, I think that right now it would behoove us to sort of maintain our resolve. It is almost always prudent to be conservative about water resources, and given the fact that many of these areas in the west are growing very, very fast in areas that are relatively dry, and don’t have access to plentiful resources, our sensitivity to drought is actually heightened. And so it’s usually prudent to be concerned about water resources, particularly given that these growing populations in the west, but given the North Atlantic warming that started to occur in 1995, more than now it behooves us to be prudent about water resources.

JB: Well, the North Atlantic, true to form is pretty much staying where it is, which is anomolously warm. The Pacific is a little bit more fickle though. And so what happened is that from 1998 to 2002 the Pacific was relatively cold, and despite the fact that it went warm again, the tropical Pacific went warm again, in 1998-1999, the drought continued. And the Pacific now is kind of in neutral mode. The Pacific is a little bit more fickle than the North Atlantic. So the tendency is for the North Atlantic warming to continue. It’s not going to reverse itself overnight. Whereas the pacific is capable of having substantial year to year variability, year to year changes.

JB: I think that the area of drought is going to continue to be fairly large, and maybe the locus of wet and dry areas is going to shift, in particular between the Pacific Northwest and the Southwest. The Pacific Northwest and the southwest tend to see opposite things. When the tropical Pacific is more El Ni?o like, or warm, the Pacific Northwest is dry and the southwest is wet. And during La Ni?a, when the tropical Pacific is cold, the Pacific Northwest is wet and the southwest is dry. So I think we’ll see shifts in the areas that are wet and dry, particularly along the West Coast. But I suspect that given the fact that the North Atlantic will likely remain warm, that much of the U.S. is going to experience dry conditions over the next few years until the North Atlantic starts to shift back into cold mode.

JB: Well, these changes that I’ve been talking about relative to the ocean are all natural. So, to this point we’re just talking about the natural variability of the ocean. And, how they relate to global warming is kind of unclear, except that the way that you might want to think about it is that the climate is going to continue to vary naturally, on the decadal to multidecadal timescales. The thing that’s different now is that you have longer and hotter growing seasons – longer and hotter summers, the temperatures are elevated because of the greenhouse effect. And we could argue about that, and in fact it’s an issue that’s been heavily politicized in recent years. But there’s very little doubt that the temperatures have stared to increase. Now what’s interesting is that usually when the U.S. is experiencing wet conditions, it’s also relatively cool. And when it’s experiencing dry conditions it’s relatively hot. There are a variety of physical reasons for this that I won’t go into, but trust me that that’s generally the case. Wet is usually associated with cool and dry with hot. What’s interesting is that we have such a very wet period from 1976 to 1995, particularly in the western U.S., and yet it was a very hot period. So now what’s happening is that increasing temperatures are actually trumping the effect that precipitation has on temperature. So, from a standpoint of what’s going on in the landscape, this is very, very interesting, because generally trees, for example, like to establish themselves as seedlings during warm, wet springs. And warm wet spring doesn’t happen very often because if a spring is wet, it’s also going to be cool. And so what’s happening now is that it’s happening more often than normal that it’s warm and wet, and it’s ideal for trees to become established. On the other hand, what’s happening now is that when you have a drought, it’s not just hot, but it’s hotter than it normally would be during droughts. So now it’s not only dry, it’s unusually hot. So there’s two things happening on the western landscape today. One is that you’ve got huge surges in the recruitment of new trees. At the same time you have huge amounts of mortality in the mature trees. So you have lots of seedlings and lots of dying mature trees at the same time. There are other things that are happening that are of interest. A large part of the water supply in the western U.S. actually comes from snow pack. And, because of the hotter temperatures, one of the things that’s happening is that less of the rainfall above certain elevations is actually occurring as snow and more of it is occurring as rain. And so the snowpack is becoming less important to the hydrology of these watersheds. And the snowpack is absolutely critical because, having snow persist on these western watersheds, the longer this snow pack persists, the more advantageous it is for the water supply. So one of the things that’s happening now with the hotter and longer growing seasons is that more of the precipitation is occurring as snow. The other thing that’s happening is that you’re melting the snow sooner. So, for example, for streamflow, snowmelt discharge that defines a lot of these western streams is actually occurring earlier in the season, and that’s not good for the water supply either. Now in terms of globally, you might want to think about the impact of the North Atlantic on Europe, for example. When the north Atlantic is naturally warm, you’re adding to the warming from the greenhouse affect. So you end up with anomolously high temperatures in the summertime in Europe. And I think that if you remember back from last year there were basically tens of thousands of human deaths because of a heat wave in Europe because of basically a combination of the natural North Atlantic warming and then the greenhouse induced warming.

JB: It potentially has a lot of impact, and it’s something that we really have to take into account. For example, one of the things that the North Atlantic might do is either mute or amplify the greenhouse induced global warming. So, during cool North Atlantic patterns, maybe it doesn’t show up as much under the North Atlantic warming, when all of a sudden you see really sky-high temperatures. But what’s interesting is that U.S. precipitation and U.S. temperature usually goes hand in hand. When it’s cool it’s because it’s wet, and when it’s hot it’s because it’s dry. And now, what’s happening is that that’s out the window. You’re seeing periods of warm-wet followed by periods of very, very hot, dry temperatures.

JB: Probably the most important thing is that in the United States, we really have not done a very good job of taking into account this decadal scale variability that’s evident not only in the instrumental record but also in the tree ring record. We’ve got to start asking the question, over and over again of “what if?” What if we saw the return of droughts like the late 16th century mega-drought? What if we saw something like the great drought of the 13th century between 1266 and 1295, that caused the abandonment by the prehistoric Anasazi Indians of the whole Four Corners area? What happens if droughts like that actually result in the complete drying of something like Lake Powell? If a drought like that happened today, what would happen to the interstate agreements, basin agreements like the Colorado River compact? Basin agreements that are international, that involve not just the United States, but Mexico? We’ve had a great experiment happen in the U.S., in just the last few years. This drought between 1999 and 2004 is a pretty nasty drought. (The year) 2002 is an unusually dry year. When you see years like 2002 in the tree ring record, they don’t occur in isolation. They don’t occur in the middle of a wet period – in other words they usually occur in the middle of a long, dry period. And so I think that that should serve as warning that perhaps we’re into something that’s going to last for a few more years. And I think that we have to use the tree ring record in particular about drought, and start asking a few more questions about how are these socio-economic systems that are linked to water in the west, how resilient are they? How can they withstand droughts that exceed droughts in the 1930s and the 1950s? It’s pretty clear that those kinds of mega-droughts have occurred in the historical record. We don’t really understand well the climatic context of those prehistoric droughts. So it’s very difficult to rule them out just because they occurred 500 years ago. And there’s a tendency to think, “well that was 500 years ago. We’re not going to see that again for a long time.” But there’s no way to rule those things out. We have to consider those droughts within the realm of expectation.

The climate system is capable of rapid and abrupt change. And I think that we’ve been lulled into complacency by a couple of really wet decades, from 1976 to 1995. And so those wet decades are really unusual – they’re not just unusual for the 20th Century, they’re unusual for in terms of the last few hundred years. And if that’s what we’re calling the norm, then I think that perhaps we’re mistaken. That’s not the norm. That’s an extremely wet period, and we should expect to see precipitation that’s less than that. And of course that’s what’s been happening over the last few years. This is not a matter of trying to scare the public – that’s not the objective here at all. It’s actually for the public to be concerned that we have to be careful about overextending our use of water and other resources because natural climate variability may determine the outcome of that overuse, when things take a turn for the worse.

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

Julio L. Betancourt
U.S. Geological Survey
Desert Laboratory
Tucson, AZ