Learn how plants respond to higher atmospheric CO2 concentrations

How does rising atmospheric CO2 affect marine organisms?

Click to locate material archived on our website by topic


Extreme Weather Events: Are they Influenced by Rising Atmospheric CO2?

3.2.2. Natural Drought Variability Seen from Long-term, Centennial-scale Studies


Beyond short-term analyses of only a few decades, a number of studies have examined droughts over centennial to millennial time scales. These studies, which comprise those reviewed in this section, allow the comparison of drought events that occurred prior to the modern buildup of anthropogenic CO2 in the air with those that occurred after it. These types of analyses reveal great detail about the breadth and depth of natural variability and are of great value in investigating the potential influence of rising CO2 on droughts.

Starting in the central United States, Tian et al. (2006) derived a 31-century high-resolution ?18O record of aridity, which they obtained from sediments extracted from Steel Lake (46°58'N, 94°41'W) in north-central Minnesota, USA. Among their findings, they note that "the region was relatively dry during the Medieval Climate Anomaly (~1400-1100 AD) and relatively wet during the Little Ice Age (~1850-1500 AD), but that the moisture regime varied greatly within each of these two periods." Their most striking finding of all, however, was the fact that "drought variability was anomalously low during the 20th century." In fact, it was so depressed, as they describe it, that "~90% of the variability values during the last 3100 years were greater than the 20th-century average."

Stambaugh et al. (2011) "used a new long tree-ring chronology developed from the central U.S. to reconstruct annual drought and characterize past drought duration, frequency, and cycles in the agriculturally-important U.S. Corn Belt region during the last millennium," which chronology they calibrated and verified against monthly values of the instrumental Palmer Hydrologic Drought Index during the summer season of June, July and August. In doing so, the six scientists report that "20th century droughts, including the Dust Bowl, were relatively unremarkable when compared to drought durations prior to the instrumental record." They note, for example, that the 19th century was the driest of the past millennium, with major drought periods occurring from about 1816 to 1844 and 1849 to 1880, during what they describe as the transition out of the Little Ice Age. Prior to that, there had been 45 years of drought in the latter part of the 17th century that were coincident with the Maunder Minimum of solar activity, which is associated with the coldest period of the current interglacial. And going back further in time, there was an approximately 35-year drought in the mid- to late-15th century during "a period of decreased radiative forcing and northern hemisphere temperatures."

Eclipsing them all, however, Stambaugh et al. write that "the approximately 61-year drought in the late 12th century (ca. AD 1148-1208) appears to be the most significant drought of the entire reconstruction," noting that it "corresponds to the single greatest megadrought in North America during the last 2000 years (Cook et al., 2007), as well as "unmatched persistent low flows in western U.S. river basins (Meko et al., 2007)." And this drought, as they describe it, occurred during the middle of the Medieval Warm Period-"an interval of warmer temperatures between approximately AD 800-1300 characterized by greater drought duration and frequency in the Northern Great Plains compared to more modern times." Thus, it is abundantly clear from Stambaugh et al.'s findings that there is nothing unusual, unnatural or unprecedented about any 20th or 21st century droughts that may have occurred throughout the agricultural heartland of the United States. It is also clear that the much greater droughts of the past millennium occurred during periods of both relative cold and relative warmth, as well as the transitions between them.

It is clear from the two prior studies that there is nothing unusual, unnatural, or unprecedented about recent droughts in the central United States. Droughts of greater duration and intensity have occurred numerous times in the past, eclipsing anything that has been observed in the modern record. Claims of increasing future drought as a result of global warming are therefore not supported by real-world data, as modern global warming, if anything, has tended to lessen drought conditions throughout the central third of the United States.

Moving to the east, Quiring (2004) introduced his study of the subject by describing the drought of 2001-2002, which by June of the latter year had produced anomalously dry conditions along most of the east coast of the country, including severe drought conditions from New Jersey to northern Florida that forced 13 states to ration water. Shortly after the drought began to subside in October of 2002, however, moist conditions returned and persisted for about a year, producing the wettest growing-season of the instrumental record. These observations, in Quiring's words, "raise some interesting questions," including the one considered here-"are moisture conditions in this region becoming more variable?"

Using an 800-year tree-ring-based reconstruction of the Palmer Hydrological Drought Index to address this question, Quiring documented the frequency, severity and duration of growing-season moisture anomalies in the southern mid-Atlantic region of the United States. Among other things, this work revealed, in Quiring's words, that "conditions during the 18th century were much wetter than they are today, and the droughts that occurred during the 16th century tended to be both longer and more severe." He therefore concluded that "the recent growing-season moisture anomalies that occurred during 2002 and 2003 can only be considered rare events if they are evaluated with respect to the relatively short instrumental record (1895-2003)," for when compared to the 800-year reconstructed record, he notes that "neither of these events is particularly unusual." In addition, Quiring reports that "although climate models predict decreases in summer precipitation and significant increases in the frequency and duration of extreme droughts, the data indicate that growing-season moisture conditions during the 20th century (and even the last 19 years) appear to be near normal (well within the range of natural climate variability) when compared to the 800-year record."

Across the continent, Gedalof et al. (2004) used a network of 32 drought-sensitive tree-ring chronologies to reconstruct mean water-year flow on the Columbia River at The Dales in Oregon since 1750. This study of the second largest drainage basin in the United States is stated by them to have been done "for the purpose of assessing the representativeness of recent observations, especially with respect to low frequency changes and extreme events." When finished, it revealed, in their words, that "persistent low flows during the 1840s were probably the most severe of the past 250 years," and that "the drought of the 1930s is probably the second most severe." More recent droughts, in the words of the researchers, "were not exceptional in the context of the last 250 years and were of shorter duration than many past events." In fact, they say that "the period from 1950 to 1987 is anomalous in the context of this record for having no notable multiyear drought events," demonstrating the fact that Pacific Northwest droughts have not become more severe or long-lasting as temperatures have risen over the course of the 20th century.

MacDonald and Tingstad (2007) examined instrumental climate records to outline historical spatiotemporal patterns of precipitation variability in the Uinta Mountains, after which they "used tree-ring width chronologies from Pinus edulis Engelm. (two-needle pinyon pine) trees growing near the northern and southern flanks of the mountains to produce an ~600-year reconstruction (AD 1405-2001) of Palmer Drought Severity Index [PDSI] for Utah Climate Division 5," which they say "allows for the placement of 20th century droughts within the longer context of natural drought variability and also allows for the detection of long-term trends in drought." The researchers report that "in the context of prolonged severe droughts," the 20th century "has been relatively moist compared to preceding centuries," and they say their PDSI reconstruction and the Uinta Basin precipitation reconstruction indicate "the early to mid-17th century in particular, and portions of the 18th and 19th centuries, experienced prolonged (>10 years) dry conditions that would be unusually severe by 20th century standards," noting that "the most striking example of widespread extended drought occurred during a ~45-year period between 1625 and 1670 when PDSI only rarely rose above negative values."

Allen et al. (2013) introduce their study by noting climate models "predict the western U.S. will experience reduced snowpack, increased temperatures, and more severe and longer duration droughts," citing Barnett et al. (2004), Cook et al. (2004) and Barnett and Pierce (2009)," and they add the models predict the warming will "intensify the effects of droughts and their economic impact," citing Rauscher et al. (2008). In a test of such claims, Allen et al. "created six new tree-ring chronologies in northern Utah, which were used with pre-existing chronologies from Utah and western Wyoming to reconstruct mean annual flow for the Logan River, the largest tributary of the regionally important Bear River," which efforts resulted in what they say is "the first extended record of streamflow in northern Utah," covering the 400-year period of AD 1605-2005.

According to the six scientists, their work reveals "the Logan River has experienced highly variable streamflow over the last four centuries," adding this variability "is only partly apparent when considering only the instrumental record." And in this regard they further note "the instrumental record does not capture the full range of natural variability," as they say has been found to be the case "in studies in surrounding basins and across the western U.S.," citing Graumlich et al. (2003), Woodhouse et al. (2006), Timilsena et al. (2007), Watson et al. (2009) Barnett et al. (2010) and Wise (2001). More specifically, they indicate their reconstructions of Logan River flow suggest "overall flows were more variable at times preceding the instrumental period," and "it is likely that past droughts and wet periods [were] more extreme than the models indicate, thereby implying the possibility that water supplies may have been more volatile in the past." It would therefore appear that, rather than causing droughts and floods to become both more frequent and severe than they have been in the past, as climate models are prone to predict for a CO2-warmed world (such as is claimed we now reside in by many of the world's climate alarmists), in the case of northern Utah and western Wyoming-and possibly much more of the intermountain U.S. west-just the opposite appears to have been the case when lengthy real-world data sets have been obtained and analyzed.

Introducing their study of "perfect drought" in Southern California, MacDonald et al. (2008) define the term as "a prolonged drought that affects southern California, the Sacramento River basin and the upper Colorado River basin simultaneously." They note the instrumental record indicates the occurrence of such droughts throughout the past century, but they "generally persist for less than five years." That they have occurred at all, however, suggests the possibility of even longer perfect droughts, which could well prove catastrophic for the region. Thus, the three researchers explored the likelihood of such droughts occurring in the future, based on dendrochronological reconstructions of the winter Palmer Drought Severity Index (PDSI) in southern California over the past thousand years, plus the concomitant annual discharges of the Sacramento and Colorado Rivers, under the logical assumption that what has occurred before may well occur again. And in doing so, MacDonald et al. found that "prolonged perfect droughts (~30-60 years), which produced arid conditions in all three regions simultaneously, developed in the mid-11th century and the mid-12th century during the period of the so-called 'Medieval Climate Anomaly'," which is also widely known as the Medieval Warm Period, leading them to conclude that "prolonged perfect droughts due to natural or anthropogenic changes in radiative forcing, are a clear possibility for the near future."

In another study, Woodhouse and Lukas (2006) developed "a network of 14 annual streamflow reconstructions, 300-600 years long, for gages in the Upper Colorado and South Platte River basins in Colorado generated from new and existing tree-ring chronologies." And in the words of the two researchers, their reconstructions indicate that "the 20th-century gage record does not fully represent the range of streamflow characteristics seen in the prior two to five centuries." Of greatest significance, in this regard, was the fact that "multi-year drought events more severe than the 1950s drought have occurred," and that "the greatest frequency of extreme low flow events occurred in the 19th century," with a "clustering of extreme event years in the 1840s and 1850s."

Covering the whole of the western United States was Woodhouse (2004), who reported what is known about natural hydroclimatic variability throughout the entire region via descriptions of several major droughts that occurred there over the past three millennia, all but the last century of which had atmospheric CO2 concentrations that never varied by more than about 10 ppm from a mean value of 280 ppm.

For comparative purposes, Woodhouse began by noting that "the most extensive U.S. droughts in the 20th century were the 1930s Dust Bowl and the 1950s droughts." The first of these droughts lasted "most of the decade of the 1930s" and "occurred in several waves," while the latter "also occurred in several waves over the years 1951-1956." Far more severe than either of these two droughts, however, was the 16th-Century Megadrought, which lasted from 1580 to 1600 and included northwestern Mexico in addition to the southwestern United States and the western Great Plains. Then there was The Great Drought, which spanned the last quarter of the 13th century and was actually the last in a series of three 13th-century droughts, the first of which may have been even more severe than the last. In addition, Woodhouse noted there was a period of remarkably sustained drought in the second half of the 12th century.

It is evident from these observations, according to Woodhouse, that "the 20th-century climate record contains only a subset of the range of natural climate variability in centuries-long and longer paleoclimatic records." It is also obvious that this subset, as it pertains to water shortage, does not even begin to approach the level of drought severity and duration experienced in prior centuries and millennia, which fact was confirmed in a separate paper published by Woodhouse and four collaborators a few years later (Woodhouse et al., 2010). This being the case, it is also clear it would take a drought much more extreme than the most extreme droughts of the 20th century to propel the western United States and adjacent portions of Canada and Mexico into a truly unprecedented state of dryness.

A similar assessment was reached by Cook et al. (2010), who prefaced their analysis by writing that "IPCC Assessment Report 4 model projections suggest that the subtropical dry zones of the world will both dry and expand poleward in the future due to greenhouse warming," and that "the US southwest is particularly vulnerable in this regard and model projections indicate a progressive drying there out to the end of the 21st century." They then state "the USA has been in a state of drought over much of the West for about 10 years now," and "while severe, this turn of the century drought has not yet clearly exceeded the severity of two exceptional droughts in the 20th century," so "while the coincidence between the turn of the century drought and projected drying in the Southwest is cause for concern, it is premature to claim that the model projections are correct."

This fact is understood when the "turn of the century drought" is compared with the two "exceptional droughts" that preceded it by a few decades. Based on gridded instrumental Palmer Drought Severity indices for tree ring reconstruction that extend back to 1900, for example, Cook et al. (2010) calculated that the turn-of-the-century drought had its greatest Drought Area Index value of 59% in the year 2002, while the Great Plains/Southwest drought covered 62% of the US in its peak year of 1954, and the Dust Bowl drought covered 77% of the US in 1934. In terms of drought duration, on the other hand, things are not quite as clear. Stahle et al. (2007) estimated that the first two droughts lasted for 12 and 14 years, respectively; Seager et al. (2005) estimated them to have lasted for 8 and 10 years; and Andreadis et al. (2005) estimated them to have lasted for 7 and 8 years, yielding means of 9 and 11 years for the two exceptional droughts, which durations are to be compared to 10 or so years for the turn-of-the-century drought, which again makes the latter drought not unprecedented compared to those that occurred earlier in the 20th century.

Real clarity, however, comes when the turn-of-the-century drought is compared to droughts of the prior millennium. Cook et al. (2010) write that "perhaps the most famous example is the 'Great Drouth' (sic) of AD 1276-1299 described by A.E. Douglass (1929, 1935)." Yet this 24-year drought was eclipsed by the 38-year drought that was found by Weakley (1965) to have occurred in Nebraska from AD 1276 to 1313, which the authors say "may have been a more prolonged northerly extension of the 'Great Drouth'." But even these multi-decade droughts truly pale in comparison to the "two extraordinary droughts discovered by Stine (1994) in California that lasted more than two centuries before AD 1112 and more than 140 years before AD 1350." And each of these megadroughts, as Cook et al. (2010) describe them, occurred, in their words, "in the so-called Medieval Warm Period." And they add that "all of this happened prior to the strong greenhouse gas warming that began with the Industrial Revolution."

Given that the above-referenced medieval megadroughts "occurred without any need for enhanced radiative forcing due to anthropogenic greenhouse gas forcing"-because, of course, there was none at that time-Cook et al. (2010) rightfully concluded "there is no guarantee that the response of the climate system to greenhouse gas forcing will result in megadroughts of the kind experienced by North America in the past." And if proponents of the CO2-induced global warming hypothesis refuse to acknowledge this possibility and continue to claim that global warming will most assuredly trigger the occurrence of medieval-like megadroughts, they will also have to acknowledge that the Medieval Warm Period of a thousand years ago had to have been much warmer than the Current Warm Period has been to date. But this acknowledgement destroys yet another of their claims, i.e., that the Earth is currently warmer than it has been for one (Mann et al., 1999) to two (Mann and Jones, 2003) millennia.

Moving up to Canada, St. George and Nielsen (2002) used "a ringwidth chronology developed from living, historical and subfossil bur oak in the Red River basin to reconstruct annual precipitation in southern Manitoba since AD 1409." According to the authors, "prior to the 20th century, southern Manitoba's climate was more extreme and variable, with prolonged intervals that were wetter and drier than any time following permanent Euro-Canadian settlement." Thus, 20th-century warming, if anything, led to more stable climatic conditions with fewer hydrologic extremes (floods and droughts) than was typical of prior Little Ice Age conditions. Consequently, St. George and Nielsen concluded that "climatic case studies in regional drought and flood planning based exclusively on experience during the 20th century may dramatically underestimate true worst-case scenarios."

Also working in Canada, Wolfe et al. (2005) conducted a multi-proxy hydro-ecological analysis of Spruce Island Lake in the northern Peace sector of the Peace-Athabasca Delta in northern Alberta. Their research revealed that hydro-ecological conditions in that region varied substantially over the past 300 years, especially in terms of multi-decadal dry and wet periods. More specifically, they found (1) recent drying in the region was not the product of Peace River flow regulation that began in 1968, but rather the product of an extended drying period that was initiated in the early to mid-1900s, (2) the multi-proxy hydro-ecological variables they analyzed were well correlated with other reconstructed records of natural climate variability, and (3) hydro-ecological conditions after 1968 have remained well within the broad range of natural variability observed over the past 300 years, with the earlier portion of the record actually depicting "markedly wetter and drier conditions compared to recent decades."

At the Pacific coast of North America (Heal Lake near the city of Victoria on Canada's Vancouver Island), Zhang and Hebda (2005) conducted dendroclimatological analyses of 121 well-preserved subfossil logs discovered at the bottom of the lake plus 29 Douglas-fir trees growing nearby that led to the development of an ~4,000-year chronology exhibiting sensitivity to spring precipitation. And in doing so, they found that "the magnitude and duration of climatic variability during the past 4,000 years are not well represented by the variation in the brief modern period." As an example of this fact, they note that spring droughts represented by ring-width departures exceeding two standard deviations below the mean in at least five consecutive years occurred in the late AD 1840s and mid-1460s, as well as the mid-1860s BC, and were more severe than any drought of the 20th century. In addition, the most persistent drought occurred during the 120-year period between about AD 1440 and 1560. Other severe droughts of multi-decadal duration occurred in the mid AD 760s-800s, the 540s-560s, the 150s-late-190s and around 800 BC. Wavelet analyses of the tree-ring chronology also revealed a host of natural oscillations on timescales of years to centuries, demonstrating that the 20th century was in no way unusual in this regard, as there were many times throughout the prior 4,000 years when it was both wetter and drier than it was during the last century of the past millennium.

Sauchyn et al. (2011), in their study of the subject, wrote that "a growing demand for the surface water resources of the Canadian Prairie Provinces has resulted in increasing vulnerability to hydrological drought," citing the studies of Schindler and Donahue (2006) and Wheaton et al. (2008); and they further note, in this regard, that "a shift in the amount and timing of streamflow represents the most serious risk from recent and projected climate warming in western Canada (Sauchyn et al., 2001)," adding that "the Saskatchewan River Basin is among Canada's most vulnerable watersheds, in terms of projected climate changes and impacts, and the sensitivity of natural systems and economic activities to Canada's most variable hydroclimate." Therefore, they felt it important to know the characteristics of past streamflow variability in order to better prepare for future droughts, as well as to determine if extreme droughts that may occur in the future might be due to CO2-induced global warming or if they are within the range of natural variability experienced in the past, when the air's CO2 concentration was both much lower and less variable than it is currently. And so the question arises: Is a mere century of real-world data sufficient for these purposes?

In a study designed to explore this important question by determining if streamflow variability recorded by the streamflow gauge at Edmonton, Alberta (Canada) over the past century (since 1912) is representative of the range of variability experienced there over the past millennium, Sauchyn et al. (2011) developed a 945-year reconstruction of the annual flow of the Northern Saskatchewan River based on tree-rings collected from seven different sites within the runoff-generating upper basin of the river (see Figure 4).


Figure 4. North Saskatchewan River reconstructed water year (October to September) flow for the period 1063-2006. Adapted from Sauchyn et al. 2011.

Clearly, the Edmonton stream-gauge record does not "represent the full extent of inter-annual to multi-decadal variability in the tree-ring data," for as noted by Sauchyn et al. (2011) "there are periods of low flow in the pre-instrumental record that are longer and more severe than those recorded by the gauge" and which "pre-date Euro-Canadian settlement of the region." Two of these extreme events were approximate 30-year droughts, one occurring in the early 1700s and another during the mid-1100s, while one of the two most prominent mega-droughts lasted for most of the 14th century, while the other occurred in the latter part of the 15th century.

Sauchyn et al. thus go on to state "there is less certainty and stationarity in western [Canadian] water supplies than implied by the instrumental record," which they say is "the conventional basis for water resource management and planning" of the region. Likewise, it is clear that their streamflow reconstruction provides a whole new-and-improved basis for determining the "uniqueness" of whatever future droughts might occur throughout the region, making it much more difficult to claim that such droughts were caused by anthropogenic CO2 emissions, since there was far less CO2 in Earth's atmosphere prior to the 1912 start-date of the region's prior streamflow history, when several way more serious droughts than those of the past century are now known to have occurred.

Still in North America, but working down in Mexico, Diaz et al. (2002) constructed a history of winter-spring (November-April) precipitation-which accounts for one-third of the yearly total- for the state of Chihuahua for the period 1647-1992, based on earlywood width chronologies of over 300 Douglas fir trees growing at four locations along the western and southern borders of Chihuahua and at two locations in the United States just above Chihuahua's northeast border. And on the basis of these reconstructions, they note that "three of the 5 worst winter-spring drought years in the past three-and-a-half centuries are estimated to have occurred during the 20th century." Although this observation tends to make the 20th century look highly anomalous in this regard, it is not; for two of those three worst drought years occurred during a period of average to slightly-above-average precipitation.

Diaz et al. also note that "the longest drought indicated by the smoothed reconstruction lasted 17 years (1948-1964)," which is again indicative of abnormally dry conditions during the 20th century. However, for several of the 17 years of that below-normal-precipitation interval, precipitation values were only slightly below normal. For all practical purposes, therefore, there were four very similar dry periods interspersed throughout the preceding two and a half centuries: one in the late 1850s and early 1860s, one in the late 1790s and early 1800s, one in the late 1720s and early 1730s, and one in the late 1660s and early 1670s.

With respect to the 20th century alone, there was also a long period of high winter-spring precipitation that stretched from 1905 to 1932; and following the major drought of the 1950s, precipitation remained at or just slightly above normal for the remainder of the record. Finally, with respect to the entire 346 years, there was no long-term trend in the data, nor was there evidence of any sustained departure from that trend over the course of the 20th century, indicating that neither 20th century anthropogenic CO2 emissions nor 20th century warming significantly impacted rainfall in the Mexican state of Chihuahua.

Cleaveland et al. (2003) constructed a winter-spring (November-March) precipitation history for the period 1386-1993 for Durango, Mexico, based on earlywood width chronologies of Douglas-fir tree rings collected at two sites in the Sierra Madre Occidental. They report that this record "shows droughts of greater magnitude and longer duration than the worst historical drought that occurred in the 1950s and 1960s." These earlier dramatic droughts included the long dry spell of the 1850s-1860s and what they called the megadrought of the mid- to late-16th century. Their work clearly demonstrates, therefore, that the worst droughts of the past 600 years did not occur during the period of greatest warmth. Instead, they occurred during the Little Ice Age, which was perhaps the coldest period of the current interglacial.

Going back further in time, Hodell et al. (1995) had provided evidence for a protracted drought during the Terminal Classic Period of Mayan civilization (AD 800-1000), based on their analysis of a sediment core retrieved in 1993 from Lake Chichanacanab in the center of the northern Yucatan Peninsula of Mexico. Subsequently, based on two additional sediment cores retrieved from the same location in 2000, Hodell et al. (2001) determined that the massive drought likely occurred in two distinct phases (750-875 and 1000-1075). Reconstructing the climatic history of the region over the past 2,600 years and applying spectral analysis to the data also revealed a significant recurrent drought periodicity of 208 years that matched well with a cosmic ray-produced 14C record preserved in tree rings, which is believed to reflect variations in solar activity; and because of the good correspondence between the two data sets, they concluded that "a significant component of century-scale variability in Yucatan droughts is explained by solar forcing."

Working in central Mexico, Therrell et al. (2006) "developed a continuous, exactly dated, tree-ring reconstruction of maize yield variability" over the period 1474 to 2001 in an effort to provide "new insight into the history of climate and food availability in the heartland of the Mesoamerican cultural province." This work was made possible by latewood-width data they derived from what they describe as "the second-most southerly native stand of Douglas-fir (Pseudtosuga menziesii) trees known in the Americas." In addition, the authors compared their reconstruction to "historical records of crop failure and famine in order to cross-validate the tree-ring and historical records."

Therrell et al.'s plot of reconstructed drought-induced maize-yield anomalies exposed a total of seven major decadal-scale yield shortfalls over the past 500 years, with a mean rate of occurrence of 1.5 per century over the 400-year period AD 1500-1900. Over the 20th century, however, there was only one such multi-year famine, and its magnitude paled in comparison to that of the average such event of the preceding four centuries. Thus, the so-called unprecedented warming of the 20th century did not produce the alarmist-predicted effect on drought in central Mexico. In fact, the threat of major drought-induced famines in this part of the world appears to have lessened with increased warming.

Moving into South America, Webster et al. (2007) removed an active stalagmite (MC01) from the entrance chamber of Macal Chasm - a cave on the Vaca Plateau west of the Rio Macal in the Cavo District of Belize near the border with Guatemala (~17°N, 89°W) - from which the authors obtained "reliably dated reflectance, color, luminescence, and C and O stable isotope records for the period from 1225 BC to the present." Upon examination of the record, they found that the interval "from AD 750 to 1150 was the most prolonged dry phase in our 3,300-year record," which period of time corresponds well with the MWP's mean time of occurrence around the globe, which period, in their words, "coincided with the collapse of the Maya civilization." More specifically, they say their data depict "a series of droughts centered at about AD 780, 910, 1074, and 1139," with "successive droughts increasing in severity."

Mundo et al. (2012) employed 43 new and updated tree-ring chronologies from a network of Araucaria araucana and Austrocedrus chilensis trees in reconstructing the October-June mean streamflow of Argentina's Neuquen River over the 654-year period AD 1346-2000. According to the eight researchers who conducted this study, in terms of the frequency, intensity and duration of droughts and pluvial events, they say "the 20th century contains some of the driest and wettest annual to decadal-scale events in the last 654 years." However - and it's a very big however - they report that "longer and more severe events were recorded in previous centuries," the significance of which becomes apparent when it is recognized that the bulk of the 554 years that preceded the 20th century were part of the much colder Little Ice Age. Therefore, it would appear that, if anything, the "unprecedented" global warming of the past century has brought Argentina's Neuquen River less extreme streamflow conditions, which is just the opposite of what model-based projections suggest should have happened.

Masiokas et al. (2012) developed the first reconstruction and quantitative analysis of variations in snow accumulation of the past eight-and-a-half centuries in the Andes between 30° and 37°S. This record was based on "instrumental rainfall and streamflow data from adjacent lowlands, a variety of documentary records, and century-long tree-ring series of precipitation-sensitive species from the western side of the Andes," representing "the first attempt to reconstruct annually-resolved, serially complete snowpack variations spanning most of the past millennium in the Southern Hemisphere," which record "allows testing the relative severity of recent 'extreme' conditions in a substantially longer context."

Based on their findings, the eight researchers who conducted the study report that "variations observed in the last 60 years are not particularly anomalous when assessed in a multi-century context," noting that both extreme high and low snowpack values "have not been unusual when assessed in the context of the past eight centuries." Indeed, they say "the most extreme dry decades are concentrated between the late 16th century and the mid-18th century," and there were "decade-long periods of high snowpack levels that equaled or probably surpassed those recorded during the past six decades."

Shifting to a different continent, Therrell et al. (2006) developed what they describe as "the first tree-ring reconstruction of rainfall in tropical Africa using a 200-year regional chronology based on samples of Pterocarpus angolensis [a deciduous tropical hardwood known locally as Mukwa] from Zimbabwe." This project revealed "a decadal-scale drought reconstructed from 1882 to 1896 matches the most severe sustained drought during the instrumental period (1989-1995)," and "an even more severe drought is indicated from 1859 to 1868 in both the tree-ring and documentary data." They report, for example, the year 1860 (which was the most droughty year of the entire period), was described in a contemporary account from Botswana (where part of their tree-ring chronology originated) as "a season of 'severe and universal drought' with 'food of every description' being 'exceedingly scarce' and the losses of cattle being 'very severe' (Nash and Endfield, 2002)," while at the other end of the moisture spectrum, Therrel et al. report "a 6-year wet period at the turn of the nineteenth century (1897-1902) exceeds any wet episode during the instrumental era." Consequently, for a large part of central southern Africa, it is clear that the supposedly unprecedented global warming of the 20th century did not result in an intensification of either extreme dry or wet periods. If anything, just the opposite appears to have occurred.

Similar findings were reported by Esper et al. (2007). In prefacing their work, they stated that "analysis of the PDSI [Palmer Drought Severity Index], a standardized measure of surface moisture conditions, revealed distinct 20th century aridity changes in vulnerable NW Africa, including a sharp downward trend towards drier conditions in the 1980s (Luterbacher et al., 2006)," but they indicated that "a high-resolution long-term reconstruction that could place current conditions in the context of the past millennium is missing for N Africa," which was exactly what the authors hoped they could remedy. More specifically, Esper et al. re-used Cedrus atlantica tree-ring data generated in the 1980s (Glueck and Stockton, 2001) and combined these measurements with a major update made in 2002, which allowed "analysis of tree growth and instrumental data during the current drought episode in comparison to PDSI estimates back to AD 1049."

The six scientists who conducted this study reported that "PDSI values were above average for most of the 1450-1980 period, which let recent drought appear exceptional." However, they say the long-term results they obtained indicate the "pluvial episode of the past millennium was preceded by generally drier conditions back to 1049," leading them to state the late 20th-century drought "appears more typical when associated with conditions before 1400." In addition, they concluded their paper by stating that the "ultimate drivers" for the medieval hydroclimate pattern that led to the earlier drought conditions in Morocco "seemed to be high solar irradiance and low volcanic forcings," citing Emile-Geay et al. (2007).

Probing some 1500 years into the past was the study of Holmes et al. (1997), who wrote that since the late 1960s, the African Sahel had experienced "one of the most persistent droughts recorded by the entire global meteorological record." However, in a high-resolution study of a sediment sequence extracted from an oasis in the Manga Grasslands of northeast Nigeria, they too determined that "the present drought is not unique and that drought has recurred on a centennial to interdecadal timescale during the last 1500 years."

Moving to Asia, Jiang et al. (2005) analyzed historical documents to produce a time series of flood and drought occurrences in eastern China's Yangtze Delta since AD 1000. Their work also revealed that alternating wet and dry episodes occurred throughout this lengthy period; and the data demonstrate that droughts and floods usually occurred in the spring and autumn seasons of the same year, with the most rapid and strongest of these fluctuations occurring during the Little Ice Age (1500-1850), as opposed to the preceding Medieval Warm Period and the following Current Warm Period.

Writing as background for their work, Cai et al. (2014) state "investigations of the natural climate background of the Chinese Loess Plateau (CLP) are crucial for understanding the processes and characteristics of climate change in this region as well as the current status of the climate." Thus, they set out to conduct a historical analysis of drought for this region, attempting to answer the following two key questions: (1) "Did the drought severity or frequency increase in response to the global warming?" and (2) Is the present drought in the region "unprecedented during the last three centuries?" The analysis was made possible by the development of a new regional tree-ring chronology from Chinese pine trees located in three different sites of the Lingkong Mountain area (112°10'-112°15' E, 36°31'-36°43' N) of the southeast CLP. The resulting chronology was positively correlated with monthly Palmer Drought Severity Index (PDSI) values obtained from meteorological data for the region over the period 1954-2005. Based on that correlation, the researchers were able to reconstruct a record of historic PDSI since 1703 AD.

Cai et al. report the existence of seven dry periods and six wet periods in the 306-year reconstruction. The driest interval occurred between 1867 and 1932, while the wettest interval followed between 1934 and 1957. With respect to the most recent drought (1993-2008), there was nothing unique or unprecedented about it. Rather, the team of researchers report it is "still within the frame of natural climate variability." In addition, multi-taper spectral analysis identified periodicities of 37.9 and 102 years in the reconstructed PDSI record, which Cai et al. say "resemble the 35-year Bruckner (Raspopov et al., 2004) and Gleissberg cycles of solar activity (Sonett et al., 1990; Braun et al., 2005), respectively." Further analysis of the PDSI reconstruction and the NOAA sunspot time series revealed a significant correlation between the two variables that Cai et al. say "convincingly [supports] the influence of solar activity on moisture variations in the Lingkong Mountain area." Thus, the results of their analysis speak for themselves in answering the two questions posed by the authors at the beginning of their study. Recent drought in the CLP is not unprecedented and is more likely the product of natural, as opposed to anthropogenic, forcings.

Kalugin et al. (2005) utilized sediment cores from Lake Teletskoye in the Altai Mountains of Southern Siberia to produce a multi-proxy climate record spanning the past 800 years. With respect to moisture and precipitation, Kalugin et al. state that the period between 1210 and 1480 was more humid than today, while the period between 1480 and 1840 was more arid. In addition, they report three episodes of multi-year drought (1580-1600, 1665-1690 and 1785-1810), which findings are in agreement with other historical data and tree-ring records from the Mongolia-Altai region (Butvilovskii, 1993; Jacoby et al., 1996; Panyushkina et al., 2000). Consequently, this study also proves problematic in attempting to support the claim that global warming will lead to more frequent and more severe droughts, as all of the major multi-year droughts detected in this study occurred during the cool phase of the 800-year record.

In another study, Kim et al. (2009) developed a 200-year history of precipitation measured at Seoul, Korea (1807 to 2006), along with the results of a number of "progressive methods for assessing drought severity from diverse points of view," starting with (1) the Effective Drought Index (EDI) developed by Byun and Wilhite (1999), which Kim et al. describe as "an intensive measure that considers daily water accumulation with a weighting function for time passage," (2) a Corrected EDI that "considers the rapid runoff of water resources after heavy rainfall" (CEDI), (3) an Accumulated EDI that "considers the drought severity and duration of individual drought events" (AEDI), and (4) a Year-accumulated negative EDI "representing annual drought severity" (YAEDI).

These researchers' precipitation history and two of their drought severity histories are presented, in that order, in Figures 5 and 6.


Figure 5. Annual precipitation history at Seoul, Korea, where the solid line represents a 30-year moving-average. Adapted from Kim et al. (2009).


Figure 6. Annual "dryness" history at Seoul, Korea, represented by YAEDI365 (Sum of daily negative EDI values divided by 365, represented by bars) and YAEDIND (Sum of daily negative EDI values divided by total days of negative EDI, represented by open circles). Adapted from Kim et al. (2009).

In viewing the results depicted in Figures 5 and 6, it is clear that the only major multi-year deviation from long-term normalcy is the decadal-scale decrease in precipitation and ensuing drought, each of which achieved their most extreme values (low in the case of precipitation, high in the case of drought) in the vicinity of AD 1900, well before the 20th century rise in atmospheric CO2 and global temperatures. The significant post-Little Ice Age warming of the planet, therefore, had essentially no effect at all on the long-term histories of either precipitation or drought at Seoul, Korea.

Sinha et al. (2011) explored the issue of drought in India. In writing as background for their study they warn the return of a severe drought to the region could pose "a particular serious threat for the predominantly agrarian-based societies of monsoon Asia, where the lives of billions of people are tightly intertwined with the annual monsoon cycle." And as a result of such concern, the group of eight researchers, hailing from China, Germany and the United States, review in great detail what is known about the history of the monsoon, relying heavily on the work of Sinha et al. (2007) and Berkelhammer et al. (2010).

Based on their review, Sinha et al. (2011) state that "proxy reconstructions of precipitation from central India, north-central China [Zhang et al., 2008], and southern Vietnam [Buckley et al., 2010] reveal a series of monsoon droughts during the mid-14th-15th centuries that each lasted for several years to decades," and they say "these monsoon megadroughts have no analog during the instrumental period." They also report that "emerging tree ring-based reconstructions of monsoon variability from SE Asia (Buckley et al., 2007; Sano et al., 2009) and India (Borgaonkar et al., 2010) suggest that the mid-14th-15th century megadroughts were the first in a series of spatially widespread megadroughts that occurred during the Little Ice Age," and they say they "appear to have played a major role in shaping significant regional societal changes at that time." Among these upheavals, they make special mention of "famines and significant political reorganization within India (Dando, 1980; Pant et al., 1993; Maharatna, 1996), the collapse of the Yuan dynasty in China (Zhang et al., 2008), the Rajarata civilization in Sri Lanka (Indrapala, 1971), and the Khmer civilization of Angkor Wat fame in Cambodia (Buckley et al., 2010)," noting that the evidence suggests that "monsoon megadroughts may have played a major contributing role in shaping these societal changes." And contrary to conventional climate-alarmist thinking on the subject, all of these droughts occurred prior to the modern increase of atmospheric CO2 and the supposedly unprecedented warming of the past few decades.

Over in Europe and concentrating on Central Scandinavia, Linderholm and Chen (2005) derived a 500-year history of winter (September-April) precipitation from tree-ring data obtained within the Northern Boreal zone of the region. This chronology indicated that below average precipitation was observed during the periods 1504-1520, 1562-1625, 1648-1669, 1696-1731, 1852-1871 and 1893-1958, with the lowest values occurring at the beginning of the record and at the beginning of the 17th century. These results clearly demonstrate that for this portion of the European continent, 20th-century global warming did not result in more frequent or more severe droughts.

In a related study conducted in east central Sweden, Linderholm and Molin (2005) analyzed two independent precipitation proxies, one derived from tree-ring data and one from a farmer's diary, to produce a 250-year record of summer (Jun-Aug) precipitation. This work indicated there had been a high degree of variability in summer precipitation on inter-annual to decadal time scales throughout the record, but with the past century exhibiting less variability than the 150 years that preceded it. One period that stood out vividly was a persistent dry episode between 1806 and 1832, when the tree-ring history revealed its longest consecutive period of below-average tree growth, which was associated with a concomitant period of drought that was documented in the farmer's diary.

Moving further in the continent toward Germany, Wilson et al. (2005) used the regional curve standardization technique to develop a summer (March-August) precipitation chronology from living and historical ring-widths of trees in the Bavarian Forest region of southeast Germany for the period 1456-2001. This technique captured low frequency variations that indicate the region was substantially drier than the long-term average during the periods 1500-1560, 1610-1730 and 1810-1870, all of which intervals were much colder than the bulk of the 20th century. Thus, these results, too, fly in the face of model-based predictions.

Another study of interest concerns the Danube River in western Europe, where several researchers had studied the precipitation histories of adjacent regions and suggested that an anthropogenic signal was present in the latter decades of the 20th century, and that it was responsible for that period's supposedly drier conditions. Determined to investigate further, Ducic (2005) examined these claims by analyzing observed and reconstructed discharge rates of the river near Orsova, Serbia over the period 1731-1990. This work revealed that the lowest 5-year discharge value in the pre-instrumental era (1831-1835) was practically equal to the lowest 5-year discharge value in the instrumental era (1946-1950), and that the driest decade of the entire 260-year period was 1831-1840. What is more, the discharge rate for the last decade of the record (1981-1990), which prior researchers had claimed was anthropogenically-influenced, was found to be "completely inside the limits of the whole series," in Ducic's words, and only 0.7% less than the 260-year mean, leading to the conclusion that "modern discharge fluctuations do not point to dominant anthropogenic influence." In fact, Ducic's correlative analysis suggests the detected cyclicity in the record could "point to the domination of the influence of solar activity."

Also of note, Pfister et al. (2006) identified extremely low water stages within the Upper Rhine River Basin via hydrological measurements made since 1808 at Basel, Switzerland, while "for the period prior to 1808, rocks emerging in rivers and lakes in the case of low water were used along with narrative evidence for assessing extreme events." This work revealed that "29 severe winter droughts are documented since 1540," which events, in their words, "occurred after a succession of four months with below-average precipitation" associated with "persistent anticyclones centered over Western Europe." Of most interest, in this regard, was their finding that "severe winter droughts were relatively rare in the 20th century compared to the former period, which is due to increased winter temperature and precipitation." And in discussing the generality of their findings, they note that "extended droughts in the winter half-year in Central Europe were more frequent, more persistent and more severe during the Little Ice Age than in the preceding 'Medieval Warm Period' and the subsequent 'warm 20th century' (Pfister, 2005)," which facts suggest a relationship just the opposite of what is typically assumed by climate models to be the case.

In summation, the aforementioned studies demonstrate the need to have long-term (millennial-scale) records of droughts in order to determine how exceptional 20th-century changes in their characteristics might have been, which can help to determine whether there is compelling reason to attribute such changes to historical increases in the atmospheric concentrations of various greenhouse gases. And considering the findings presented above, real-world evidence suggests that the global warming of the past century or so has not led to either a greater frequency or severity of drought. Indeed, even the worst droughts in recorded meteorological history do not seem to have been any worse (in fact, they are actually much milder) than droughts that occurred in the historic past. And as a result, there is little reason to put any credence whatsoever in the potential for global warming to lead to more frequent or severe droughts.

Back to the Table of Contents