- Email: [email protected]
Climate variability and change in the drylands of Western North America
Global and Planetary Change 64 (2008) 111–118
Contents lists available at ScienceDirect
Global and Planetary Change j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / g l o p l a c h a
Climate variability and change in the drylands of Western North America M.K. Hughes a,⁎, H.F. Diaz b a b
Laboratory of Tree-Ring Research, University of Arizona, Tucson, AZ 85721, USA Cooperative Institute for Research in Environmental Sciences (CIRES) NOAA/ESRL/PSD1, 325 Broadway, Boulder, CO 80305, USA
a r t i c l e
i n f o
Article history: Received 24 March 2008 Accepted 28 July 2008 Available online 19 October 2008 Keywords: Western North America semi-arid regions natural and anthropogenic climatic change climate variability paleoclimatic reconstructions drought
a b s t r a c t We argue that it is important to expand the consideration of climate in the context of provision of ecosystem services in drylands. In addition to climate change, it is necessary to include climate variability on timescales relevant to human and ecological considerations, namely interannual to decadal and multidecadal. The period of global instrumental record (about a century and a half long at the very most) is neither an adequate nor an unbiased sample of the range and character of natural climate variability that might be expected with the climate system configured as it is now. We base this on evidence from W. N. America, where there has recently been a major multi-year drought, of a scale and intensity that has occurred several times in the last 2000 years, and on attempts to provide explanations of these phenomena based on physical climatology. Ensembles of runs of forced climate system models suggest the next 50 years will bring much more extensive and intense drought in the continental interior of North America. The trajectory followed by the supply of ecosystem services will be contingent not only on the genotypes available and the antecedent soil, economic and social conditions but also on climate variability and change. The critical features of climate on which patterns of plant growth and water supply depend may vary sharply during and between human generations, resulting in very different experiences and hence, expectations. © 2008 Elsevier B.V. All rights reserved.
1. Introduction 1.1. Climate, desertification and ecosystem services It is commonplace to link the degradation of soil and vegetation characteristics of desertification with particular climates. These links have ranged from the taxonomic, in which particular climates are associated with desert conditions (Kottek et al., 2006), to the dynamic, in which the surface conditions of degraded vegetation and soil themselves interact with the atmospheric environment to reinforce the climatic conditions conducive to such degradation (e.g. Charney, 1975). Neither of these views takes account of the role of humans and other organisms. In contrast, the Millennium Ecosystem Assessment (MEA) (2005) takes a more nuanced view, in which desertification is a subset of the possible changes that might occur in a landscape's capacity to provide a range of ecosystem services. In order to explore the range of possible trajectories of the landscape in this context, MEA argues that it is necessary to consider both climate (in this case anthropogenic climate change) and biodiversity loss. Both kinds of change independently and in combination can affect the provision of ecosystem services. Moreover, the ecological change associated with ⁎ Corresponding author. Tel.: +1 520 621 2191; fax: +1 520 621 8229. E-mail addresses: [email protected] (M.K. Hughes), [email protected] (H.F. Diaz). 0921-8181/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.gloplacha.2008.07.005
these two factors will be further influenced by the effects of increasing atmospheric carbon dioxide concentrations on water use efficiency for some plants, and thereby affect the composition and structure of the vegetation. The MEA concluded, with “medium certainty” that, “… although climate change may increase aridity and desertification risk in many areas, the consequent effects on services driven by biodiversity loss and, hence, on desertification, are difficult to predict” (page 18). We add a further element to this complex picture. In the earlier static views, climates were quantitative descriptions of conditions derived from some period of years, usually thirty, which were assumed to be representative of what might reasonably be expected at that place. In the MEA view, observed and expected climate change is added to this. We aim to convince the reader that it is important to expand the consideration of climate in this context to include climate variability on timescales relevant to human and ecological considerations, namely interannual to decadal and multidecadal. We base this on what may be the most important finding of the recently burgeoning field of high-resolution Paleoclimatology (Hughes, 2002). This is that the period of global instrumental record (about a century and a half long at the very most) is neither an adequate nor an unbiased sample of the range and character of natural climate variability that might be expected with the climate system configured as it is now. Put another way, patterns of variation have existed in the past 1000–2000 years that have not been seen in the 20th century. Some of these patterns involve extreme climate
112
M.K. Hughes, H.F. Diaz / Global and Planetary Change 64 (2008) 111–118
conditions on time scales from interannual to multi-century, particularly droughts, likely to have major ecological and societal impacts were they to recur now. We know no reason why they should not. 2. Western North America as a representative region 2.1. Geography Western North America is one of the largest contiguous regions of drylands in a highly developed region. As a result of this, it is a useful object for study of the interactions of climate change and climate variability with desertification, because it has been particularly well observed and recorded. The drylands of Western North America contain many kinds of desert vegetation and of desert climate, and numerous examples of the effects of intense resource use. So, they may be considered representative of many situations found around the world. 2.2. Climate teleconnections Climatic conditions in this region are linked to those of other mid-latitude regions of descending air. Cayan et al. (1998) found decadal variability in precipitation in the Southwestern United States …“to be aligned with opposing precipitation fluctuations in North Africa” (their page 3148). They discussed specific climatic mechanisms that might lead to this alignment, and indeed might cause decadal-scale variability throughout the dry mid-latitudes. It is also worth noting that such decadal variability accounted for 20– 50% of the variance of annual precipitation in Western North America. 2.3. A wealth of natural archives A further, compelling reason for our focus on Western North America is its extraordinary wealth of developed natural archives of climate variability in recent millennia, including abundant ancient trees growing under climatic stress, geomorphologic features associated with glacier activity and lake level changes, and laminated marine sediments. This permits an unusual degree of cross-checking between completely independent natural archives. The annual resolution and multi-millennial length of many of these records allows them to yield information on climatic fluctuations on all time scales from interannual to multi-millennial. 3. Climate change There is growing evidence that the climate of the Western U.S. has undergone a number of changes in the last three decades. These include a widespread and substantial decline in precipitation together with sustained warming (Fig. 1), with some amplification of warming trends at higher elevations (Diaz and Eischeid, 2007). These climate trends combine to produce a smaller fraction of precipitation as snow, earlier snowmelt, and changes in stream flow (e.g. Cayan et al., 2001). Warmer summers also contribute to greater evaporative loss. This has resulted in an increase of the area that would meet the Koeppen “desert” classification during a recent severe drought in the region (Fig. 2). Using a multivariable detection and attribution methodology, Barnett et al. (2008, page 1080) demonstrated that “the majority of the observed low-frequency changes in the hydrological cycle…over the western United States from 1950 to 1999 are due to human-caused climate changes from greenhouse gases and aerosols”. They note that the changes already observed “differ in length and strength from trends expected as a result of natural variability…and differ in the specific ways expected of human-induced effects”.
4. Climate variability 4.1. Time scales of variability Using the instrumental record, climate variability in the Western United States, in particular in precipitation, has been demonstrated to occur on interannual and decadal time scales (Diaz, 1983; Dettinger et al., 1998). This variability exists in the spatiotemporal domain, so that, for example, there are “regional north–south contrasts that appear at many timescales” (Dettinger et al., 1998, page 3109). Much of this variability has been linked to antecedent conditions over the Pacific Basin, for example the El Niño-Southern Oscillation phenomenon on interannual time scales (Bradley et al., 1987; Kiladis and Diaz, 1989), and indices of sea surface temperature on decadal time scales in both the Pacific (Mantua et al., 1997) and Atlantic Oceans (McCabe et al., 2004). These marine influences may interact to modify their expression in the hydroclimate of different parts of the continent (Gershunov and Barnett, 1998). In order to detect the existence of climate variability on longer timescales, multi-decadal to centennial, for example, it is necessary to use the kinds of natural archives mentioned above. Moreover, because of the size of the region (~30° of latitude by ~30° of longitude), it is necessary to first consider the record from sub-regions. 4.2. Paleoclimatic evidence 4.2.1. California and the Great Basin Giant sequoia (Sequoiadendron giganteum (Lindley) Buchholz 1939) tree rings from the western flanks of the Sierra Nevada show variations in the frequency of extreme single year droughts in the Central Valley of California over the last 2100 years (Hughes and Brown, 1992; Brown et al., 1992; Hughes et al., 1996). The incidence of such droughts on a century time scale has varied more than threefold, with highest frequencies in the 3rd and 4th, 8th and 9th, and 15th and 16th centuries. The 20th century frequency was slightly below the 2100 year mean. Not only tree rings show that there is a greater tendency for droughts to be intense and persistent between AD 400 and AD 1600. Stine (1994) identified extreme low stands in Mono Lake, a closed basin on the California/Nevada border, lasting from the early 10th century to the end of the 11th and from the beginning of the 13th to the middle of the 14th, coinciding with droughts seen in the rings of drought-sensitive bristlecone pine (Pinus longaeva D.K. Bailey 1970) from the neighboring White Mountains (LaMarche, 1974; Hughes and Graumlich, 1996), in the Sierra Nevada (Graumlich 1993; Graybill and Funkhouser 2000) and in the Great Basin (Hughes and Funkhouser 1998). Stine (1994) inferred that these low stands of Mono Lake resulted from a precipitation deficit of 35% or more over many decades, far more intense and greater in duration than at any time since European settlement in the region. Graham and Hughes (2007) used a completely independent approach to reconstruct the level of Mono Lake for the past 2000 years. They calculated the inflow to the lake from moisture sensitive tree-ring series in the Sierra Nevada. From this they reconstructed natural changes in the level of the lake, which has no outflow, using a water balance model for the lake. They identified two low stands of very similar magnitudes, and duration, to those proposed by Stine (2004) (Fig. 3). Graham and Hughes (2007) attributed the 70–100 years offset between their reconstructed low stands and those described by Stine (2004) partly to problems in their modeling, but “primarily to bias in the calibration of the 14C ages of the fossil vegetation” (their page 1207). It is notable that these were by far the deepest low stands in their reconstruction, and that their relative timing and depth corresponds so closely to Stine's observations. Further evidence that these multidecadal to century-scale features in the tree-ring reconstructions represent real fluctuations in hydroclimate comes from comparison of the tree-ring based reconstruction
M.K. Hughes, H.F. Diaz / Global and Planetary Change 64 (2008) 111–118
113
Fig. 1. a)Total linear trend change [trend coeff ⁎ (N−1)] in annual precipitation (mm) for the period AD 1979–2004. Derived from 4-km resolution PRISM data (Daly et al., 1994; 2002). PRISM Group, Oregon State University, http://www.prism.oregonstate.edu/.b) Mean annual (January–December) temperature anomalies in the Western Conterminous United States (11 States) (1895–2006), in degrees Celsius. Data from 4-km PRISM dataset as in Fig. 1a.
from bristlecone pine (Hughes and Graumlich, 1996; Hughes and Funkhouser, 1998) with δ18O in Globigerina bulloides in sediments from the Santa Barbara Basin over the last 3500 years (Kennett and Kennett 2000) as well as with other tree-ring records and with natural archives from other parts of the Pacific Basin (Graham et al., 2007). The timing, amplitude and teleconnections of most of these far-flung records are consistent with current knowledge of physical climatology. A different pattern emerges from the rate of change of δ18O in the sediments of Pyramid Lake at the northern end of the California/ Nevada border, but this record corresponds remarkably to centuryscale fluctuations in a tree-ring reconstruction of streamflow fed from the same mountains as the lake (Benson et al., 1999).
4.2.2. The Upper Basin of the Colorado River and the Southwest The gaged flow of major rivers represents an integrated measure of moisture availability in their catchments. Meko et al. (2007) have extended and updated earlier tree-ring based reconstructions of the flow of the Colorado River (Stockton and Jacoby, 1976). They have calculated annual flow at the point where the river passes from the Upper Basin (broadly corresponding to states that are net contributors to the flow) to the Lower Basin (roughly corresponding to states that are net consumers of the flow). The Upper Basin lies several hundred kilometers to the east and north of the Sierra Nevada/Great Basin regions referred to in the previous paragraph. This record shows that the reconstructed flow has almost equaled or been lower than the
114
M.K. Hughes, H.F. Diaz / Global and Planetary Change 64 (2008) 111–118
Fig. 2. Expansion of the Köppen desert type during the drought of 2000–2004, shown in dark shading. Desert category expanded by over 2% in this 5-year period—from about 10% to 12% of the total western US area. The historical extent of the desert type is shown in light shading.
lowest flow of the observed record of natural flow (1906–2004) on eight occasions since AD 800. Several of these were more extreme and persistent than any low-flow event in the observed record, most
notably those in the 900s, mid-1100s (the most extreme), late 1500s (see also Stahle et al., 2000), and late 1800/early 1900s. This is evidence of decadal-scale precipitation deficits of the order of 15–20%
Fig. 3. Simulated level of Mono Lake (meters above m.s.l.) for the past 2000 years. The symbols show elevation and calibrated 14C “death dates” (from Stine (1994) for trees (circles), shrubs (squares) and grass (triangles) with 1σ uncertainties (bars). Open symbols indicate second low stand. The vertical gray lines indicate the nominal extent of the first (solid line) and second (dashed line) low stands according to Stine (1994). The two lines of evidence indicate a precipitation deficit of ~ 35% sustained over many decades during the reconstructed low stands. After Fig. 9 in Graham and Hughes (2007).
M.K. Hughes, H.F. Diaz / Global and Planetary Change 64 (2008) 111–118
over a sub-continental-scale catchment. The Colorado River record also shows periods of decadal elevated flow, in several cases immediately following one of the persistent low-flow periods and associated drought. Such a couplet of episodes is reconstructed for the late 16th (dry) and early 17th (wet) centuries, a period designated by Biondi et al. (2000, page 76) as “the near-1600 dry/wet knockout”. They referred to the ecologic and geomorphic consequences of a sustained period of drought (vegetation morbidity/mortality, increase in vulnerability of soils to washout) followed by a much wetter period (particularly the effects on sediment loadings in rivers and the stand structure in vegetation). This sequence is reflected in reconstructions of cool season (November–April) precipitation since AD 1000 in the NOAA Climate divisions of Arizona (7 divisions) and New Mexico (8 divisions) (Ni et al., 2002). Consecutive 30-year period means may differ by as much as 39% (Fig. 4). 4.2.3. Great Plains There is evidence for greater incidence of sustained, severe drought before approximately AD 1600 than since in the northern Great Plains, from reconstructions of lake-water salinity (Fritz et al., 2000) and records of eolian activity (Muhs et al., 1997). Droughts lasting 20 years or more in the Southern Plains were reconstructed by Woodhouse (2003) using tree-ring data for the 13th and 16th centuries. She commented “That is something we have not come close to experiencing in the twentieth century” (page 106). It should be noted that her study region is characterized by extensive dune fields that are intermittently mobile (during dry periods) and stabilized (during wetter periods). 4.2.4. A subcontinental view Building on the pioneering work of Stockton and Meko (1983), Meko et al. (1993) and Cook et al. (1999). Cook et al. (2004) used reconstructed gridded Palmer Drought Severity Index (PDSI) for 286 gridpoints on a 2.5 × 2.5° grid covering North America. They then plotted a time series of the percentage of the western part of the continent experiencing drought with PDSI less than −1. Three notable features are evident. First, for most of the 20th century, the proportion of Western North America experiencing drought according to this criterion is close to or less than the 1200 year long term mean. Second, all four periods designated as having a significantly greater fraction of the area experiencing drought fell before AD 1300 (centered at AD 936, 1034, 1150 and 1253). Third, the long term mean percentage of the area experiencing drought is ~ 42% from AD 900–1300, ~38% for the whole period, and 30% for the 20th century. This continental-scale evidence for extensive, extreme episodes of drought outside the range seen in the instrumental period complements the detailed regional records discussed in the previous paragraphs. In particular, it supports
Fig. 4. Cool season (November–April) precipitation reconstruction (mm.) for Arizona climate division 2. (Ni et al., 2002). Gray line, reconstruction; varying black line—5-year running mean; Fine horizontal black line—mean for period AD 1000–1988; heavy horizontal black lines—30-year means for 1571–1600 (mn 124.2 mm, SD 43.8), 1601– 1630 (162.1, 33.9), 1871–1900 (135.1, 30.4), 1901–1930 (187.5, 58.0). Further details at http://www.climas.arizona.edu/research/paleoclimate/product/AZ2/reconstruction. html.
115
the case for multicentury “regimes” of climate over large regions of the Earth's surface, in this case the Pacific Basin and the mid-latitudes of the Americas (Stine, 1994; Hughes and Funkhouser, 1998; Graham et al., 2007). All these records point to recurrent and marked patterns of interannual to at least multi-century time-scale variability in climatic conditions in Western North America in the late Holocene. Depending on the time scale, some of these can be linked to hemispheric atmospheric circulation patterns (1-year extreme droughts) (Namias, 1978), some to very large-scale atmosphere/ocean interactions such as ENSO (Ropelewski and Halpert, 1987) and the PDO (Pacific Decadal Oscillation)(Mantua et al., 1997; Cayan et al., 1998; Dettinger et al., 1998). Multicentury-scale regime-shifts over the whole of the Pacific Basin and its margins have been invoked based solely on tree-ring evidence from North America (Cook et al., 2004) or on a very diverse set of natural archives from across the Pacific and the Americas (Graham et al., 2007). These results may best be explained by a greater (lesser) frequency of La Niña-like conditions before/after the mid-2nd millennium AD (Graham et al., 2007). Cook et al. (2004) tentatively link this inferred shift in frequency to the so-called Medieval Warm Period, although there is no consensus on the existence of any such period, or even on whether it is a meaningful concept (Hughes and Diaz, 1994; Bradley et al., 2003; Osborn and Briffa, 2005). 5. Climate variability and climate change The instrumental record and natural archives of climate over the past two millennia show Western North America to be a region in which robust spatiotemporal patterns of interannual to multidecadal and even multicentury variability are seen in hydroclimate. These time scales are relevant to many ecologic and geomorphic processes, and hence must be considered when trying to understand desertification and the associated changes in the availability of ecosystem services. Recent changes seem to be attributable to anthropogenic climate change, acknowledging the context of natural variability (Barnett et al., 2008). What might happen in the next few decades? It has been pointed out that the increased frequency and persistence of La Niñalike conditions seen before the mid-second millennium AD can be forced by anomalous heating over the Pacific (Mann et al., 2005; Cook et al., 2004; Graham et al., 2007; Seager et al., 2007a). Increased radiative forcing as a result of anthropogenic climate change could have a similar effect, modified in some important respects, including higher temperatures (Seager et al., 2007b). The result would be a drier climate for much of the North American continent. This conclusion is supported by the ensemble of climate model results, noting that the models somewhat underestimate the extent of extreme water deficit conditions for the instrumental period (Fig. 5). It follows from this that they may underestimate the severity of conditions in the mid-21st century. Even so, they project a dramatic extension of the area susceptible to land degradation. Seager et al. (2007b, page 3) write that “The most severe future droughts will still occur during persistent La Niña events but they will be worse than any since the Medieval period because the La Niña conditions will be perturbing a base state that is drier than any experienced recently”. There is ample evidence that the patterns of hydroclimatic variability revealed by paleoclimatic records in Western North America have had substantial impacts on the components of ecosystems on which modern ecosystem services depend. For example, Swetnam and Betancourt (1998) present extensive evidence for ecological responses to climate variability in this region on time scales from annual to decadal and on spatial scales from less than 100 km2 to 1,000,000 km2. The responses they review include regionally synchronized fires, insect outbreaks and pulses in tree demography. Biondi et al. (2000) examine one decadal-scale sequence in detail, the drought/pluvial sequence centered on AD 1600. They
116
M.K. Hughes, H.F. Diaz / Global and Planetary Change 64 (2008) 111–118
Fig. 5. Top panel—observed annual precipitation minus annual potential evapotranspiration (mm.) (P-PET) for the conterminous United States for the period AD 1895–2005. Middle panel: P-PET mm for AD 1895–2005 ensemble model results from 19 climate models for 42 A1B simulations prepared for the IPCC Fourth Assessment Report. Bottom panel: projected percentage change by mid-21st century from mean conditions of AD 1895–2005. Courtesy of Marty Hoerling, NOAA ESRL.
interpret changes in sediment input to the Santa Barbara Basin as indicating that the tree-ring reconstructed drought of the late 16th century “disrupted vegetation cover over large areas, allowing for major erosion during floods in the following years”, namely the first decades of the 17th century. Similarly, Graham et al. (2007) document major ecosystem changes coincident and consistent with the major climatic regime shift they propose for the middle of the second millennium AD. These changes include increased fire as recorded by diverse records, including charcoal in lacustrine sediments and dendrochronologically-dated fire scars in trees, salinity in San Francisco Bay, and vegetation composition also from sediments.
Graham et al. (2007) also noted the coincidence of drought and mobilization of dune fields covering a significant part of the Great Plains, consistent with changing salinity in lakes in this region. Note that, when stabilized, these areas provide a major and extensive rangeland resource, particularly in the Nebraska Sand Hills referred to by Muhs et al. (1997). 6. Final thoughts Much of W. N. America has recently been in a major multi-year drought, of a scale and intensity that has occurred several times in the
M.K. Hughes, H.F. Diaz / Global and Planetary Change 64 (2008) 111–118
last 2000 years, especially before the middle of the second millennium AD. Were one of these naturally occurring multi-decadal drought episodes to recur, the consequences for the availability of several important ecosystem services would be extreme. Consider the results of a multidecade interval with 35% deficiency in precipitation for natural ecosystems, rain-fed agriculture, water supply and ground water recharge in the rapidly growing American West. Ensembles of runs of forced climate system models suggest the next 50 years will bring much more extensive and intense drought in the continental interior of North America. “Desert climate” could invade much of the Great Plains, with its associated effects on vegetation and soils, and challenges to society and ecosystems. Appropriate responses will be made harder to define by the character of hydroclimatic variability in such semi-arid and arid regions. Consider, for example, the effects on decision making of a decadalscale “pluvial” (period of unusually high precipitation) in a dry region superimposed on a multidecadal drying trend. Such an event would surely make it more difficult to promote the changes in policy, behavior and engineering needed for life in a drier climate. Although it may presently be difficult to model regional climate change reliably, it should be noted that the explanations available so far for decadal to century-scale hydroclimatic variability in recent centuries are rather similar to the mechanisms proposed for nearterm (next few decades) regional climate change. Writing in the context of water resources, Milly et al. (2008, page 573) describe stationarity as a fundamental idea in the training and practice of water-resource engineering. It is the assumption that “natural systems fluctuate within an unchanging envelope of variability”, and that an adequate estimate of these parameters may be obtained from the instrumental record. They note that natural variability has usually been discounted, on the basis that it has been sufficiently small to be disregarded for engineering and policy purposes. It may be that similar, less formally expressed, assumptions underlie ways of thinking about the interactions between climate and ecological systems. We argue that in both cases, such disregard of the scale of natural variability, especially in drylands, is misguided. Milly et al. (2008) urge the abandonment of the assumption of stationarity in light of recent and projected climate change. We argue that the record of hydroclimatic variability in Western North America over the past two millennia shows the stationarity assumption to have been flawed from the beginning. Interactions between people and the land take place on these same time scales, interannual to multidecadal. The trajectory followed by the supply of ecosystem services will be contingent not only on the genotypes available and the antecedent soil, economic and social conditions. The critical features of climate on which patterns of plant growth and water supply depend may vary sharply during and between human generations, resulting in very different experiences and hence, expectations. Acknowledgements MKH was in receipt of a CIRES Visiting Fellowship at the University of Colorado, Boulder, when this paper was prepared. His contribution was also supported by grant #ATM- 0551986 from the Paleoclimatology program of the U.S. National Science Foundation. He is grateful to the organizers of the 2007 Wengen Workshop for the opportunity to participate. We are grateful to Marty Hoerling of NOAA/ESRL for Fig. 5, and to Jon Eischeid for extensive assistance. References Barnett, T.P., Pierce, D.W., Hidalgo, H.G., Bonfils, C., Santer, B.D., Das, T., Bala, G., Wood, A.W., Nozawa, T., Mirin, A.A., Cayan, D.R., Dettinger, M.D., 2008. Human-induced changes in the hydrology of the Western United States. Science 319, 1080–10833. Benson, L., Lund, S., Paillet, F., Kashgarian, M., Smoot, J., Mensing, S., Dibb, J., 1999. A 2800-yr history of oscillations in surface-water supply to the Central Valley and to the Bay Area of Northern California. Abstracts of the Fall 1999 AGU Meeting. EOS.
117
Biondi, F., Isaacs, C., Hughes, M.K., Cayan, D.R., Berger, W.H., 2000. The near-1600 dry/ wet knockout: linking terrestrial and near-shore ecosystems. Proceedings of the Twenty-fourth Annual Climate Diagnostics and Prediction Workshop. November 5– 9, 1999, Tucson, AZ. US Department of Commerce, NOAA, pp. 76–79. Bradley, R.S., Diaz, H.F., Kiladis, G.N., Eischeid, J.K., 1987. ENSO signal in continental temperatures and precipitation records. Nature 327, 497–501. Bradley, R.S., Hughes, M.K., Diaz, H.F., 2003. Climate in Medieval times. Science 302, 404–405. Brown, P.B., Hughes, M.K., Swetnam, T.W., Caprio, A.R., 1992. Giant sequoia ring-width chronologies from the central Sierra Nevada. Tree-Ring Bulletin 52, 1–14. Cayan, D.R., Dettinger, M.D., Diaz, H.F., Graham, N.E., 1998. Decadal variability of precipitation over Western North America. Journal of Climate 11, 3148–3166. Cayan, D.R., Kammerdiener, S.A., Dettinger, M.D., Caprio, J.M., Peterson, D.H., 2001. Changes in the onset of spring in the Western United States. Bulletin of the American Meteorological Society 82, 399–415. Charney, J., 1975. Dynamics of deserts and drought in the Sahel. The physical basis of climate and climate modelling. World Meteorological Organization, Geneva, Switzerland, pp. 171–175. Cook, E.R., Meko, D.M., Stahle, D.W., Cleaveland, M.K., 1999. Drought reconstructions for the continental United States. Journal of Climate 12, 1145–1162. Cook, E.R., Woodhouse, C.A., Eakin, C.M., Meko, D.M., Stahle, D.W., 2004. Long-term aridity changes in the western United States. Science 306, 1015–1018. Daly, C., Neilson, R.P., Phillips, D.L., 1994. A statistical topographic model for mapping climatological precipitation over mountainous terrain. Journal of Climate and Applied Methodology 33, 140–158. Daly, C., Gibson, W.P., Taylor, G.H., Johnson, G.L., Pasteris, P., 2002. A knowledge-based approach to the statistical mapping of climate. Climate Research 22, 99–113. Dettinger, M.D., Cayan, D.R., Diaz, H.F., Meko, D.M., 1998. North–South precipitation patterns in western North America on interannual-to-decadal time scales. Journal of Climate 11, 3095–3111. Diaz, H.F., 1983. Some aspects of major dry and wet periods in the contiguous United States, 1895–1981. Journal of Climate and Applied Meteorology 22, 3–16. Diaz, H.F., Eischeid, J.K., 2007. Disappearing “Alpine Tundra” Köppen climatic type in the Western United States. Geophysical Research Letters 34, L18707. doi:10.1029/ 2007GL031253. Fritz, S.C., Ito, E., Yu, Z., Laird, K., Engstrom, D.R., 2000. Hydrologic variation in the Northern Great Plains during the last two millennia. Quaternary Research 53, 175–184. Gershunov, A., Barnett, T.P., 1998. Interdecadal modulation of ENSO teleconnections. Bulletin of the American Meteorological Society 79 (12), 2715–2725. Graham, N.E., Hughes, M.K., 2007. Reconstructing the Mediaeval low stands of Mono Lake, Sierra Nevada, California. The Holocene 17, 1197–1210. Graham, N.E., Hughes, M.K., Ammann, C.M., Cobb, K.M., Hoerling, M.P., Xu, T., Kennett, D.J., Kennett, J.P., Rein, B., Stott, L., Wigand, P.E., 2007. Tropical Pacific-mid-latitude teleconnections in Medieval times. Climatic Change. doi:10.1007/s10584-007-9239-2. Graumlich, L.J., 1993. A 1000-year record of temperature and precipitation in the Sierra Nevada. Quaternary Research 39, 249–255. Graybill, D.A., Funkhouser, G., 2000. Dendroclimatic reconstructions during the past millennium in the southern Sierra Nevada and Owens Valley, California. In: Lavenberg, R. (Ed.), Southern California Climate: Trends and Extremes of the Past 2000 years. Natural History Museum of Los Angeles County, Los Angeles, CA. Hughes, M.K., 2002. Dendrochronology in climatology—the state of the art. Dendrochronologia 20, 95–116. Hughes, M.K., Brown, P.M., 1992. Drought frequency in central California since 101 B.C. recorded in giant sequoia tree rings. Climate Dynamics 6, 161–167. Hughes, M.K., Diaz, H.F., 1994. Was there a Medieval Warm Period, and if so, where and when? Climatic Change 26, 109–142. Hughes, M.K., Graumlich, L.J., 1996. Multimillennial dendroclimatic records from western North America. In: Bradley, R.S., Jones, P.D., Jouzel, J. (Eds.), Climatic Variations and Forcing Mechanisms of the Last 2000 years. Springer Verlag, Berlin, pp. 109–124. Hughes, M.K., Funkhouser, G., 1998. Extremes of moisture availability reconstructed from tree rings for recent millennia in the Great Basin of Western North America. In: Beniston, M., Innes, J.L. (Eds.), The Impacts of Climate Variability on Forests. Springer, Berlin, pp. 99–107. Hughes, M.K., Touchan, R., Brown, P., 1996. A multimillennial network of giant sequoia chronologies for dendroclimatology. In: Dean, J., Meko, D.M., Swetnam, T.W. (Eds.), Tree Rings, Environment and Humanity, Proceedings of the International Conference. Arizona, Radiocarbon, Tucson, pp. 225–234. Kennett, D.J., Kennett, J.P., 2000. Competitive and cooperative responses to climatic instability in coastal southern California. American Antiquity 65, 379–395. Kiladis, G.N., Diaz, H.F., 1989. Global climatic anomalies associated with extremes in the Southern Oscillation. Journal of Climate 2, 1069–1090. Kottek, M., Grieser, J., Beck, C., Rudolf, B., Rubel, F., 2006. World map of the KöppenGeiger climate classification updated. Meteorologische Zeitschrift 15, 259–263. LaMarche, V.C., 1974. Paleoclimatic inferences from long tree-ring records. Science 183, 1043–1088. Mann, M.E., Cane, M.A., Zebiak, S.E., Clement, A., 2005. Volcanic and solar forcing of the Tropical Pacific over the past 1000 years. Journal of Climate 18, 447–456. Mantua, N.J., Hare, S.R., Zhang, Y., Wallace, J.M., Francis, R.C., 1997. A Pacific interdecadal climate oscillation with impacts on salmon production. Bulletin of the American Meteorological Society 78, 1069–1079. McCabe, G.J., Palecki, M.A., Betancourt, J.L., 2004. Pacific and Atlantic Ocean influences on multidecadal drought frequency in the United States. Proceedings of the National Academy of Sciences 101 (12), 4136–4141. Meko, D.M., Cook, E.R., Stahle, D.W., Stockton, C.W., Hughes, M.K., 1993. Spatial patterns of tree-growth anomalies in the United States and southeastern Canada. Journal of Climate 6, 1773–1786.
118
M.K. Hughes, H.F. Diaz / Global and Planetary Change 64 (2008) 111–118
Meko, D.M., Woodhouse, C.A., Baisan, C.A., Knight, T., Lukas, J.J., Hughes, M.K., Salzer, M.W., 2007. Medieval drought in the upper Colorado River Basin. Geophysical Research Letters 34. doi:10.1029/2007GL029988. Millennium Ecosystem Assessment, 2005. Ecosystems and Human Well-being: Desertification Synthesis. World Resources Institute, Washington, DC. Milly, P.C.D., Betancourt, J., Falkenmark, M., Hirsch, R.M., Kundzewicz, Z.W., Lettenmaier, D.P., Stouffer, R.J., 2008. Stationarity is dead: whither water management? Science 319, 573–574. Muhs, D.R., Stafford, T.W., Swinehart, J.B., Cowherd, S.A., Bush, R.F., Madole, R.F., Maat, P.B., 1997. Late Holocene eolian activity in the mineralogically mature Nebraska Sand Hills. Quaternary Research 48, 162–176. Namias, J., 1978. Multiple causes of the North American abnormal winter 1976–77. Monthly Weather Review 106, 279–295. Ni, F., Cavazos, T., Hughes, M.K., Comrie, A.C., Funkhouser, G., 2002. Cool season precipitation in the Southwestern United States since AD 1000: comparison of linear and nonlinear techniques for reconstruction. International Journal of Climatology 22, 1645–1662. Osborn, T., Briffa, K., 2005. The spatial extent of 20th-century warmth in the context of the past 1200 years. Science 311, 841–844. Ropelewski, C.F., Halpert, M.S., 1987. Global and regional sale precipitation patterns associated with the El Niño-Southern oscillation. Monthly Weather Review 115, 1606–1626.
Seager, R., Graham, N., Herweijer, C., Gordon, A.L., Kushnir, Y., Cook, E., 2007a. Blueprints for Medieval hydroclimate. Quaternary Science Reviews 26, 2322–2336. Seager, R., Ting, M., Held, I., Kushnir, Y., Lu, J., Vecchi, G., Huang, H.-P., Harnik, N., Leetmaa, A., Lau, N.-C., Li, C., Velez, J., Naik, N., 2007b. Model projections of an imminent transition to a more arid climate in Southwestern North America. Science 316 (5828), 1181–1184. Stahle, D.W., Cook, E.R., Cleaveland, M.K., Therrell, M.D., Meko, D.M., Grissino-Mayer, H.D., Watson, E., Luckman, B., 2000. Tree-ring Data Document 16th Century Megadrought Over North America, 81. EOS, p. 121,125. Stine, S., 1994. Extreme and persistent drought in California and Patagonia during mediaeval time. Nature 269, 546–549. Stockton, C.W., Jacoby, G.C., 1976. Long-term surface water supply and streamflow levels in the upper Colorado River basin. Lake Powell Research Project Bulletin 18 70pp. Stockton, C.W., Meko, D.M., 1983. Drought recurrence in the Great Plains as reconstructed from long-term tree-ring records. Journal of Climate and Applied Meteorology 2, 17–29. Swetnam, T.W., Betancourt, J.L., 1998. Mesoscale disturbance and ecological response to decadal climate variability in the American Southwest. Journal of Climate 11, 3128–3147. Woodhouse, C.A., 2003. Droughts of the past, implications for the future? In: Smith, S.L. (Ed.), The future of the Southern Plains. University of Oklahoma Press, Norman, OK, pp. 95–113.
Contents lists available at ScienceDirect
Global and Planetary Change j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / g l o p l a c h a
Climate variability and change in the drylands of Western North America M.K. Hughes a,⁎, H.F. Diaz b a b
Laboratory of Tree-Ring Research, University of Arizona, Tucson, AZ 85721, USA Cooperative Institute for Research in Environmental Sciences (CIRES) NOAA/ESRL/PSD1, 325 Broadway, Boulder, CO 80305, USA
a r t i c l e
i n f o
Article history: Received 24 March 2008 Accepted 28 July 2008 Available online 19 October 2008 Keywords: Western North America semi-arid regions natural and anthropogenic climatic change climate variability paleoclimatic reconstructions drought
a b s t r a c t We argue that it is important to expand the consideration of climate in the context of provision of ecosystem services in drylands. In addition to climate change, it is necessary to include climate variability on timescales relevant to human and ecological considerations, namely interannual to decadal and multidecadal. The period of global instrumental record (about a century and a half long at the very most) is neither an adequate nor an unbiased sample of the range and character of natural climate variability that might be expected with the climate system configured as it is now. We base this on evidence from W. N. America, where there has recently been a major multi-year drought, of a scale and intensity that has occurred several times in the last 2000 years, and on attempts to provide explanations of these phenomena based on physical climatology. Ensembles of runs of forced climate system models suggest the next 50 years will bring much more extensive and intense drought in the continental interior of North America. The trajectory followed by the supply of ecosystem services will be contingent not only on the genotypes available and the antecedent soil, economic and social conditions but also on climate variability and change. The critical features of climate on which patterns of plant growth and water supply depend may vary sharply during and between human generations, resulting in very different experiences and hence, expectations. © 2008 Elsevier B.V. All rights reserved.
1. Introduction 1.1. Climate, desertification and ecosystem services It is commonplace to link the degradation of soil and vegetation characteristics of desertification with particular climates. These links have ranged from the taxonomic, in which particular climates are associated with desert conditions (Kottek et al., 2006), to the dynamic, in which the surface conditions of degraded vegetation and soil themselves interact with the atmospheric environment to reinforce the climatic conditions conducive to such degradation (e.g. Charney, 1975). Neither of these views takes account of the role of humans and other organisms. In contrast, the Millennium Ecosystem Assessment (MEA) (2005) takes a more nuanced view, in which desertification is a subset of the possible changes that might occur in a landscape's capacity to provide a range of ecosystem services. In order to explore the range of possible trajectories of the landscape in this context, MEA argues that it is necessary to consider both climate (in this case anthropogenic climate change) and biodiversity loss. Both kinds of change independently and in combination can affect the provision of ecosystem services. Moreover, the ecological change associated with ⁎ Corresponding author. Tel.: +1 520 621 2191; fax: +1 520 621 8229. E-mail addresses: [email protected] (M.K. Hughes), [email protected] (H.F. Diaz). 0921-8181/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.gloplacha.2008.07.005
these two factors will be further influenced by the effects of increasing atmospheric carbon dioxide concentrations on water use efficiency for some plants, and thereby affect the composition and structure of the vegetation. The MEA concluded, with “medium certainty” that, “… although climate change may increase aridity and desertification risk in many areas, the consequent effects on services driven by biodiversity loss and, hence, on desertification, are difficult to predict” (page 18). We add a further element to this complex picture. In the earlier static views, climates were quantitative descriptions of conditions derived from some period of years, usually thirty, which were assumed to be representative of what might reasonably be expected at that place. In the MEA view, observed and expected climate change is added to this. We aim to convince the reader that it is important to expand the consideration of climate in this context to include climate variability on timescales relevant to human and ecological considerations, namely interannual to decadal and multidecadal. We base this on what may be the most important finding of the recently burgeoning field of high-resolution Paleoclimatology (Hughes, 2002). This is that the period of global instrumental record (about a century and a half long at the very most) is neither an adequate nor an unbiased sample of the range and character of natural climate variability that might be expected with the climate system configured as it is now. Put another way, patterns of variation have existed in the past 1000–2000 years that have not been seen in the 20th century. Some of these patterns involve extreme climate
112
M.K. Hughes, H.F. Diaz / Global and Planetary Change 64 (2008) 111–118
conditions on time scales from interannual to multi-century, particularly droughts, likely to have major ecological and societal impacts were they to recur now. We know no reason why they should not. 2. Western North America as a representative region 2.1. Geography Western North America is one of the largest contiguous regions of drylands in a highly developed region. As a result of this, it is a useful object for study of the interactions of climate change and climate variability with desertification, because it has been particularly well observed and recorded. The drylands of Western North America contain many kinds of desert vegetation and of desert climate, and numerous examples of the effects of intense resource use. So, they may be considered representative of many situations found around the world. 2.2. Climate teleconnections Climatic conditions in this region are linked to those of other mid-latitude regions of descending air. Cayan et al. (1998) found decadal variability in precipitation in the Southwestern United States …“to be aligned with opposing precipitation fluctuations in North Africa” (their page 3148). They discussed specific climatic mechanisms that might lead to this alignment, and indeed might cause decadal-scale variability throughout the dry mid-latitudes. It is also worth noting that such decadal variability accounted for 20– 50% of the variance of annual precipitation in Western North America. 2.3. A wealth of natural archives A further, compelling reason for our focus on Western North America is its extraordinary wealth of developed natural archives of climate variability in recent millennia, including abundant ancient trees growing under climatic stress, geomorphologic features associated with glacier activity and lake level changes, and laminated marine sediments. This permits an unusual degree of cross-checking between completely independent natural archives. The annual resolution and multi-millennial length of many of these records allows them to yield information on climatic fluctuations on all time scales from interannual to multi-millennial. 3. Climate change There is growing evidence that the climate of the Western U.S. has undergone a number of changes in the last three decades. These include a widespread and substantial decline in precipitation together with sustained warming (Fig. 1), with some amplification of warming trends at higher elevations (Diaz and Eischeid, 2007). These climate trends combine to produce a smaller fraction of precipitation as snow, earlier snowmelt, and changes in stream flow (e.g. Cayan et al., 2001). Warmer summers also contribute to greater evaporative loss. This has resulted in an increase of the area that would meet the Koeppen “desert” classification during a recent severe drought in the region (Fig. 2). Using a multivariable detection and attribution methodology, Barnett et al. (2008, page 1080) demonstrated that “the majority of the observed low-frequency changes in the hydrological cycle…over the western United States from 1950 to 1999 are due to human-caused climate changes from greenhouse gases and aerosols”. They note that the changes already observed “differ in length and strength from trends expected as a result of natural variability…and differ in the specific ways expected of human-induced effects”.
4. Climate variability 4.1. Time scales of variability Using the instrumental record, climate variability in the Western United States, in particular in precipitation, has been demonstrated to occur on interannual and decadal time scales (Diaz, 1983; Dettinger et al., 1998). This variability exists in the spatiotemporal domain, so that, for example, there are “regional north–south contrasts that appear at many timescales” (Dettinger et al., 1998, page 3109). Much of this variability has been linked to antecedent conditions over the Pacific Basin, for example the El Niño-Southern Oscillation phenomenon on interannual time scales (Bradley et al., 1987; Kiladis and Diaz, 1989), and indices of sea surface temperature on decadal time scales in both the Pacific (Mantua et al., 1997) and Atlantic Oceans (McCabe et al., 2004). These marine influences may interact to modify their expression in the hydroclimate of different parts of the continent (Gershunov and Barnett, 1998). In order to detect the existence of climate variability on longer timescales, multi-decadal to centennial, for example, it is necessary to use the kinds of natural archives mentioned above. Moreover, because of the size of the region (~30° of latitude by ~30° of longitude), it is necessary to first consider the record from sub-regions. 4.2. Paleoclimatic evidence 4.2.1. California and the Great Basin Giant sequoia (Sequoiadendron giganteum (Lindley) Buchholz 1939) tree rings from the western flanks of the Sierra Nevada show variations in the frequency of extreme single year droughts in the Central Valley of California over the last 2100 years (Hughes and Brown, 1992; Brown et al., 1992; Hughes et al., 1996). The incidence of such droughts on a century time scale has varied more than threefold, with highest frequencies in the 3rd and 4th, 8th and 9th, and 15th and 16th centuries. The 20th century frequency was slightly below the 2100 year mean. Not only tree rings show that there is a greater tendency for droughts to be intense and persistent between AD 400 and AD 1600. Stine (1994) identified extreme low stands in Mono Lake, a closed basin on the California/Nevada border, lasting from the early 10th century to the end of the 11th and from the beginning of the 13th to the middle of the 14th, coinciding with droughts seen in the rings of drought-sensitive bristlecone pine (Pinus longaeva D.K. Bailey 1970) from the neighboring White Mountains (LaMarche, 1974; Hughes and Graumlich, 1996), in the Sierra Nevada (Graumlich 1993; Graybill and Funkhouser 2000) and in the Great Basin (Hughes and Funkhouser 1998). Stine (1994) inferred that these low stands of Mono Lake resulted from a precipitation deficit of 35% or more over many decades, far more intense and greater in duration than at any time since European settlement in the region. Graham and Hughes (2007) used a completely independent approach to reconstruct the level of Mono Lake for the past 2000 years. They calculated the inflow to the lake from moisture sensitive tree-ring series in the Sierra Nevada. From this they reconstructed natural changes in the level of the lake, which has no outflow, using a water balance model for the lake. They identified two low stands of very similar magnitudes, and duration, to those proposed by Stine (2004) (Fig. 3). Graham and Hughes (2007) attributed the 70–100 years offset between their reconstructed low stands and those described by Stine (2004) partly to problems in their modeling, but “primarily to bias in the calibration of the 14C ages of the fossil vegetation” (their page 1207). It is notable that these were by far the deepest low stands in their reconstruction, and that their relative timing and depth corresponds so closely to Stine's observations. Further evidence that these multidecadal to century-scale features in the tree-ring reconstructions represent real fluctuations in hydroclimate comes from comparison of the tree-ring based reconstruction
M.K. Hughes, H.F. Diaz / Global and Planetary Change 64 (2008) 111–118
113
Fig. 1. a)Total linear trend change [trend coeff ⁎ (N−1)] in annual precipitation (mm) for the period AD 1979–2004. Derived from 4-km resolution PRISM data (Daly et al., 1994; 2002). PRISM Group, Oregon State University, http://www.prism.oregonstate.edu/.b) Mean annual (January–December) temperature anomalies in the Western Conterminous United States (11 States) (1895–2006), in degrees Celsius. Data from 4-km PRISM dataset as in Fig. 1a.
from bristlecone pine (Hughes and Graumlich, 1996; Hughes and Funkhouser, 1998) with δ18O in Globigerina bulloides in sediments from the Santa Barbara Basin over the last 3500 years (Kennett and Kennett 2000) as well as with other tree-ring records and with natural archives from other parts of the Pacific Basin (Graham et al., 2007). The timing, amplitude and teleconnections of most of these far-flung records are consistent with current knowledge of physical climatology. A different pattern emerges from the rate of change of δ18O in the sediments of Pyramid Lake at the northern end of the California/ Nevada border, but this record corresponds remarkably to centuryscale fluctuations in a tree-ring reconstruction of streamflow fed from the same mountains as the lake (Benson et al., 1999).
4.2.2. The Upper Basin of the Colorado River and the Southwest The gaged flow of major rivers represents an integrated measure of moisture availability in their catchments. Meko et al. (2007) have extended and updated earlier tree-ring based reconstructions of the flow of the Colorado River (Stockton and Jacoby, 1976). They have calculated annual flow at the point where the river passes from the Upper Basin (broadly corresponding to states that are net contributors to the flow) to the Lower Basin (roughly corresponding to states that are net consumers of the flow). The Upper Basin lies several hundred kilometers to the east and north of the Sierra Nevada/Great Basin regions referred to in the previous paragraph. This record shows that the reconstructed flow has almost equaled or been lower than the
114
M.K. Hughes, H.F. Diaz / Global and Planetary Change 64 (2008) 111–118
Fig. 2. Expansion of the Köppen desert type during the drought of 2000–2004, shown in dark shading. Desert category expanded by over 2% in this 5-year period—from about 10% to 12% of the total western US area. The historical extent of the desert type is shown in light shading.
lowest flow of the observed record of natural flow (1906–2004) on eight occasions since AD 800. Several of these were more extreme and persistent than any low-flow event in the observed record, most
notably those in the 900s, mid-1100s (the most extreme), late 1500s (see also Stahle et al., 2000), and late 1800/early 1900s. This is evidence of decadal-scale precipitation deficits of the order of 15–20%
Fig. 3. Simulated level of Mono Lake (meters above m.s.l.) for the past 2000 years. The symbols show elevation and calibrated 14C “death dates” (from Stine (1994) for trees (circles), shrubs (squares) and grass (triangles) with 1σ uncertainties (bars). Open symbols indicate second low stand. The vertical gray lines indicate the nominal extent of the first (solid line) and second (dashed line) low stands according to Stine (1994). The two lines of evidence indicate a precipitation deficit of ~ 35% sustained over many decades during the reconstructed low stands. After Fig. 9 in Graham and Hughes (2007).
M.K. Hughes, H.F. Diaz / Global and Planetary Change 64 (2008) 111–118
over a sub-continental-scale catchment. The Colorado River record also shows periods of decadal elevated flow, in several cases immediately following one of the persistent low-flow periods and associated drought. Such a couplet of episodes is reconstructed for the late 16th (dry) and early 17th (wet) centuries, a period designated by Biondi et al. (2000, page 76) as “the near-1600 dry/wet knockout”. They referred to the ecologic and geomorphic consequences of a sustained period of drought (vegetation morbidity/mortality, increase in vulnerability of soils to washout) followed by a much wetter period (particularly the effects on sediment loadings in rivers and the stand structure in vegetation). This sequence is reflected in reconstructions of cool season (November–April) precipitation since AD 1000 in the NOAA Climate divisions of Arizona (7 divisions) and New Mexico (8 divisions) (Ni et al., 2002). Consecutive 30-year period means may differ by as much as 39% (Fig. 4). 4.2.3. Great Plains There is evidence for greater incidence of sustained, severe drought before approximately AD 1600 than since in the northern Great Plains, from reconstructions of lake-water salinity (Fritz et al., 2000) and records of eolian activity (Muhs et al., 1997). Droughts lasting 20 years or more in the Southern Plains were reconstructed by Woodhouse (2003) using tree-ring data for the 13th and 16th centuries. She commented “That is something we have not come close to experiencing in the twentieth century” (page 106). It should be noted that her study region is characterized by extensive dune fields that are intermittently mobile (during dry periods) and stabilized (during wetter periods). 4.2.4. A subcontinental view Building on the pioneering work of Stockton and Meko (1983), Meko et al. (1993) and Cook et al. (1999). Cook et al. (2004) used reconstructed gridded Palmer Drought Severity Index (PDSI) for 286 gridpoints on a 2.5 × 2.5° grid covering North America. They then plotted a time series of the percentage of the western part of the continent experiencing drought with PDSI less than −1. Three notable features are evident. First, for most of the 20th century, the proportion of Western North America experiencing drought according to this criterion is close to or less than the 1200 year long term mean. Second, all four periods designated as having a significantly greater fraction of the area experiencing drought fell before AD 1300 (centered at AD 936, 1034, 1150 and 1253). Third, the long term mean percentage of the area experiencing drought is ~ 42% from AD 900–1300, ~38% for the whole period, and 30% for the 20th century. This continental-scale evidence for extensive, extreme episodes of drought outside the range seen in the instrumental period complements the detailed regional records discussed in the previous paragraphs. In particular, it supports
Fig. 4. Cool season (November–April) precipitation reconstruction (mm.) for Arizona climate division 2. (Ni et al., 2002). Gray line, reconstruction; varying black line—5-year running mean; Fine horizontal black line—mean for period AD 1000–1988; heavy horizontal black lines—30-year means for 1571–1600 (mn 124.2 mm, SD 43.8), 1601– 1630 (162.1, 33.9), 1871–1900 (135.1, 30.4), 1901–1930 (187.5, 58.0). Further details at http://www.climas.arizona.edu/research/paleoclimate/product/AZ2/reconstruction. html.
115
the case for multicentury “regimes” of climate over large regions of the Earth's surface, in this case the Pacific Basin and the mid-latitudes of the Americas (Stine, 1994; Hughes and Funkhouser, 1998; Graham et al., 2007). All these records point to recurrent and marked patterns of interannual to at least multi-century time-scale variability in climatic conditions in Western North America in the late Holocene. Depending on the time scale, some of these can be linked to hemispheric atmospheric circulation patterns (1-year extreme droughts) (Namias, 1978), some to very large-scale atmosphere/ocean interactions such as ENSO (Ropelewski and Halpert, 1987) and the PDO (Pacific Decadal Oscillation)(Mantua et al., 1997; Cayan et al., 1998; Dettinger et al., 1998). Multicentury-scale regime-shifts over the whole of the Pacific Basin and its margins have been invoked based solely on tree-ring evidence from North America (Cook et al., 2004) or on a very diverse set of natural archives from across the Pacific and the Americas (Graham et al., 2007). These results may best be explained by a greater (lesser) frequency of La Niña-like conditions before/after the mid-2nd millennium AD (Graham et al., 2007). Cook et al. (2004) tentatively link this inferred shift in frequency to the so-called Medieval Warm Period, although there is no consensus on the existence of any such period, or even on whether it is a meaningful concept (Hughes and Diaz, 1994; Bradley et al., 2003; Osborn and Briffa, 2005). 5. Climate variability and climate change The instrumental record and natural archives of climate over the past two millennia show Western North America to be a region in which robust spatiotemporal patterns of interannual to multidecadal and even multicentury variability are seen in hydroclimate. These time scales are relevant to many ecologic and geomorphic processes, and hence must be considered when trying to understand desertification and the associated changes in the availability of ecosystem services. Recent changes seem to be attributable to anthropogenic climate change, acknowledging the context of natural variability (Barnett et al., 2008). What might happen in the next few decades? It has been pointed out that the increased frequency and persistence of La Niñalike conditions seen before the mid-second millennium AD can be forced by anomalous heating over the Pacific (Mann et al., 2005; Cook et al., 2004; Graham et al., 2007; Seager et al., 2007a). Increased radiative forcing as a result of anthropogenic climate change could have a similar effect, modified in some important respects, including higher temperatures (Seager et al., 2007b). The result would be a drier climate for much of the North American continent. This conclusion is supported by the ensemble of climate model results, noting that the models somewhat underestimate the extent of extreme water deficit conditions for the instrumental period (Fig. 5). It follows from this that they may underestimate the severity of conditions in the mid-21st century. Even so, they project a dramatic extension of the area susceptible to land degradation. Seager et al. (2007b, page 3) write that “The most severe future droughts will still occur during persistent La Niña events but they will be worse than any since the Medieval period because the La Niña conditions will be perturbing a base state that is drier than any experienced recently”. There is ample evidence that the patterns of hydroclimatic variability revealed by paleoclimatic records in Western North America have had substantial impacts on the components of ecosystems on which modern ecosystem services depend. For example, Swetnam and Betancourt (1998) present extensive evidence for ecological responses to climate variability in this region on time scales from annual to decadal and on spatial scales from less than 100 km2 to 1,000,000 km2. The responses they review include regionally synchronized fires, insect outbreaks and pulses in tree demography. Biondi et al. (2000) examine one decadal-scale sequence in detail, the drought/pluvial sequence centered on AD 1600. They
116
M.K. Hughes, H.F. Diaz / Global and Planetary Change 64 (2008) 111–118
Fig. 5. Top panel—observed annual precipitation minus annual potential evapotranspiration (mm.) (P-PET) for the conterminous United States for the period AD 1895–2005. Middle panel: P-PET mm for AD 1895–2005 ensemble model results from 19 climate models for 42 A1B simulations prepared for the IPCC Fourth Assessment Report. Bottom panel: projected percentage change by mid-21st century from mean conditions of AD 1895–2005. Courtesy of Marty Hoerling, NOAA ESRL.
interpret changes in sediment input to the Santa Barbara Basin as indicating that the tree-ring reconstructed drought of the late 16th century “disrupted vegetation cover over large areas, allowing for major erosion during floods in the following years”, namely the first decades of the 17th century. Similarly, Graham et al. (2007) document major ecosystem changes coincident and consistent with the major climatic regime shift they propose for the middle of the second millennium AD. These changes include increased fire as recorded by diverse records, including charcoal in lacustrine sediments and dendrochronologically-dated fire scars in trees, salinity in San Francisco Bay, and vegetation composition also from sediments.
Graham et al. (2007) also noted the coincidence of drought and mobilization of dune fields covering a significant part of the Great Plains, consistent with changing salinity in lakes in this region. Note that, when stabilized, these areas provide a major and extensive rangeland resource, particularly in the Nebraska Sand Hills referred to by Muhs et al. (1997). 6. Final thoughts Much of W. N. America has recently been in a major multi-year drought, of a scale and intensity that has occurred several times in the
M.K. Hughes, H.F. Diaz / Global and Planetary Change 64 (2008) 111–118
last 2000 years, especially before the middle of the second millennium AD. Were one of these naturally occurring multi-decadal drought episodes to recur, the consequences for the availability of several important ecosystem services would be extreme. Consider the results of a multidecade interval with 35% deficiency in precipitation for natural ecosystems, rain-fed agriculture, water supply and ground water recharge in the rapidly growing American West. Ensembles of runs of forced climate system models suggest the next 50 years will bring much more extensive and intense drought in the continental interior of North America. “Desert climate” could invade much of the Great Plains, with its associated effects on vegetation and soils, and challenges to society and ecosystems. Appropriate responses will be made harder to define by the character of hydroclimatic variability in such semi-arid and arid regions. Consider, for example, the effects on decision making of a decadalscale “pluvial” (period of unusually high precipitation) in a dry region superimposed on a multidecadal drying trend. Such an event would surely make it more difficult to promote the changes in policy, behavior and engineering needed for life in a drier climate. Although it may presently be difficult to model regional climate change reliably, it should be noted that the explanations available so far for decadal to century-scale hydroclimatic variability in recent centuries are rather similar to the mechanisms proposed for nearterm (next few decades) regional climate change. Writing in the context of water resources, Milly et al. (2008, page 573) describe stationarity as a fundamental idea in the training and practice of water-resource engineering. It is the assumption that “natural systems fluctuate within an unchanging envelope of variability”, and that an adequate estimate of these parameters may be obtained from the instrumental record. They note that natural variability has usually been discounted, on the basis that it has been sufficiently small to be disregarded for engineering and policy purposes. It may be that similar, less formally expressed, assumptions underlie ways of thinking about the interactions between climate and ecological systems. We argue that in both cases, such disregard of the scale of natural variability, especially in drylands, is misguided. Milly et al. (2008) urge the abandonment of the assumption of stationarity in light of recent and projected climate change. We argue that the record of hydroclimatic variability in Western North America over the past two millennia shows the stationarity assumption to have been flawed from the beginning. Interactions between people and the land take place on these same time scales, interannual to multidecadal. The trajectory followed by the supply of ecosystem services will be contingent not only on the genotypes available and the antecedent soil, economic and social conditions. The critical features of climate on which patterns of plant growth and water supply depend may vary sharply during and between human generations, resulting in very different experiences and hence, expectations. Acknowledgements MKH was in receipt of a CIRES Visiting Fellowship at the University of Colorado, Boulder, when this paper was prepared. His contribution was also supported by grant #ATM- 0551986 from the Paleoclimatology program of the U.S. National Science Foundation. He is grateful to the organizers of the 2007 Wengen Workshop for the opportunity to participate. We are grateful to Marty Hoerling of NOAA/ESRL for Fig. 5, and to Jon Eischeid for extensive assistance. References Barnett, T.P., Pierce, D.W., Hidalgo, H.G., Bonfils, C., Santer, B.D., Das, T., Bala, G., Wood, A.W., Nozawa, T., Mirin, A.A., Cayan, D.R., Dettinger, M.D., 2008. Human-induced changes in the hydrology of the Western United States. Science 319, 1080–10833. Benson, L., Lund, S., Paillet, F., Kashgarian, M., Smoot, J., Mensing, S., Dibb, J., 1999. A 2800-yr history of oscillations in surface-water supply to the Central Valley and to the Bay Area of Northern California. Abstracts of the Fall 1999 AGU Meeting. EOS.
117
Biondi, F., Isaacs, C., Hughes, M.K., Cayan, D.R., Berger, W.H., 2000. The near-1600 dry/ wet knockout: linking terrestrial and near-shore ecosystems. Proceedings of the Twenty-fourth Annual Climate Diagnostics and Prediction Workshop. November 5– 9, 1999, Tucson, AZ. US Department of Commerce, NOAA, pp. 76–79. Bradley, R.S., Diaz, H.F., Kiladis, G.N., Eischeid, J.K., 1987. ENSO signal in continental temperatures and precipitation records. Nature 327, 497–501. Bradley, R.S., Hughes, M.K., Diaz, H.F., 2003. Climate in Medieval times. Science 302, 404–405. Brown, P.B., Hughes, M.K., Swetnam, T.W., Caprio, A.R., 1992. Giant sequoia ring-width chronologies from the central Sierra Nevada. Tree-Ring Bulletin 52, 1–14. Cayan, D.R., Dettinger, M.D., Diaz, H.F., Graham, N.E., 1998. Decadal variability of precipitation over Western North America. Journal of Climate 11, 3148–3166. Cayan, D.R., Kammerdiener, S.A., Dettinger, M.D., Caprio, J.M., Peterson, D.H., 2001. Changes in the onset of spring in the Western United States. Bulletin of the American Meteorological Society 82, 399–415. Charney, J., 1975. Dynamics of deserts and drought in the Sahel. The physical basis of climate and climate modelling. World Meteorological Organization, Geneva, Switzerland, pp. 171–175. Cook, E.R., Meko, D.M., Stahle, D.W., Cleaveland, M.K., 1999. Drought reconstructions for the continental United States. Journal of Climate 12, 1145–1162. Cook, E.R., Woodhouse, C.A., Eakin, C.M., Meko, D.M., Stahle, D.W., 2004. Long-term aridity changes in the western United States. Science 306, 1015–1018. Daly, C., Neilson, R.P., Phillips, D.L., 1994. A statistical topographic model for mapping climatological precipitation over mountainous terrain. Journal of Climate and Applied Methodology 33, 140–158. Daly, C., Gibson, W.P., Taylor, G.H., Johnson, G.L., Pasteris, P., 2002. A knowledge-based approach to the statistical mapping of climate. Climate Research 22, 99–113. Dettinger, M.D., Cayan, D.R., Diaz, H.F., Meko, D.M., 1998. North–South precipitation patterns in western North America on interannual-to-decadal time scales. Journal of Climate 11, 3095–3111. Diaz, H.F., 1983. Some aspects of major dry and wet periods in the contiguous United States, 1895–1981. Journal of Climate and Applied Meteorology 22, 3–16. Diaz, H.F., Eischeid, J.K., 2007. Disappearing “Alpine Tundra” Köppen climatic type in the Western United States. Geophysical Research Letters 34, L18707. doi:10.1029/ 2007GL031253. Fritz, S.C., Ito, E., Yu, Z., Laird, K., Engstrom, D.R., 2000. Hydrologic variation in the Northern Great Plains during the last two millennia. Quaternary Research 53, 175–184. Gershunov, A., Barnett, T.P., 1998. Interdecadal modulation of ENSO teleconnections. Bulletin of the American Meteorological Society 79 (12), 2715–2725. Graham, N.E., Hughes, M.K., 2007. Reconstructing the Mediaeval low stands of Mono Lake, Sierra Nevada, California. The Holocene 17, 1197–1210. Graham, N.E., Hughes, M.K., Ammann, C.M., Cobb, K.M., Hoerling, M.P., Xu, T., Kennett, D.J., Kennett, J.P., Rein, B., Stott, L., Wigand, P.E., 2007. Tropical Pacific-mid-latitude teleconnections in Medieval times. Climatic Change. doi:10.1007/s10584-007-9239-2. Graumlich, L.J., 1993. A 1000-year record of temperature and precipitation in the Sierra Nevada. Quaternary Research 39, 249–255. Graybill, D.A., Funkhouser, G., 2000. Dendroclimatic reconstructions during the past millennium in the southern Sierra Nevada and Owens Valley, California. In: Lavenberg, R. (Ed.), Southern California Climate: Trends and Extremes of the Past 2000 years. Natural History Museum of Los Angeles County, Los Angeles, CA. Hughes, M.K., 2002. Dendrochronology in climatology—the state of the art. Dendrochronologia 20, 95–116. Hughes, M.K., Brown, P.M., 1992. Drought frequency in central California since 101 B.C. recorded in giant sequoia tree rings. Climate Dynamics 6, 161–167. Hughes, M.K., Diaz, H.F., 1994. Was there a Medieval Warm Period, and if so, where and when? Climatic Change 26, 109–142. Hughes, M.K., Graumlich, L.J., 1996. Multimillennial dendroclimatic records from western North America. In: Bradley, R.S., Jones, P.D., Jouzel, J. (Eds.), Climatic Variations and Forcing Mechanisms of the Last 2000 years. Springer Verlag, Berlin, pp. 109–124. Hughes, M.K., Funkhouser, G., 1998. Extremes of moisture availability reconstructed from tree rings for recent millennia in the Great Basin of Western North America. In: Beniston, M., Innes, J.L. (Eds.), The Impacts of Climate Variability on Forests. Springer, Berlin, pp. 99–107. Hughes, M.K., Touchan, R., Brown, P., 1996. A multimillennial network of giant sequoia chronologies for dendroclimatology. In: Dean, J., Meko, D.M., Swetnam, T.W. (Eds.), Tree Rings, Environment and Humanity, Proceedings of the International Conference. Arizona, Radiocarbon, Tucson, pp. 225–234. Kennett, D.J., Kennett, J.P., 2000. Competitive and cooperative responses to climatic instability in coastal southern California. American Antiquity 65, 379–395. Kiladis, G.N., Diaz, H.F., 1989. Global climatic anomalies associated with extremes in the Southern Oscillation. Journal of Climate 2, 1069–1090. Kottek, M., Grieser, J., Beck, C., Rudolf, B., Rubel, F., 2006. World map of the KöppenGeiger climate classification updated. Meteorologische Zeitschrift 15, 259–263. LaMarche, V.C., 1974. Paleoclimatic inferences from long tree-ring records. Science 183, 1043–1088. Mann, M.E., Cane, M.A., Zebiak, S.E., Clement, A., 2005. Volcanic and solar forcing of the Tropical Pacific over the past 1000 years. Journal of Climate 18, 447–456. Mantua, N.J., Hare, S.R., Zhang, Y., Wallace, J.M., Francis, R.C., 1997. A Pacific interdecadal climate oscillation with impacts on salmon production. Bulletin of the American Meteorological Society 78, 1069–1079. McCabe, G.J., Palecki, M.A., Betancourt, J.L., 2004. Pacific and Atlantic Ocean influences on multidecadal drought frequency in the United States. Proceedings of the National Academy of Sciences 101 (12), 4136–4141. Meko, D.M., Cook, E.R., Stahle, D.W., Stockton, C.W., Hughes, M.K., 1993. Spatial patterns of tree-growth anomalies in the United States and southeastern Canada. Journal of Climate 6, 1773–1786.
118
M.K. Hughes, H.F. Diaz / Global and Planetary Change 64 (2008) 111–118
Meko, D.M., Woodhouse, C.A., Baisan, C.A., Knight, T., Lukas, J.J., Hughes, M.K., Salzer, M.W., 2007. Medieval drought in the upper Colorado River Basin. Geophysical Research Letters 34. doi:10.1029/2007GL029988. Millennium Ecosystem Assessment, 2005. Ecosystems and Human Well-being: Desertification Synthesis. World Resources Institute, Washington, DC. Milly, P.C.D., Betancourt, J., Falkenmark, M., Hirsch, R.M., Kundzewicz, Z.W., Lettenmaier, D.P., Stouffer, R.J., 2008. Stationarity is dead: whither water management? Science 319, 573–574. Muhs, D.R., Stafford, T.W., Swinehart, J.B., Cowherd, S.A., Bush, R.F., Madole, R.F., Maat, P.B., 1997. Late Holocene eolian activity in the mineralogically mature Nebraska Sand Hills. Quaternary Research 48, 162–176. Namias, J., 1978. Multiple causes of the North American abnormal winter 1976–77. Monthly Weather Review 106, 279–295. Ni, F., Cavazos, T., Hughes, M.K., Comrie, A.C., Funkhouser, G., 2002. Cool season precipitation in the Southwestern United States since AD 1000: comparison of linear and nonlinear techniques for reconstruction. International Journal of Climatology 22, 1645–1662. Osborn, T., Briffa, K., 2005. The spatial extent of 20th-century warmth in the context of the past 1200 years. Science 311, 841–844. Ropelewski, C.F., Halpert, M.S., 1987. Global and regional sale precipitation patterns associated with the El Niño-Southern oscillation. Monthly Weather Review 115, 1606–1626.
Seager, R., Graham, N., Herweijer, C., Gordon, A.L., Kushnir, Y., Cook, E., 2007a. Blueprints for Medieval hydroclimate. Quaternary Science Reviews 26, 2322–2336. Seager, R., Ting, M., Held, I., Kushnir, Y., Lu, J., Vecchi, G., Huang, H.-P., Harnik, N., Leetmaa, A., Lau, N.-C., Li, C., Velez, J., Naik, N., 2007b. Model projections of an imminent transition to a more arid climate in Southwestern North America. Science 316 (5828), 1181–1184. Stahle, D.W., Cook, E.R., Cleaveland, M.K., Therrell, M.D., Meko, D.M., Grissino-Mayer, H.D., Watson, E., Luckman, B., 2000. Tree-ring Data Document 16th Century Megadrought Over North America, 81. EOS, p. 121,125. Stine, S., 1994. Extreme and persistent drought in California and Patagonia during mediaeval time. Nature 269, 546–549. Stockton, C.W., Jacoby, G.C., 1976. Long-term surface water supply and streamflow levels in the upper Colorado River basin. Lake Powell Research Project Bulletin 18 70pp. Stockton, C.W., Meko, D.M., 1983. Drought recurrence in the Great Plains as reconstructed from long-term tree-ring records. Journal of Climate and Applied Meteorology 2, 17–29. Swetnam, T.W., Betancourt, J.L., 1998. Mesoscale disturbance and ecological response to decadal climate variability in the American Southwest. Journal of Climate 11, 3128–3147. Woodhouse, C.A., 2003. Droughts of the past, implications for the future? In: Smith, S.L. (Ed.), The future of the Southern Plains. University of Oklahoma Press, Norman, OK, pp. 95–113.