« PreviousContinue »
Observed Climate Variability and Change
8 respectively). Mann et al. (1998) reconstructed global patters Historical documentary data are valuable sources of of annual surface temperature several centuries back in time. They information about past climate (e.g., Brown and Issar, 1998; calibrated a combined terrestrial (aree ring, ice core and historical Bradley, 1999). However, their use requires great care, as such documentary indicator) and marine (coral) mulci-proxy climate documents may be biased towards describing only the more network against dominant pattems of 20th century global surface extreme events, and are, in certain cases, prone to the use of temperature. Averaging the reconstructed temperature patterns inconsistent language between different writers and different over the far more data-rich Northern Hemisphere half of the epochs, and to errors in dating. As for all proxy information, global domain, they estimated the Northern Hemisphere mean historical documents require careful calibration and verification temperature back to AD 1400, a reconstruction which had signifagainst modem instrumental data. Two areas particularly strong icant skill in independent cross-validation tests. Self-consistent in historical documents describing climate are Europe and China. estimates were also made of the uncertainties. This work has now In Europe, attempts have been made to extend long climate series been extended back to AD 1000 (Figure 2.20, based on Mann er back in time using a combination of documentary evidence and al., 1999). The uncertainties (the shaded region in Figure 2.20) fragmentary instrumental records (e.g., Pfister, 1995; Pfister et expand considerably in earlier centuries because of the sparse al., 1998). Additional information about past climate change has network of proxy data. Taking into account these substantial also been obtained purely from documentary records in Europe uncertainties, Mann et al. (1999) concluded that the 1990s were (e.g., Martin-Vide and Barriendos, 1995; Brázdil, 1996; Pfister et likely to have been the warmest decade, and 1998 the warmest al., 1996, 1998, 1999; Pfister and Brázdil, 1999; Rodrigo el year, of the past millennium for at least the Northern Hemisphere. 1999). In China, regional instrumental temperature series have Jones et al. (1998) came to a similar conclusion from largely been extended back over much of the past milleonium using independent data and an entirely independent methodology. documentary data combined with inferences from ice cores and Crowley and Lowery (2000) reached the similar conclusion that tree rings (Wang et al., 1998a, 1998b; Wang and Gong, 2000). medieval temperatures were no warmer than mid-20th century
temperatures. Borehole data (Pollack et al., 1998) independently Mountain glacier moraines
support this conclusion for the past 500 years although, as The position of moraines or till left behind by receding glaciers discussed earlier (Section 126.96.36.199), detailed interpretations can provide information on the advances (and, less accurately, the comparison with long-term trends from such of such data are retreats) of mountain glaciers. Owing to the complex balance perilous owing to loss of temporal resolution back in time. between local changes in melting and ice accumulation, and the The largely independent multi-proxy Northern Hemisphere effects of topography which influence mountain glaciers (see temperature reconstructions of Jones et al. (1998) and Maon et Section 188.8.131.52), it is difficult to reconstruci regional (as opposed al. (1999) are compared in Figure 2.21, together with an to global) climate changes from the extent of mountain glaciers independent (extra-tropical, warm-season) Northern Hemisphere alone (Oerlemans, 1989). For example, both increased winter temperature estimate by Briffa (2000) based on tree-ring density precipitation (through greater accumulation) and lower summer data. The estimated uncertainties shown are those for the temperatures (through decreased melting or “ablation”) can lead smoothed Mann et al. series. Significant differences between the to more positive glacial mass balances. The inertia of large three reconstructions are evident during the 17th and early 19th glaciers dictates that they respond to climate change relatively centuries where either the Briffa et al. or Jones et al. series lie slowly, with delays of decades or occasionally centuries. For outside the estimated uncertainties in the Mann et al. series. smaller, fast moving glaciers in regions where precipitation and Much of these differences appear to result from the different accumulation are moderate, temperature changes are usually the latitudinal and seasonal emphases of the temperature estimates. dominant factor influencing mountain glacier masses and This conclusion is supported by the observation that the Mann et lengths. Here glacier moraine evidence in combination with other al. hemispheric temperature average, when restricted to just the lines of evidence can provide reliable information on past extra-tropical (30 to 70°N band) region of the Northern regional temperature changes (Salinger, 1995; Holzhauser and Hemisphere, shows greater similarity in its trend over the past Zumbühl, 1996; Raper et al., 1996; Salinger et al., 1996). few centuries to the Jones et al. reconstruction. The differences
between these reconstructions emphasise the importance of 184.108.40.206 Multi-prory synthesis of recent temperature change regional and seasonal variations in climate change. These are Since the SAR there have been several attempts to combine discussed in the next section. various types of high-resolution proxy climate indicators to create large-scale palaeoclimate reconstructions that build on earlier
2.3.3 Was there a "Little Ice Age" and a "Medieval Warm work by e.g., Bradley and Jones (1993); Hughes and Diaz (1994)
Period”? and Mann et al. (1995). Overpeck et al. (1997) and Fisher (1997) have sought to combine information from ice cores, varved lake The terms "Little Ice Age” and “Medieval Warm Period” have sediment cores, and tree rings to reconstruct high latitude climate been used to describe two past climate epochs in Europe and trends for past centuries. Jones et al. (1998) estimated extra- neighbouring regions during roughly the 17th to 19th and 11th to tropical Northem and Southern Hemisphere warm-season temper- 14th centuries, respectively. The timing, however, of these cold ature changes during the past millennium using a sparse set of and warm periods has recently been demonstrated to vary extra-tropical warm-season temperature proxy indicators (10 and geographically over the globe in a considerable way (Bradley and
Observed Climate Variability and Change
Northern Hemisphere anomaly (°C)
relative to 1961 to 1990
Year Figure 2.20: Millennial Northern Hemisphere (NH) temperature reconstruction (blue) and instrumental data (red) from AD 1000 to 1999, adapled from Mann et al. (1999). Smoother version of NH series (black), linear trend from AD 1000 to 1850 (purple-dashed) and two standard error limits (grey shaded) are shown.
Northern Hemisphere anomaly (°C)
relative to 1961 to 1990
Year Figure 2.21: Comparison of warm-season (Jones et al., 1998) and annual mean (Mann et al., 1998, 1999) multi-proxy-based and warm season tree-ring-based (Briffa, 2000) millennial Northern Hemisphere temperature reconstructions. The recent instrumental annual mean Northern Hemisphere temperature record to 1999 is shown for comparison. Also shown is an extra-tropical sampling of the Mann et al. (1999) temperature patiem reconstructions more directly comparable in its latitudinal sampling to the Jones et al. series. The self-consistently estimated two standard error limits (shaded region) for the smoothed Mann et al. (1999) series are shown. The horizontal zero line denotes the 1961 to 1990 reference period mean temperature. All series were smoothed with a 40-year Hamming-weights lowpass filter, with boundary constraints imposed by padding the series with its mean values during the first and last 25 years.
Observed Climate Variability and Change
Jones, 1993; Hughes and Diaz, 1994; Crowley and Lowery, 2000). beginning around 1568, and with 1573 the first unusually cold Evidence from mountain glaciers does suggest increased glacia- summer (Pfister, 1995). tion in a number of widely spread regions outside Europe prior to The evidence for temperature changes in past centuries in the 20th century, including Alaska, New Zealand and Patagonia the Southern Hemisphere is quite sparse. What evidence is (Grove and Switsur, 1994). However, the timing of maximum available at the hemispheric scale for summer (Jones et al., 1998) glacial advances in these regions differs considerably, suggesting and annual mean conditions (Mann et al., 2000b) suggests that they may represent largely independent regional climate markedly different behaviour from the Northern Hemisphere. changes, not a globally-synchronous increased glaciation (see The only obvious similarity is the unprecedented warmth of the Bradley, 1999). Thus current evidence does not support globally late 20th century. Speleothem evidence (isotopic evidence from synchronous periods of anomalous cold or warmth over this calcite deposition in stalagmites and stalactites) from South timeframe, and the conventional terms of “Little Ice Age" and Africa indicates anomalously cold conditions only prior to the "Medieval Warm Period” appear to have limited utility in 19th century, while speleothem (records derived from analysing describing trends in hemispheric or global mean temperature stalagmites and stalagtites) and glacier evidence from the changes in past centuries. With the more widespread proxy data Southern Alps of New Zealand suggests cold conditions during and multi-proxy reconstructions of temperature change now the mid-17th and mid-19th centuries (Salinger, 1995). available, the spatial and temporal character of these putative Dendroclimatic evidence from nearby Tasmania (Cook et al., climate epochs can be reassessed.
2000) shows no evidence of unusual coldness at these times. Mann et al. (1998) and Jones et al. (1998) support the idea Differences in the seasons most represented by this proxy that the 15th to 19th centuries were the coldest of the millennium information prevent a more direct comparison. over the Northern Hemisphere overall. However, viewed As with the "Little Ice Age", the posited “Medieval Warm hemispherically, the “Little Ice Age" can only be considered as a Period" appears to have been less distinct, more moderate in modest cooling of the Northern Hemisphere during this period of amplitude, and somewhat different in timing at the hemispheric less than 1°C relative to late 20th century levels (Bradley and scale than is typically inferred for the conventionally-defined Jones, 1993: Jones et al., 1998; Mann et al., 1998; 1999; Crowley European epoch. The Northern Hemisphere mean temperature and Lowery, 2000). Cold conditions appear, however, to have been estimates of Jones et al. (1998), Mann et al. (1999), and Crowley considerably more pronounced in particular regions. Such regional and Lowery (2000) show temperatures from the 11th to 14th variability can be understood in part as reflecting accompanying centuries to be about 0.2°C warmer than those from the 15th to changes in atmospheric circulation. The "Little Ice Age" appears 19th centuries, but rather below mid-20th century temperatures. to have been most clearly expressed in the North Atlantic region as The long-term hemispheric trend is best described as a modest altered patterns of atmospheric circulation (O'Brien et al., 1995). and irregular cooling from AD 1000 to around 1850 to 1900, Unusually cold, dry winters in central Europe (e.g., 1 to 2°C below followed by an abrupt 20th century warming. Regional evidence normal during the late 17th century) were very likely to have been is, however, quite variable. Crowley and Lowery (2000) show associated with more frequent flows of continental air from the that western Greenland exhibited anomalous warmth locally only north-east (Wanner et al., 1995; Pfister, 1999). Such conditions are around AD 1000 (and to a lesser extent, around AD 1400), with consistent (Luterbacher et al., 1999) with the negative or enhanced quite cold conditions during the latter part of the 19th century, easterly wind phase of the NAO (Sections 220.127.116.11 and 2.6.5), while Scandinavian summer temperatures appeared relatively which implies both warm and cold anomalies over different warm only during the 11th and early 12th centuries. Crowley and regions in the North Atlantic sector. Such strong influences on Lowery (2000) find no evidence for warmth in the tropics. European temperature demonstrate the difficulty in extrapolating Regional evidence for medieval warmth elsewhere in the the sparse early information about European climate change to the Northern Hemisphere is so variable that eastem, yet not western, hemispheric, let alone global, scale. While past changes in the China appears to have been warm by 20th century standards from NAO have likely had an influence in eastern North America, the 9th to 13th centuries. The 12th and 14th centuries appear to changes in the El Niño phenomenon (see also Section 2.6), are have been mainly cold in China (Wang et al., 1998a,b; Wang and likely to have had a particularly significant influence on regional Gong, 2000). The restricted evidence from the Southern temperature patterns over North America.
Hemisphere, e.g., the Tasmanian tree-ring temperature The hemispherically averaged coldness of the 17th century reconstruction of Cook et al. (1999), shows no evidence for a largely reflected cold conditions in Eurasia, while cold distinct Medieval Warm Period. hemispheric conditions in the 19th century were more associated Medieval warmth appears, in large part, to have been with cold conditions in North America (Jones et al., 1998; Mann restricted to areas in and neighbouring the North Atlantic. This et al., 2000b). So, while the coldest decades of the 19th century may implicate the role of ocean circulation-related climate appear to have been approximately 0.6 to 0.7°C colder than the variability. The Bermuda rise sediment record of Keigwin (1996) latter decades of the 20th century in the hemispheric mean (Mann suggests warm medieval conditions and cold 17th to 19th century el al., 1998), the coldest decades for the North American conditions in the Sargasso Sea of the tropical North Atlantic. A continent were closer to 1.5°C colder (Mann et al., 2000b). In sediment record just south of Newfoundland (Keigwin and addition, the timing of peak coldness was often specific to partic- Pickart, 1999), in contrast, indicates cold medieval and warm 16th ular seasons. In Switzerland, for example, the first particularly to 19th century upper ocean temperatures. Keigwin and Pickan cold winters appear to have been in the 1500s, with cold springs (1999) suggest that these temperature contrasts were associated
Observed Climate Variability and Change
vith changes in ocean currents in the North Atlantic. They argue have been the warmest decade and year, respectively, in the chat the "Little Ice Age” and “Medieval Warm Period" in the Northern Hemisphere. Independent estimates of hemispheric and Atlantic region may in large measure reflect century-scale global ground temperature trends over the past five centuries from changes in the North Atlantic Oscillation (see Section 2.6). Such sub-surface information contained in borehole data confirm the regional changes in oceanic and atmospheric processes, which are conclusion that late 20th century warmth is anomalous in a longalso relevant to the natural variability of the climate on millennial term context. Decreasing temporal resolution back in time of these and longer time-scales (see Section 2.4.2), are greatly diminished estimates and potential complications in inferring surface air or absent in their influence on hemispheric or global mean temperature trends from sub-surface ground temperature measuretemperatures.
ments precludes, however, a meaningful direct comparison of the
borehole estimates with high-resolution temperature estimates 2.3.4 Volcanic and Solar Effects in the Recent Record
based on other proxy climate data. Because less data are available,
less is known about annual averages prior to 1,000 years before the Recent studies comparing reconstructions of surface tempera- present and for conditions prevailing in most of the Southem ture and natural (solar and volcanic) radiative forcing (e.g., Lean Hemisphere prior to 1861. et al., 1995; Crowley and Kim, 1996, 1999; Overpeck et al., 1997; Mann et al., 1998; Damon and Peristykh, 1999; Free and
2.4 How Rapidly did Climate Change in the Distant Past? Robock, 1999; Waple et al., 2005) suggest that a combination of solar and volcanic influences have affected large-scale tempera
2.4.1 Background ture in past centuries. The primary features of the Northern Hemisphere mean annual temperature histories of Mann et al. Only during the 1980s was the possibility of rapid climatic (1999a) and Crowley and Lowery (2000) from AD 1000 to 1900 changes occurring at the time-scale of human life more or less have been largely reproduced : based on experiments using an fully recognised, largely due to the Greenland ice core drilled at Energy Balance Model forced by estimates of these natural Dye 3 in Southem Greenland (Dansgaard et al., 1982, 1989). A radiative forcings (Crowley, 2000; Mann, 2000) making the possible link between such events and the mode of operation of the argument that the “Little Ice Age” and “Medieval Warm Period", ocean was then subsequently suggested (Oeschger et al., 1984; at the hemispheric mean scale, are consistent with estimates of Broecker et al., 1985; see Broecker, 1997, for a recent review). naturally-forced climate variability. Several studies indicate that The SAR reviewed the evidence of such changes since the peak of the combined effect of these influences has contributed a small the last inter-glacial period about 120 ky BP (thousands of years component to the warming of the 20th century. Most of these Before Present). It concluded that: (1) large and rapid climatic studies isolate greenbouse radiative forcing as being dominant changes occurred during the last Ice Age and during the transition during late 20th century warming (see Crowley, 2000). This towards the present Holocene; (2) temperatures were far less argues against a close empirical relationship between certain variable during this latter period; and (3) suggestions that rapid sun-climate parameters and large-scale temperature that has changes may have also occurred during the last inter-glacial been claimed for the 20th century (Hoyt and Schatten, 1997). required confirmation. The reader is referred to Chapter 6 for a detailed discussion of These changes are now best documented from ice core, these radiative forcings, and to Chapter 12 for comparisons of deep-sea sediment and continental records. Complementary and observed and model simulations of recent climate change. generally discontinuous information comes from coral and lake
level data. The time-scale for the Pleistocene deep-sea core 2.3.5 Summary
record is based on the orbitally tuned oxygen isotope record
from marine sediments (Martinson et al., 1987), constrained by Since the SAR there have been considerable advances in our two radiometrically dated horizons, the peak of the last interknowledge of temperature change over the last millennium. It is glacial (about 124 ky BP) and the Brunhes Matuyama reversal likely that temperatures were relatively warm in the Northern of the Earth's magnetic field at about 780 ky BP. “C-dating is Hemisphere as a whole during the earlier centuries of the millen- also used in the upper 50 ky BP; the result is a deep-sea core nium, but it is much less likely that a globally-synchronous, well chronology believed to be accurate to within a few per cent for defined interval of “Medieval warmth” existed, comparable to the the last million years. "C-dating is also used for dating near global warmth of the late 20th century. Marked warmth continental records as well as the counting of annual layers in seems to have been confined to Europe and regions neighbouring tree rings and varved lake records, whereas ice-core-chronthe North Atlantic. Relatively colder hemispheric or global-scale ologies are obtained by combining layer counting, glaciological conditions did appear to set in after about AD 1400 and persist models and comparison with other dated records. The use of through the 19th century, but peak coldness is observed during globally representative records, such as changes in continental substantially different epochs in different regions. By contrast, the ice volume recorded in the isotopic composition of deep-sea warming of the 20th century has had a much more convincing sediments, or changes in atmospheric composition recorded in global signature (see Figure 2.9). This is consistent with the air bubbles trapped in ice cores, now allow such local records to palaeoclimate evidence that the rate and magnitude of global or be put into a global perspective. Studies still largely focus on the hemispheric surface 20th century warming is likely to have been more recent glacial-interglacial cycle (the last 120 to 130 ky). the largest of the millennium, with the 1990s and 1998 likely to Table 2.4 is a guide to terminology.
Surface Temperature Reconstructions for the Last 2,000 Years
Committee on Surface Temperature Reconstructions for the Last 2,000 Years
Board on Atmospheric Sciences and Climate
Division on Earth and Life Studies
NATIONAL RESEARCH COUNCIL
OF THE NATIONAL ACADEMIES
THE NATIONAL ACADEMIES PRESS