The Medieval Warm Period in the Arctic

Since the IPCC AR5 is again making news in talking about the Medieval Warm Period, this review from Craig Idso at CO2Science.org seems appropriate. -Anthony

Medieval Warm Period (Arctic) — Summary

This review begins with the study of Dahl-Jensen et al. (1998), who used temperature measurements from two Greenland Ice Sheet boreholes to reconstruct the temperature history of this portion of the earth over the past 50,000 years. Their data indicated that after the termination of the glacial period, temperatures steadily rose to a maximum of 2.5°C warmer than at present during the Holocene Climatic Optimum (4,000 to 7,000 years ago). The Medieval Warm Period (MWP) and Little Ice Age (LIA) were also observed in the record, with temperatures 1°C warmer and 0.5-0.7°C cooler than at the time of their writing, respectively. After the Little Ice Age, they report that temperatures once again rose, but that they had “decreased during the last decades,” thereby indicating that the MWP in this part of the Arctic was significantly warmer than it was just before the turn of the century.

Wagner and Melles (2001) also worked on Greenland, where they extracted a 3.5-m-long sediment core from a lake (Raffels So) on an island (Raffles O) located just off Liverpool Land on the east coast of Greenland, which they analyzed for a number of properties related to the past presence of seabirds there, obtaining a 10,000-year record that tells much about the region’s climatic history. Key to the study were biogeochemical data, which, in the words of the two researchers, reflect “variations in seabird breeding colonies in the catchment which influence nutrient and cadmium supply to the lake.”

These data revealed sharp increases in the values of the parameters they represented between about 1100 and 700 years before present (BP), indicative of the summer presence of significant numbers of seabirds during that “medieval warm period,” as Wagner and Melles described it, which had been preceded by a several-hundred-year period (the Dark Ages Cold Period) with little to no bird presence. And after that “medieval warm period,” their data suggested another absence of birds during what they called “a subsequent Little Ice Age,” which they said was “the coldest period since the early Holocene in East Greenland.”

The Raffels So data also showed signs of a resettlement of seabirds during the last century, as indicated by an increase of organic matter in the lake sediment and confirmed by bird counts. However, values of the most recent measurements of seabird numbers were not as great as those inferred for the earlier Medieval Warm Period, which result indicates that higher temperatures prevailed during much of the period from 1100 to 700 years BP than those that had been observed over the most recent hundred years.

A third Greenland study was conducted by Kaplan et al. (2002), who derived a climatic history of the Holocene by analyzing the physical-chemical properties of sediments obtained from a small lake in the southern sector of Greenland. This work revealed that the interval from 6000 to 3000 years BP was marked by warmth and stability, but that the climate cooled thereafter until its culmination in the Little Ice Age. From 1300-900 years BP, however, there was a partial amelioration during the Medieval Warm Period, which was associated with an approximate 1.5°C rise in temperature.

In a non-Greenland Arctic study, Jiang et al. (2002) analyzed diatom assemblages from a high-resolution core extracted from the seabed of the north Icelandic shelf, which led to their reconstruction of a 4600-year history of summer sea surface temperature at that location. Starting from a maximum value of about 8.1°C at 4400 years BP, the climate was found to have cooled fitfully for about 1700 years and then more consistently over the final 2700 years of the record. The most dramatic departure from this long-term decline was centered on about 850 years BP, during the Medieval Warm Period, when the temperature rose by more than 1°C above the line describing the long-term downward trend to effect an almost complete recovery from the colder temperatures of the Dark Ages Cold Period, after which temperatures continued their descent into the Little Ice Age, ending with a final most recent value of approximately 6.3°C. Hence, their data clearly showed that the Medieval Warm Period in this part of the Arctic was significantly warmer than it is there now.

Moving on, Moore et al. (2001) analyzed sediment cores from Donard Lake, Baffin Island, Canada, producing a 1240-year record of average summer temperatures for this Arctic region. Over the entire period from AD 750-1990, temperatures averaged 2.9°C. However, anomalously warm decades with summer temperatures as high as 4°C occurred around AD 1000 and 1100, while at the beginning of the 13th century, Donard Lake witnessed “one of the largest climatic transitions in over a millennium,” as “average summer temperatures rose rapidly by nearly 2°C from 1195-1220 AD, ending in the warmest decade in the record” with temperatures near 4.5°C.

This rapid warming of the 13th century was followed by a period of extended warmth that lasted until an abrupt cooling event occurred around 1375, which made the following decade one of the coldest in the record. This event signaled the onset of the Little Ice Age, which lasted for 400 years, until a gradual warming trend began around 1800, which was followed by a dramatic cooling event in 1900 that brought temperatures back to levels similar to those of the Little Ice Age. This cold regime lasted until about 1950, whereupon temperatures warmed for about two decades but then tended downwards again, all the way to the end of the record in 1990. Thus, in this part of the Arctic, the Medieval Warm Period was also warmer than it is there currently.

The following year, Grudd et al. (2002) assembled tree-ring widths from 880 living, dead, and subfossil northern Swedish pines into a continuous and precisely dated chronology covering the period 5407 BC to AD 1997. The strong association between these data and summer (June-August) mean temperatures of the last 129 years of the period then enabled them to produce a 7400-year history of summer mean temperature for northern Swedish Lapland. The most dependable portion of this record, based upon the number of trees that were sampled, consisted of the last two millennia, which Grudd et al. said “display features of century-timescale climatic variation known from other proxy and historical sources, including a warm ‘Roman’ period in the first centuries AD and a generally cold ‘Dark Ages’ climate from about AD 500 to about AD 900.” They also noted that “the warm period around AD 1000 may correspond to a so-called ‘Mediaeval Warm Period,’ known from a variety of historical sources and other proxy records.” Last of all, they stated that “the climatic deterioration in the twelfth century can be regarded as the starting point of a prolonged cold period that continued to the first decade of the twentieth century,” which “Little Ice Age,” in their words, is also “known from instrumental, historical and proxy records.” Going back even further in time, the tree-ring record displays several more of these relatively warmer and colder periods. And in a telling commentary on current climate-alarmist claims, they report that “the relatively warm conditions of the late twentieth century do not exceed those reconstructed for several earlier time intervals.” In fact, the warmth of many of the earlier warm intervals significantly exceeded the warmth of the late 20th century.

Seppa and Birks (2002) used a recently developed pollen-climate reconstruction model and a new pollen stratigraphy from Toskaljavri – a tree-line lake in the continental sector of northern Fenoscandia (located just above 69°N latitude) – to derive quantitative estimates of annual precipitation and July mean temperature. And as they described it, their reconstructions “agree with the traditional concept of a ‘Medieval Warm Period’ (MWP) and ‘Little Ice Age’ in the North Atlantic region (Dansgaard et al., 1975) and in northern Fennoscandia (Korhola et al., 2000).” In addition, they reported there was “a clear correlation between [their] MWP reconstruction and several records from Greenland ice cores,” and that “comparisons of a smoothed July temperature record from Toskaljavri with measured borehole temperatures of the GRIP and Dye 3 ice cores (Dahl-Jensen et al., 1998) and the ð¹⁸O record from the Crete ice core (Dansgaard et al., 1975) show the strong similarity in timing of the MWP between the records.” Last of all, they noted that “July temperature values during the Medieval Warm Period (ca. 1400-1000 cal yr B.P.) were ca. 0.8°C higher than at present,” where present means the last six decades of the 20th century.

Noting that temperature changes in high latitudes are (1) sensitive indicators of global temperature changes, and that they can (2) serve as a basis for verifying climate model calculations, Naurzbaev et al. (2002) developed a 2,427-year proxy temperature history for the part of the Taimyr Peninsula of northern Russia that lies between 70°30′ and 72°28′ North latitude, based on a study of ring-widths of living and preserved larch trees, while further noting that “it has been established that the main driver of tree-ring variability at the polar timber-line [where they conducted their study] is temperature (Vaganov et al., 1996; Briffa et al., 1998; Schweingruber and Briffa, 1996).” And in doing so, they found that “the warmest periods over the last two millennia in this region were clearly in the third [Roman Warm Period], tenth to twelfth [Medieval Warm Period] and during the twentieth [Current Warm Period] centuries.”

With respect to the second of these periods, they emphasize that “the warmth of the two centuries AD 1058-1157 and 950-1049 attests to the reality of relative mediaeval warmth in this region.” Their data also reveal three other important pieces of information: (1) the Roman and Medieval Warm Periods were both warmer than the Current Warm Period has been to date, (2) the “beginning of the end” of the Little Ice Age was somewhere in the vicinity of 1830, and (3) the Current Warm Period peaked somewhere in the vicinity of 1940.

All of these observations are at odds with what is portrayed in the thousand-year Northern Hemispheric hockeystick temperature history of Mann et al. (1998, 1999) and its thousand-year global extension developed by Mann and Jones (2003), wherein (1) the Current Warm Period is depicted as the warmest such era of the past two millennia, (2) recovery from the Little Ice Age does not begin until after 1910, and (3) the Current Warm Period experiences it highest temperatures in the latter part of the 20th century’s final decade.

Advancing two years closer to the present, Knudsen et al. (2004) documented climatic changes over the last 1200 years by means of high-resolution multi-proxy studies of benthic and planktonic foraminiferal assemblages, stable isotopes, and ice-rafted debris found in three sediment cores retrieved from the North Icelandic shelf. This work revealed that “the time period between 1200 and around 7-800 cal. (years) BP, including the Medieval Warm Period, was characterized by relatively high bottom and surface water temperatures,” after which “a general temperature decrease in the area marks the transition to … the Little Ice Age.” They also note that “minimum sea-surface temperatures were reached at around 350 cal. BP, when very cold conditions were indicated by several proxies.” Thereafter, they report that “a modern warming of surface waters … is not registered in the proxy data,” and that “there is no clear indication of warming of water masses in the area during the last decades,” even in sea surface temperatures measured over the period 1948-2002.

Fast-forwarding another two years,Grinsted et al. (2006) developed “a model of chemical fractionation in ice based on differing elution rates for pairs of ions … as a proxy for summer melt (1130-1990),” based on data obtained from a 121-meter-long ice core they extracted from the highest ice field in Svalbard (Lomonosovfonna: 78°51’53″N, 17°25’30″E), which was “validated against twentieth-century instrumental records and longer historical climate proxies.” This history indicated that “in the oldest part of the core (1130-1200), the washout indices [were] more than 4 times as high as those seen during the last century, indicating a high degree of runoff.” In addition, they said they had performed regular snow pit studies near the ice core site since 1997 (Virkkunen, 2004) and that “the very warm 2001 summer resulted in similar loss of ions and washout ratios as the earliest part of the core.” They then stated that “this suggests that the Medieval Warm Period in Svalbard summer conditions [was] as warm (or warmer) as present-day, consistent with the Northern Hemisphere temperature reconstruction of Moberg et al. (2005).” In addition, they concluded that “the degree of summer melt was significantly larger during the period 1130-1300 than in the 1990s,” which likewise suggests that a large portion of the Medieval Warm Period was significantly warmer than the peak warmth (1990s) of the Current Warm Period.

Moving ahead two more years, Besonen et al. (2008) derived thousand-year histories of varve thickness and sedimentation accumulation rate for Canada’s Lower Murray Lake (81°20’N, 69°30’W), which is typically covered for about eleven months of each year by ice that reaches a thickness of 1.5 to 2 meters at the end of each winter. With respect to these parameters, they write – citing seven other studies – that “field-work on other High Arctic lakes clearly indicates that sediment transport and varve thickness are related to temperatures during the short summer season that prevails in this region, and we have no reason to think that this is not the case for Lower Murray Lake.”

So what did they find? As the six scientists describe it, the story told by both the varve thickness and sediment accumulation rate histories of Lower Murray Lake is that “the twelfth and thirteenth centuries were relatively warm,” and in this regard we note their data indicate that Lower Murray Lake and its environs were often much warmer during this time period (AD 1080-1320) than they were at any point in the 20th century, which has also been shown to be the case for Donard Lake (66.25°N, 62°W) by Moore et al. (2001).

Working concurrently on a floating platform in the middle of a small lake (Hjort So) on an 80-km-long by 10.5-km-wide island (Store Koldewey) just off the coast of Northeast Greenland, Wagner et al. (2008) recovered two sediment cores of 70 and 252 cm length, the incremental portions of which they analyzed for grain-size distribution, macrofossils, pollen, diatoms, total carbon, total organic carbon, and several other parameters, the sequences of which were dated by accelerator mass spectrometry, with radiocarbon ages translated into calendar years before present. This work revealed, as they describe it, an “increase of the productivity-indicating proxies around 1,500-1,000 cal year BP, corresponding with the medieval warming,” while adding that “after the medieval warming, renewed cooling is reflected in decreasing amounts of total organic carbon, total diatom abundance, and other organisms, and a higher abundance of oligotrophic to meso-oligotrophic diatom taxa.” And, as they continue, “this period, the Little Ice Age, was the culmination of cool conditions during the Holocene and is documented in many other records from East and Northeast Greenland, before the onset of the recent warming [that] started ca. 150 years ago.”

In addition to the obvious importance of their finding evidence for the Medieval Warm Period, the six researchers’ statement that the Little Ice Age was the culmination, or most extreme sub-set, of cool conditions during the Holocene, suggests that it would not be at all unusual for such a descent into extreme coolness to be followed by some extreme warming, which further suggests there is nothing unusual about the degree of subsequent warming experienced over the 20th century, especially in light of the fact that the earth has not yet achieved the degree of warmth that held sway over most of the planet throughout large portions of that prior high-temperature period.

One year later, based on the use of a novel biomarker (IP25), which they described as a mono-unsaturated highly-branched isoprenoid that is synthesized by sea ice diatoms that have been shown to be stable in sediments below Arctic sea ice, Vare et al. (2009) used this new climatic reconstruction tool – together with “proxy data obtained from analysis of other organic biomarkers, stable isotope composition of bulk organic matter, benthic foraminifera, particle size distributions and ratios of inorganic elements” – to develop a spring sea ice record for that part of the central Canadian Arctic Archipelago. And in doing so, they discovered evidence for a decrease in spring sea ice between approximately 1200 and 800 years before present (BP), which they associated with “the so-called Mediaeval Warm Period.”

Contemporaneously, Norgaard-Pedersen and Mikkelsen (2009), working with a sediment core retrieved in August 2006 from the deepest basin of Narsaq Sound in southern Greenland, analyzed several properties of the materials thus obtained from which they were able to infer various “glacio-marine environmental and climatic changes” that had occurred over the prior 8,000 years. This work revealed the existence of two periods (2.3-1.5 ka and 1.2-0.8 ka) that appeared to coincide roughly with the Roman and Medieval Warm Periods, while they identified the colder period that followed the Medieval Warm Period as the Little Ice Age and the colder period that preceded it as the Dark Ages Cold Period. And citing the works of Dahl-Jensen et al. (1998), Andresen et al. (2004), Jensen et al. (2004) and Lassen et al. (2004), the two Danish scientists said that the cold and warm periods identified in those researchers’ studies “appear to be more or less synchronous to the inferred cold and warm periods observed in the Narsaq Sound record,” providing ever more evidence for the reality of the naturally-occurring phenomenon that governs this millennial-scale oscillation of climate.

One year later, Vinther et al. (2010) analyzed 20 ice core records from 14 different sites, all of which stretched at least 200 years back in time, as well as near-surface air temperature data from 13 locations along the southern and western coasts of Greenland that covered approximately the same time interval (1784-2005), plus a similar temperature data set from northwest Iceland (said by the authors to be employed “in order to have some data indicative of climate east of the Greenland ice sheet”). This work demonstrated that winter ð¹⁸O was “the best proxy for Greenland temperatures.” And based on that determination and working with three longer ice core ð¹⁸O records (DYE-3, Crete and GRIP), they developed a temperature history that extended more than 1400 years back in time.

This history revealed, in the words of the seven scientists, that “temperatures during the warmest intervals of the Medieval Warm Period” – which they defined as occurring some 900 to 1300 years ago – “were as warm as or slightly warmer than present day Greenland temperatures.” As for what this result implies, they state that further warming of present day Greenland climate “will result in temperature conditions that are warmer than anything seen in the past 1400 years,” which, of course, has not happened yet. Furthermore, Vinther et al. readily acknowledge that the independent “GRIP borehole temperature inversion suggests that central Greenland temperatures are still somewhat below the high temperatures that existed during the Medieval Warm Period.”

About this same time, Kobashi et al. (2010) had a paper published in which they had written that “in Greenland, oxygen isotopes of ice (Stuiver et al., 1995) have been extensively used as a temperature proxy, but the data are noisy and do not clearly show multi-centennial trends for the last 1,000 years in contrast to borehole temperature records that show a clear ‘Little Ice Age’ and ‘Medieval Warm Period’ (Dahl-Jensen et al., 1998).” However, they went on to note that nitrogen (N) and argon (Ar) isotopic ratios – ¹⁵N/¹⁴N and ⁴⁰Ar/³⁶Ar, respectively – can be used to construct a temperature record that “is not seasonally biased, and does not require any calibration to instrumental records, and resolves decadal to centennial temperature fluctuations.”

After describing the development of the new approach, they used it to construct a history of the last thousand years of central Greenland surface air temperature, based on values of the isotopic ratios of nitrogen and argon previously derived by Kobashi et al. (2008) from air bubbles trapped in the GISP2 ice core that had been extracted from central Greenland, obtaining the result depicted in the figure below.

Central Greenland surface temperature reconstruction for the last millennium. Adapted from Kobashi et al. (2010).

This figure depicts the central Greenland surface temperature reconstruction produced by the six scientists; and as best as can be determined from this representation, the peak temperature of the latter part of the Medieval Warm Period – which actually began some time prior to the start of their record, as demonstrated by the work of Dansgaard et al. (1975), Jennings and Weiner (1996), Johnsen et al. (2001) and Vinther et al. (2010) – was approximately 0.33°C greater than the peak temperature of the Current Warm Period, and about 1.67°C greater than the temperature of the last decades of the 20th century.

One year closer to the present, and noting that the varve thicknesses of annually-laminated sediments laid down by Hvitarvatn, a proglacial lake in the central highlands of Iceland, is controlled by the rate of glacial erosion and efficiency of subglacial discharge from the adjacent Langjokull ice cap, Larsen et al. (2011) employed a suite of environmental proxies contained within those sediments to reconstruct the region’s climate variability and glacial activity over the past 3000 years, which proxies included varve thickness, varve thickness variance, ice-rafted debris, total organic carbon (mass flux and bulk concentration), and the C:N ratio of sedimentary organic matter. And when all was said and done, this effort indicated that “all proxy data reflect a shift toward increased glacial erosion and landscape destabilization from ca 550 AD to ca 900 AD and from ca 1250 AD to ca 1950 AD, separated by an interval of relatively mild conditions,” and they state that “the timing of these intervals coincides with the well-documented periods of climate change commonly known as the Dark Ages Cold Period, the Medieval Warm Period, and the Little Ice Age.”

In the case of the Medieval Warm Period, they additionally note that “varve thickness decreases after 950 AD and remains consistently low through Medieval time with slightly thinner annual laminations than for any other multi-centennial period in the past 3000 years,” which suggests that the MWP was the warmest period of the past three millennia, while they say that “the LIA was the most severe multi-centennial cold interval of the late Holocene” and “likely since regional deglaciation 10,000 years ago.”

Finally, for those desiring additional brief reports on the Medieval Warm Period in the Arctic, go to www.co2science.org and search for Hill et al. (2001), Joynt and Wolfe (2001), Hantemirov and Shiyatov (2002), Andersson et al. (2003), Helama et al. (2005), Mazepa (2005), Weckstrom et al. (2006), Jiang et al. (2007), Zabenskie and Gajewski (2007), Grudd (2008), Justwan et al. (2008), Scire et al. (2008), Axford et al. (2009), Bjune et al. (2009), Cook et al. (2009), Fortin and Gajewski (2010), Büntgen et al. (2011), Divine et al. (2011), Ran et al. (2011), Velle et al. (2011), D’Andrea et al. (2012) and Esper et al. (2012), full references for which articles are included in the Reference section below.

In concluding this summary, it is clear that the suite of measurements described in the studies reviewed above continues to indicate that the Arctic – which climate models suggest should be super-sensitive to greenhouse-gas-induced warming – is still not even as warm as it was several centuries ago during portions of the Medieval Warm Period, when there was much less CO2 and methane in the air than there is today, which facts further suggest that the planet’s more modest current warmth need not be the result of historical increases in these two trace greenhouse gases.

Source: http://www.co2science.org/subject/m/summaries/mwparctic.php (with references)