The Medieval Warm Period

It has frequently been suggested that the period encompassing the ninth to the fourteenth centuries A.D. experienced a climate warmer than that prevailing around the turn of the twentieth century. This epoch has become known as the Medieval Warm Period, since it coincides with the Middle Ages in Europe. In this review a number of lines of evidence are considered, (including climate-sensitive tree rings, documentary sources, and montane glaciers) in order to evaluate whether it is reasonable to conclude that climate in medieval times was, indeed, warmer than the climate of more recent times. Our review indicates that for some areas of the globe (for example, Scandinavia, China, the Sierra Nevada in California, the Canadian Rockies and Tasmania), temperatures, particularly in summer, appear to have been higher during some parts of this period than those that were to prevail until the most recent decades of the twentieth century. These warmer regional episodes were not strongly synchronous. Evidence from other regions (for example, the Southeast United States, southern Europe along the Mediterranean, and parts of South America) indicates that the climate during that time was little different to that of later times, or that warming, if it occurred, was recorded at a later time than has been assumed. Taken together, the available evidence does not support a global Medieval Warm Period, although more support for such a phenomenon could be drawn from high-elevation records than from low-elevation records.

The available data exhibit significant decadal to century scale variability throughout the last millennium. A comparison of 30-year averages for various climate indices places recent decades in a longer term perspective.

It is hypothesised that the Medieval Warm Period was preceded and followed by periods of moraine deposition associated with glacier expansion. Improvements in the methodology of radiocarbon calibration make it possible to convert radiocarbon ages to calendar dates with greater precision than was previously possible. Dating of organic material closely associated with moraines in many montane regions has reached the point where it is possible to survey available information concerning the timing of the medieval warm period. The results suggest that it was a global event occurring between about 900 and 1250 A.D., possibly interrupted by a minor readvance of ice between about 1050 and 1150 A.D.

Available evidence for climatic conditions in the southern Canadian Rockies around the period of the Early Medieval Warm Period is presented and reviewed. Treelines appear to have been above present levels during the 14th–17th centuries and there is limited evidence of higher treelines ca. 1000 ¹⁴C yr B.P. (ca. 1000 A.D.). During the 13th century at least three glaciers were advancing over mature forest in valley floor sites, 0.5-1.0 km upvalley of Little Ice Age maximum positions attained in the 18th and 19th centuries. Tree-ring width chronologies from treeline sites show suppressed growth in the early 12th century and for several periods in the 12th–14th centuries. The only tree-ring chronology presently spanning the 900–1300 A.D. interval has generally wider ringwidths between 950 and 1100 A.D. suggesting conditions were more favourable at that time. Forested sites overrun by glaciers in the 12th–14th centuries have only been deglaciated within the present century.

A tree-ring reconstruction of summer temperatures from northern Patagonia shows distinct episodes of higher and lower temperature during the last 1000 yr. The first cold interval was from A.D. 900 to 1070, which was followed by a warm period A.D. 1080 to 1250 (approximately coincident with the Medieval Warm Epoch). Afterwards a long, cold-moist interval followed from A.D. 1270 to 1660, peaking around 1340 and 1640 (contemporaneously with early Little Ice Age events in the Northern Hemisphere). In central Chile, winter rainfall variations were reconstructed using tree rings back to the year A.D. 1220. From A.D. 1220 to 1280, and from A.D. 1450 to 1550, rainfall was above the long-term mean. Droughts apparently occurred between A.D. 1280 and 1450, from 1570 to 1650, and from 1770 to 1820. In northern Patagonia, radiocarbon dates and tree-ring dates record two major glacial advances in the A.D. 1270–1380 and 1520–1670 intervals. In southern Patagonia, the initiation of the Little Ice Age appears to have been around A.D. 1300, and the culmination of glacial advances between the late 17th to the early 19th centuries.

Most of the reconstructed winter-dry periods in central Chile are synchronous with cold summers in northern Patagonia, resembling the present regional patterns associated with the El Niño-Southern Oscillation (ENSO). The years A.D. 1468–69 represent, in both temperature and precipitation reconstructions from treerings, the largest departures during the last 1000 yr. A very strong ENSO event was probably responsible for these extreme deviations. Tree-ring analysis also indicates that the association between a weaker southeastern Pacific subtropical anticyclone and the occurence of El Niño events has been stable over the last four centuries, although some anomalous cases are recognized.

In the set of climate reconstructions from tree-rings available for Europe, Scandinavia and North Africa, there are very few reconstructions relating to the Middle Ages, one of the main reasons being the scarcity of continuous and reliable tree-ring series. The five longest temperature reconstructions covering the period 950–1500 are presented here. A sixth reconstruction is proposed which concerns the mean April to September temperature at the geographical point 45°; N-10° E (Northeastern Italy), and a comparison is made with the five other reconstructions.

Several questions concerning the Medieval Warm Period (MWP), an interval (A.D. 900 to 1300) of elevated temperatures first identified in northern Europe, are addressed with paleoenvironmental and archaeological data from the southern Colorado Plateau in the southwestern United States. Low and high frequency variations in alluvial groundwater levels, floodplain aggradation and degradation, effective moisture, dendroclimate, and human adaptive behavior fail to exhibit consistent patterns that can be attributed to either global or regional expressions of the MWP. There is some suggestion, however, that climatic factors related to the MWP may have modified the regional patterns to produce minor anomalies in variables such as the number of intense droughts, the occurrence of specific droughts in the twelfth and thirteenth centuries, the prevalence of low temporal variability in dendroclimate, and the coherence of some low and high frequency environmental variables and aspects of human adaptive behavior. These results suggest that the MWP does not represent warming throughout the world. Rather, it was a complex phenomenon that probably was expressed differently in different regions.

The zenith of Anasazi Pueblo Indian occupation in the northern Colorado Plateau region of the southwestern U.S.A. coincides with the Little Climatic Optimum or Medieval Warm Period (A.D. 900-1300), and its demise coincides with the commencement of the Little Ice Age. Indexes of winter (jet-stream derived) and summer (monsoon derived) precipitation and growing season length were developed for the La Plata Mountains region of southwestern Colorado. The results show that during the height of the Little Climatic Optimum (A.D. 1000–1100) the region was characterized by a relatively long growing season and by a potential dry farming zone or elevational belt (currently located between 2,000 m and 2,300 m elevation) that was twice as wide as present and could support Anasazi upland dry farming down to at least 1,600 m, an elevation that is quite impossible to dry farm today because of insufficient soil moisture. This expanded dry-farm belt is attributable to a more vigorous circulation regime characterized by both greater winter and summer precipitation than that of today. Between A.D. 1100 and 1300 the potential dry-farm belt narrowed and finally disappeared with the onset of a period of markedly colder and drier conditions than currently exist. Finally, when the Little Ice Age terminated in the mid A.D. 1800s and warmer, wetter conditions returned to the region, another group of farmers (modern Anglos) were able to dry farm the area.

Decreased solar activity correlates with positive cosmogenic isotope anomalies, and with cool, wet climate in temperate regions of the world. The relationship of isotope anomalies to climate may be the opposite for areas influenced by monsoonal precipitation, i.e., negative anomalies may be wet and warm. Petersen (1988) has found evidence for increased summer precipitation in the American Southwest that can be shown to be coincident with negative ¹⁴C anomalies during the Medieval Warm Period. The present study compares palynological indicators of lake level for the Southwest with Petersen’s data and with the ¹⁴C isotope chronology. Percentages of aquatic pollen and algae from three sites within the Arizona Monsoon record greater lake depth or fresher water from A.D. 700–1350, between the Roman IV and Wolf positive isotope anomalies, thereby supporting Petersens’s findings. Maximum summer moisture coincides with maximum population density of prehistoric people of the Southwest. However, water depth at a more northern site was low at this time, suggesting a climateisotope relationship similar to that of other temperate regions. Further analysis of latitudinal patterns is hampered by inadequate ¹⁴C dating.

The collected documentary records of the cultivation of citrus trees and Boehmeria nivea (a perennial herb) have been used to produce distribution maps of these plants for the eighth, twelfth and thirteenth centuries A.D. The northern boundary of citrus and Boehmeria nivea cultivation in the thirteenth century lay to the north of the modern distribution. During the last 1000 years, the thirteenthcentury boundary was the northernmost. This indicates that this was the warmest time in that period. On the basis of knowledge of the climatic conditions required for planting these species, it can be estimated that the annual mean temperature in south Henan Province in the thirteenth century was 0.9-1.0 °C higher than at present. A new set of data for the latest snowfall date in Hangzhou from A.D. 1131 to 1264 indicates that this cannot be considered a cold period, as previously believed.

A long δ¹³C chronology was developed from bristlecone pine (Pinus longaeva) at the Methuselah Walk site in the White Mountains of California. The chronology represents cellulose from five-year ring groups pooled from multiple radii of multiple trees. The most dramatic isotopic event in the chronology appears from A.D. 1080–1129, when δ¹³C values are depressed to levels ∼ 2σ below the mean for the period A.D. 925–1654. This isotopic excursion appears to represent a real event and is not an artifact of sampling circumstances; in fact, a similar excursion occurs in a previously-reported, independent δ¹³C chronology from bristlecone pine. By carbon isotope fractionation models, the shift to low δ¹³C values is consistent with abundant soil moisture, permitting leaf stomata to remain open, and allowing ready access of CO₂ from which carbon fixation may discriminate more effectively against ¹³C in favor of ¹²C. According to this model, the ¹³C-depleted 50-yr isotopic excursion represents the wettest period in the White Mountains in the past 1000 yr, during which isotope-reconstructed July Palmer Drought Severity Indices averaged ∼ +2.2.

Paleoclimatic studies of the Medieval Solar Maximum (c. A.D. 1100–1250, corresponding with the span of the Medieval Warm Epoch) may prove useful because it provides a better analog to the present solar forcing than the intervening era. The Medieval Solar Activity Maximum caused the cosmogenic isotope production minimum during the 12th and 13th Centuries A.D. reflected by Δ¹⁴C and ¹⁰Be records stored in natural archives. These records suggest solar activity has returned to Medieval Solar Maximum highs after a prolonged period of reduced solar activity. Climate forcing by increased solar activity may explain some of this century’s temperature rise without assuming unacceptably high climate sensitivity. By analogy with the Medieval Solar Activity Maximum, the contemporary solar activity maximum may be projected to last for 150 years. The maximum temperature increase forced by increased solar activity stays well below the predicted doubled atmospheric CO₂ greenhouse forcing.

We document the characteristic time scales of variability for seven climate indices whose time-dependent behavior is sensitive to some aspect of the El Niño/Southern Oscillation (ENSO). The ENSO sensitivity arises from the location of these long-term records on the periphery of the Indian and Pacific Oceans. Three of the indices are derived principally from historical sources, three others consist of tree-ring reconstructions (one of summer temperature, and the other two of winter rainfall), and one is an annual record of oxygen isotopic composition for a high-elevation glacier in Peru. Five of the seven indices sample at least portions of the Medieval Warm Period (∼ A.D. 950 to 1250).

In general, spectral power on time scales of ∼ 2–6 yr is statistically significant and persists throughout most of the time intervals sampled by the different indices. Assuming that the ENSO phenomenon is the source of much of the variability at these time scales, this indicates that ENSO has been an important part of interannual climatic variations over broad areas of the circum-Pacific region throughout the last millennium. Significant coherence values were found for El Niño and reconstructed Sierra Nevada winter precipitation at ∼ 2–4 yr throughout much of their common record (late 1500s to present) and between 6 and 7 yr from the mid-18th to the early 20th century.

At decadal time scales each record generally tends to exhibit significant spectral power over different periods at different times. Both the Quelccaya Ice Cap δ¹⁸O series and the Quinn El Niño event record exhibit significant spectral power over frequencies ∼ 35 to 45 yr; however, there is low coherence between these two series at those frequencies over their common record. The Sierra Nevada winter rainfall reconstruction exhibits consistently strong variability at periods of ∼ 30–60 yr.