SUPPLEMENTARY INFORMATION, Moberg et al.

Supplementary Note 1

Proxy data selection

We strived at using records with complete data in the 2000-year reconstruction period to avoid jumpwise inhomogeneities caused by a time-varying number of proxy series. Such long records are scarce; in particular the available records end in various years in the 20th century, wherefore it is impossible to reconstruct temperatures up to the end of the last century using only proxy data. Records that have been interpreted as reflecting annual mean temperatures were preferred, but a strict application of such a criterion would have led to a too limited amount of data. Therefore, data representing seasonal (mainly summer) temperatures were also allowed. This is not necessarily a large problem, as summer temperatures averaged over the entire hemisphere are strongly correlated with hemispheric annual mean temperatures1, but it could be serious as differential trends in summer and winter temperatures in the last millennium have been observed at least for some locations in China and Europe2. In any case, all previous NH temperature reconstructions also suffer from a warm-season bias2, and the problem seems unavoidable until more winter-sensitive proxies have been developed.

Low-resolution proxY DATA

The following eleven proxy data series (numbered as in Fig. 1 and Table 1) were used as input to the low-frequency component of the reconstruction. Time series plots of each series is shown in Fig. S1.

1. A stacked ice melt record derived from three ice cores from Agasssiz Ice cap, Ellesmere Island, Canada3-4. It shows changes in percent summer melt reflecting summer temperatures. We used an arithmetic average of data from the files:
ftp://ftp.ngdc.noaa.gov/paleo/icecore/polar/agassiz/pcmelt77_5yr.txt

ftp://ftp.ngdc.noaa.gov/paleo/icecore/polar/agassiz/pcmelt84_5yr.txt

ftp://ftp.ngdc.noaa.gov/paleo/icecore/polar/agassiz/pcmelt87_5yr.txt

2. Annual temperature reconstruction based on ice borehole temperatures from the Greenland Ice Core Project (GRIP) borehole3. Data source: digitized from Fig. 4B in ref. 5.

3. Summer temperature reconstruction derived from pollen composition in annually laminated sediments from Conroy Lake in the eastern US6. We used Conroy Lake data from: ftp://ftp.ngdc.noaa.gov/paleo/pollen/recons/liadata.txt

4. Spring temperature reconstruction estimated from variations in the Mg/Ca composition in fossil shells from Chesapeake Bay in the eastern U.S.7. Data source: ftp://ftp.ngdc.noaa.gov/paleo/contributions_by_author/cronin2003/cronin2003.txt

5-6. Two annual sea surface temperature (SST) reconstructions, from the Sargasso Sea8 and the Caribbean Sea9 respectively, estimated from the 18O ratio in planctonic foraminifera. Data sources:

ftp://ftp.ngdc.noaa.gov/paleo/contributions_by_author/keigwin1996/fig4bdata

ftp://ftp.ngdc.noaa.gov/paleo/contributions_by_author/nyberg2002/nyberg2002.txt

7. Summer temperature reconstruction based on diatom composition in sediments from Lake Tsuolbmajavri in northern Finland10. Data source: Atte Korhola and Jan Weckström (personal communication).

8. Annual mean temperature reconstruction based on a stalagmite 18O record from cave Søylegrotta, near Mo i Rana in northern Norway11. Data source: Stein-Erik Lauritzen (personal communication).

9. Summer temperature reconstruction for the Beijing region, China, based on observed correlation between stalagmite layer thickness (from the Shihua cave) and temperature12. Data source:

ftp://ftp.ngdc.noaa.gov/paleo/speleothem/china/shihua_tan2003.txt

10. A Chinese multi-proxy composite regional annual mean temperature reconstruction13 based on 9 individual records of various origins. We used the “Weighted” version in ref. 13. Data source:

ftp://ftp.ngdc.noaa.gov/paleo/contributions_by_author/yang2002/china_temp.txt

11. A combination of two marine sediment records from the Arabian Sea14-15 in which the percentage of the foraminifera Globigerina bulloides reflects the extent of ocean up-welling, which is determined by the strength of monsoons, which in turn indirectly reflect both summer and winter large-scale temperature changes through the differential seasonal heating and cooling of the Asian continent and surrounding oceans16. We used data from Core 723A15 for the early years up to 1390 A.D. and data from Core RC273014 from 1391 to 1986 A.D. (c.f. Fig. 3c in ref. 15). Although this record reflects temperatures only indirectly, it was included to improve the balance in the geographical distribution of proxy sites. Data sources:

ftp://ftp.ngdc.noaa.gov/paleo/contributions_by_author/anderson2002/anderson2002.txt

ftp://ftp.ngdc.noaa.gov/paleo/contributions_by_author/gupta2003/gupta2003.txt

Tree-ring data

Seven tree-ring temperature reconstructions (labeled as in Fig. 1) were used as input to the high-frequency component of the reconstruction; four from northern Eurasia and three from the western U.S. Their time series are plotted in Fig. S2. The northern Eurasian records are obtained from conifer trees (Pinus silvestris from northern Sweden17 (D in Fig. 1), Larix siberica from the Yamal Peninsula18 (E), Larix gmelinii from the Taimyr Peninsula19 (F) and Larix cajanderi Mayr from the Indigirka River region in north-eastern Yakutia20 (G)) growing close to the northern tree line limit where summer temperature is the major growth-limiting factor. The western U.S. records (A-C) are derived from Bristlecone Pine (Pinus longaeva) growing in Nevada and California (Indian Garden, Methuselah Walk, White Mountain Master; obtained from the International Tree Ring Data Bank, and are primarily interpreted as reflecting annual precipitation variability21. It has also been suggested that their signal at decadal and longer timescales is sensitive to variability in warm-season temperatures22 although this relationship is questionable21. We performed correlation analysis between the average of the three western U.S. tree-ring series and with temperatures averaged for the region 30-45N, 100-130W, and found a correlation of –0.57 for 10-year averages of annual mean temperatures (–0.22 at annual resolution). This is weaker than for the three northern Eurasian series from Torneträsk, Yamal and Taimyr, which after averaging to one regional series have a correlation of +0.79 with 10-year annual mean temperatures (+0.35 at annual resolution) in the region 60-75N, 5-130E, but still strong enough to suggest a negative response of tree-ring growth to annual mean temperatures at decadal timescales; i.e. tree growth is limited by dry and warm conditions and favoured by moist and cold conditions.

REFERENCES

  1. Esper, J., Frank, D. C. & Wilson, R. J. S. Climate reconstructions: Low-frequency ambition and high-frequency ratification. EOS85 (12), 113 (2004).
  2. Jones, P. D., Osborn, T. J. & Briffa, K. R. Changes in the Northern Hemisphere annual cycle: Implications for paleoclimatology? J. Geophys. Res. 108, D18, 4588, doi: 10.1029/2003JD003695 (2003).
  3. Fisher, D. A. & Koerner, R. M. Signal and Noise in four ice-core records from the Agassiz Ice Cap, Ellesmere Island, Canada: Details of the last millennium for stable isotopes, melt and solid conductivity.Holocene4, 113-120 (1994).
  4. Fisher, D. A., Koerner, R. M. & Reeh, N. Holocene climatic records from Agassiz ice cap, Ellesmere Island, NWT, Canada. Holocene5, 19-24 (1995).
  5. Dahl-Jensen, D. et al. Past temperatures directly from the Greenland ice sheet. Science 282, 268-271 (1998).
  6. Gajewski, K. Late Holocene climate changes in eastern North America estimated from pollen data. Quat. Res.29, 255-262 (1988).
  7. Cronin, T. M., Dwyer, G. S., Kamiya, T., Schwede, S. & Willard, D. A. Medieval warm period, Little ice age and 20th century temperature from Chesapeake Bay. Global Planet. Change 36, 17-29 (2003).
  8. Keigwin, L. D. The Little ice age and Medieval warm period in the Sargasso Sea. Science274, 1504-1508 (1996).
  9. Nyberg, J., Malmgren, B. A., Kuijpers, A. & Winter, A. A centennial-scale variability of tropical North Atlantic surface hydrography during the late Holocene.Palaeogeogr. Palaeoclimatol. Palaeoecol.183, 25-41 (2002).
  10. Korhola, A., Weckström, J., Holmström, L. & Erästö, P. A quantitative Holocene climatic record from diatoms in northern Fennoscandia. Quat. Res.54, 284-294 (2000).
  11. Lauritzen, S.-E. & Lundberg, J. Calibration of the speleothem delta function: an absolute temperature record for the Holocene in northern Norway. Holocene9, 659-669 (1999).
  12. Tan, M. et al.Cyclic rapid warming on centennial-scale revealed by a 2650-year stalagmite record of warm season temperature. Geophys. Res. Lett.30, 1617, doi: 10.1029/2003GL017352 (2003)
  13. Yang, B., Braeuning, A., Johnson, K. R. & Yafeng, S. General characteristics of temperature variation in China during the last two millennia. Geophys. Res. Lett.29 (9) 1324, doi:10.1029/GL014485 (2002).
  14. Anderson, D. M., Overpeck, J. T. & Gupta, A. K. Increase in the Asian SW Monsoon During the Past Four Centuries. Science297, 596-599 (2002).
  15. Gupta, A. K., Anderson, D. M. & Overpeck, J. T. Abrupt changes in the Asian southwest monsoon during the Holocene and their links to the North Atlantic Ocean.Nature421, 354-357 (2003).
  16. Overpeck, J., Anderson, D., Trumbore, S. & Prell, W. The southwest Indian Monsoon over the last 18 000 years. Clim. Dyn.12, 213-225 (1996).
  17. Grudd, H. et al. A 7400-year tree-ring chronology in northern Swedish Lapland: natural climatic variability expressed on annual to millennial timescales. Holocene12, 657-666 (2002).
  18. Hantemirov, R. M. &Shiyatov, S. G. A continuous multimillennial ring-width chronology in Yamal, northwestern Russia. Holocene12, 717-726 (2002).
  19. Naurzbaev, M. M.,Vaganov, E. A., Sidorova, O. V. & Schweingruber, F. H. Summer temperatures in eastern Taimyr inferred from a 2427-year late-Holocene tree-ring chronology and earlier floating series. Holocene12, 727-736 (2002).
  20. Sidorova O.V., & Naurzbaev, M. M. Reakzija na klimatitcheskie izmenemija listvennizi Kajandera na verchnei granitze lesa i v doline reki Indigirki (Response of Larix Kajanderi to climatic changes at the upper timberline and in the Indigirka River valley). Lesovedenie, 2, 73-75 (2002).
  21. Hughes, M. K. & Funkhouser, G. Frequency-dependent climate signal in upper and lower forest border tree rings in the mountains of the Great Basin. Clim. Change59, 233-244 (2003).
  22. LaMarche, V. C. Frequency-dependent relationships between tree-ring series along an ecological gradient and some dendroclimatic implications. Tree-Ring Bull.34, 1-20 (1974).

We gratefully acknowledge Håkan Grudd, Rashit Hantemirov, Atte Korhola, Stein-Erik Lauritzen, Mukhtar Naurzbaev, Eugene Vaganov and Jan Weckström for contributing with their proxy data series.

1