Solar forcing of climate during the last millennium recorded in lake sediments from northern Sweden

Abstract

We report on a sediment record from a small lake within the subarctic wetland complex Stordalen in northernmost Sweden covering the last 1000 years. Variations in the content of minerogenic material are found to follow reconstructed variations in the activity of the Sun between the 13th and 18th centuries. Periods of low solar activity are associated with minima in minerogenic material and vice versa. A comparison between the sunspot cycle and a long instrumental series of summer precipitation further reveals a link between the 11 yr solar cycle and summer precipitation variability since around 1960. Solar minima are in this period associated with minima in summer precipitation, whereas the amount of summer precipitation increases during periods with higher solar activity. Our results suggest that the climate responds to both the 11 yr solar cycle and to long-term changes in solar activity and in particular solar minima, causing dry conditions with resulting decreased runoff.

Keywords

climate lake sediments minerogenic material precipitation runoff solar activity

Introduction

A large and growing body of evidence suggests that the climate on Earth is intrinsically related to variations in the activity of the Sun. Holocene climate changes on decadal, centennial and millennial timescales have been suggested to be forced by variations in the activity of the Sun (e.g. Bond et al., 2001; Büntgen et al., 2006; Haigh, 2003; Haltia-Hovi et al., 2007; Magny, 2004; van Geel et al., 1996; Wang et al., 2005). Yet, because of the assumed small changes in total solar irradiance, amplifying mechanisms appear to be necessary to explain the amplitude of climate variability suggested by the paleorecords. An important non-linear process could be connected to the solar UV variability that influences the heat budget and circulation in the stratosphere (e.g. Haigh et al., 2005). Mechanisms have been suggested for how this signal is transferred to the troposphere (Matthes et al., 2006). The stratosphere–troposphere link, which has only recently become a focus of research, might provide a clue for the understanding of the solar influence on climate, since the effects of solar variability are much more clearly visible in the stratospheric circulation than in the troposphere (e.g. Labitzke, 2006). Furthermore, a series of papers demonstrates linkages between variability in the Sun’s activity and the North Atlantic Oscillation (NAO) and the Arctic Oscillation (AO) (Gimeno et al., 2003; Kodera, 2002; Lukianova and Alekseev, 2004; Ogi et al., 2003b). In agreement with findings from other studies, this indicates that solar modulation of climate may be locally amplified through changes in atmospheric circulation (Shindell et al., 2001).

Here we present a new record of last millennium variations in the amount of minerogenic material in a sediment sequence from Lake Inre Harrsjön in northern Sweden (Figure 1). Our results show an overall positive correlation between temporal variations in the content of minerogenic material, solar activity and temperatures of Atlantic water masses in the Norwegian Sea. A comparison between sunspot numbers and a record of last century summer precipitation from the nearby Abisko Scientific Research Station (ANS) indicates a periodic relationship between summer precipitation variability, the solar cycle and the NAO, with minima in the 11 yr solar cycle being related to summer precipitation minima and more negative phases of the NAO.

Figure 1.

Left: map of Fennoscandia showing the location of the study area (black) and the location of the oxygen isotope record representing temperature variability in Atlantic waters along the Norwegian coast from a sediment sequence in the eastern Norwegian Sea (grey) (Sejrup et al., 2010). Right: aerial photo of Lake Inre Harrsjön (available at Abisko Scientific Research Station). The lighter colored area southeast of the lake represents the Stordalen peatland.

Study site

Lake Inre Harrsjön (68°21′N, 19°03′E) is located c. 10 km east of the small village of Abisko in northern Sweden just south of Lake Torneträsk (Figure 1). The area is characterized by a steep precipitation gradient from the Norwegian coast to the west with an oceanic climate, to a more continental climate further east. Since 1913 weather has been monitored at the Abisko Scientific Research Station (Callaghan et al., 2010). Between 1913 and 2005 the mean annual temperature was −0.6°C, and the mean total annual precipitation was 305 mm, of which on average c. 40% fell in the summer.

The studied lake covers an area of 0.02 km², with a maximum depth of 5 m and a pH of c. 7. The lake is situated c. 350 m a.s.l. and has a catchment (c. 0.3 km²) that consists of mires and sub-alpine birch forest. The Stordalen mire southeast of the lake is composed of a slightly elevated ombrotrophic area containing permafrost and wet fen areas with no permafrost. The lake has no apparent inlet but is, through a narrow channel, connected to a system of lakes that extend to the west and northwest. A larger lake to the southeast (Lake Villasjön) drains water to the west through fen areas north and south of the central peat plateau. Groundwater recharge occurs from two springs south and northeast of Lake Inre Harrsjön. Organic sedimentation in Lake Inre Harrsjön was initiated 2650 cal. BP, likely in response to permafrost formation and frost heave of the surroundings (Kokfelt et al., 2010).

The initiation of peat deposition in the adjacent Stordalen mire has been dated to the mid Holocene (Kokfelt et al., 2010; Sonesson, 1972). The maximum lateral expansion of poor fen and/or bog communities was by Malmer and Wallén (1996) found to have been reached as late as c. 300 cal. BP, whereas Kokfelt et al. (2010) dated poor fen formation to c. 675 cal. BP and ombrotrophication to c. 120 cal. BP. Both these transitions likely indicate permafrost formation phases. Wet minerotrophic areas expanded during the last decades, leading to significant changes in carbon cycling of the mire (Christensen et al., 2004; Johansson et al., 2006; Malmer et al., 2005).

Material and methods

We studied a sediment sequence from Lake Inre Harsjön with respect to variations in the content of minerogenic material over the last 1000 years. The sediment sequence is a composite of a 26 cm surface gravity core sequence retrieved in 2004 and a 90 cm sequence retrieved in 2005 with a Russian corer (10 cm diameter). The uppermost 26 cm was sampled in 0.5 cm sections whereas the Russian core sequence was sampled in 1 cm sections. Significant variations in water and organic matter content of the sequences allowed the correlation to a parallel sediment sequence that has been dated using ²¹⁰Pb dating and ¹⁴C dating of terrestrial macrofossils combined with Bayesian modeling (Kokfelt et al., 2010). The temporal resolution of each sample was between 3 and 16 years.

The ignition residue (IR) content was calculated as the residue after ignition at 550 and 925°C (Kokfelt et al., 2010) relative to the weight of the total dry sediment sample (Figure 2a). The IR consists mainly of mineral matter, originating partly from allochthonous minerogenic material and partly from biogenic silica (BSi). BSi may in turn originate from both allochthonous sources (Kokfelt et al., 2009b, 2010) and from autochthonous production in the lake. Temporal changes of the BSi content in the sediments of Lake Inre Harrsjön have previously been analysed on a parallel sediment sequence but with a lower resolution (Kokfelt et al., 2010) than the IR record presented here (Figure 2a). In order to reduce uncertainties about allochthonous and autochthonous sources, the BSi record was interpolated and subtracted from the IR record (BSi corrected IR, hereafter IR-_BSi in the text) (Figure 2a). This remaining IR-_BSi fraction is considered to represent a predominantly minerogenic and allochthonous component of the sediment, brought into the lake either by runoff from the catchment, as a result of shore erosion, in suspension from the connected lakes or by wind.

Figure 2.

(a) Green: the ignition residue (IR) record from Lake Inre Harrsjön shown against ages based on ²¹⁰Pb dating and AMS ¹⁴C dates obtained on terrestrial macrofossils (Kokfelt et al., 2010). Blue: biogenic silica (BSi) content measured on a parallel sequence (Kokfelt et al., 2010). Black: the BSi corrected IR record; IR-_BSi. (b) Black: the IR-_BSi record (detrended) from Lake Inre Harrsjön. Dashed: the IR-_BSi record shown with the adjusted age model based on tuning with the solar modulation. Red: solar modulation function inferred from ¹⁴C production rates and neutron monitor data (Muscheler et al., 2007). (c) Black (solid and dashed): the IR-_BSi record from Lake Inre Harrsjön with the original and the adjusted age model, respectively. Black arrows indicate calibrated age control points based on ¹⁴C datings on terrestrial macrofossils (Kokfelt et al., 2009a, 2010). The black bar represents the time interval in which age control is based on 17 ²¹⁰Pb dates (Kokfelt et al., 2009a). Pink: a 5 yr running average of δ¹⁸O in foraminifera from a sediment record in the eastern Norwegian Sea (Figure 1), representing the temperature and transport of warm Atlantic waters entering the Nordic Basin (Sejrup et al., 2010)

Solar activity for the last 1000 years has been reconstructed based on ¹⁴C production rates and neutron monitor data (Muscheler et al., 2007).

Last millennium solar variability signature in a lake sediment record

A comparison of variations in solar activity and detrended IR-_BSi in the lake sediment sequence since ad 1000 is shown in Figure 2b. The two records show an overall and significant positive correlation (r²=0.2958). A period of low solar activity around ad 1230 as well as the Wolf, Spörer and Maunder solar minima between c. ad 1270 and 1715 have counterparts in the IR-_BSi record and are characterized by a low content of IR-_BSi of the sediment. The 18th and 19th centuries show relatively little variation in the IR-_BSi record. Overall, the correlation between longer-term variability in solar activity and IR-_BSi is seen best between the 13th and the early 18th centuries (r²=0.318 between ad 1200 and 1750), a period that is characterized by large variations in solar activity. Some of the disagreement in the timing between IR-_BSi and solar proxy data in this period might be explained by dating uncertainties. The age model between ad 1500 and 1950 is more uncertain since it is only based on a model connecting a ²¹⁰Pb record in top of the sediment sequence with a range of ¹⁴C dates from lower levels (Kokfelt et al., 2010). We therefore tested how much an adjustment within the errors if the age model could improve the correlation between IR-_BSi and solar activity reconstruction. This was done by tuning maximum and minimum values of IR-_BSi to the peak and minimum values of solar modulation around ad 1600 and 1690, respectively, and by a corresponding linear stretching and compression of the timescale. The adjustment resulted in an alternative age model and in an improved positive correlation between IR-_BSi and solar modulation to r²=0.4488 for the whole period and to r²=0.627 between ad 1200 and 1750.

A potential link to temperature of Atlantic water masses

A recent study demonstrates that variations in the July–August–September (JAS) temperature of Atlantic water masses entering the Norwegian Sea via the North Atlantic Slope Current (NwASC) have been strongly correlated with changes in solar activity over the last 1000 years (Sejrup et al., 2010). Our lake record shares striking similarities with this reconstruction (Figure 2c), where temperatures of the relatively shallow (~50 m deep) Atlantic water along the Norwegian Coast correlate positively with the amount of IR-_BSi in the Lake Inre Harrsjön record. Certain conspicuous features, such as an overall declining trend until the early 18th century, and isolated peaks around 1190 and 1745 can be found in both records. The records diverge after c. 1900. The apparent timing offset between ad 1500 and 1700 could again be an effect of increased dating uncertainties of the lake record for this section (Figure 2c) (Kokfelt et al., 2010). The chronology of the marine record is much better constrained, as this is based on ²¹⁰Pb and ¹³⁷Cs dating, identification of historical tephra and wiggle matching of a large number of ¹⁴C datings (Sejrup et al., 2010).

Solar influence on summer precipitation variability in the second half of the 20th century

In order to gain a better understanding of how the climate in the region could have responded to solar variability, we compare the long instrumental series of summer precipitation from Abisko Scientific Research Station (68°20′N, 19°02′E) since 1913 (Callaghan et al., 2010) with the sunspot number record (downloaded from ftp://ftp.ngdc.noaa.gov/STP/SOLAR_DATA/SUNSPOT_NUMBERS/). We find that minima in summer precipitation occurred during minima in the solar 11 yr cycle since around 1960, while in periods of high solar activity the average summer precipitation amount increased (Figure 3a). An 11 yr running correlation between summer precipitation and sunspot numbers reveals an overall positive correlation after 1960 that, however, is statistically significant at the 95% level only from 1975 to 1977 and from 1997 to 1998 (i.e. bracketing the years 1970–1982 and 1992–2003) (Figure 4). Another important feature of the comparison between the summer precipitation and the solar cycle is a correspondence between a period of generally low solar activity between 1962 and 1977, when the sunspot maximum in the 11 yr cycle around 1969 was low relative to surrounding maxima. During these 15 years the summer precipitation was 17% and 25% lower than during the 15 years before and after. We also note that during this period, solar minima were associated with cold summers. In contrast to the second half of the 20th century, a connection between summer precipitation and solar activity is not apparent during the first half of the 20th century.

Figure 3.

(a) Sunspot record and summer precipitation around Abisko since 1913 (Callaghan et al., 2010; Kokfelt et al., 2009a). (b) Sunspot record and the winter North Atlantic Oscillation (NAO) index (updated after Hurrell, 1995).

Figure 4.

11 yr running correlation between summer (JJA) precipitation and sunspot numbers. Coarsely dashed lines indicate significance at 95% and finely dashed line indicates significance at 99%.

A NAO-mediated link between solar activity and climate in northern Fennoscandia?

An important concept for the description of European precipitation and temperature patterns is the winter NAO, defined as the normalized pressure difference between Iceland and the Azores (Hurrell, 1995). In the positive mode, the pressure difference between Iceland and the Azores is high and causes strong westerlies and efficient transport of Atlantic air and moisture to northern Europe, whereas in the negative mode northern Europe experiences drier and colder winter conditions (Hurrell, 1995). Although the NAO has the largest impact on the winter climate, the winter NAO has also been found to impact on the summertime climate the following year (Ogi et al., 2003a) and a series of papers has linked the winter NAO to variations in solar activity (Gimeno et al., 2003; Kodera, 2002; Lukianova and Alekseev, 2004). The winter NAO further affects the transport of Atlantic water along the Norwegian coast where the stronger southwesterlies that prevail during a positive state of the NAO result in a wider and warmer branch at the surface in the Barents Sea (Ingvaldsen, 2005). The winter NAO index is shown together with the sunspot data in Figure 3b. It can be seen that, although there is no straightforward relationship between sunspot numbers and the winter NAO index, there is a tendency towards a negative state of the winter NAO during solar minima in the second half of the 20th century. Together with the overall positive (although not consistently statistically significant) correlation between summer precipitation and sunspots after 1960 (Figure 3a), our results indicate that summer precipitation may, to some extent, have been affected by solar variability via changes in atmospheric winter circulation (the NAO).

Discussion and conclusions

Many studies have shown effects of variations in solar activity on climate conditions in the Northern Hemisphere. Instrumental data and modeling results suggest that the NAO is a concept that could help to understand the link between solar variations, the dynamical response of the atmosphere and observed climate changes. The spatial extent of the NAO has been found to vary according to the phase of the solar cycle, extending into the stratosphere during solar maxima, whereas the NAO is confined to the eastern Atlantic sector in the troposphere during solar minima (Kodera, 2002). Lukianova and Alekseev (2004) found that the NAO after c. 1940 varied with the same rhythm as the aa index, which is a geomagnetic index linked to the magnetic activity of the Sun. The solar influence on the NAO has further been found to affect Northern Hemisphere temperatures, with a positive correlation between temperature and the winter NAO during solar maxima phases and a zero or negative correlation between temperature and the winter NAO during solar minima phases (Gimeno et al., 2003). Although the NAO is most important for winter climate, the high-latitude summer climate in the Northern Hemisphere is influenced by the NAO of the previous winter through the effect on SST, sea-ice extent and snow cover anomalies memorized from the winter season (Ogi et al., 2003a).

From our study we conclude that solar variability exerted an important control of climate variability in around Abisko between the early 13th and 18th centuries and after 1960. As the IR-_BSi component of the sediment is considered a mainly allochthonous component of the sediment, we interpret the variations in the IR-_BSi record from Lake Inre Harrsjön as a result of changes in runoff, where low solar activity is related to decreased runoff and vice versa. The comparison between sunspot numbers and local climate data showed that minima in summer precipitation co-occurred with minima in the 11 yr sunspot cycle during the second part of the 20th century, and that a period of relatively dry summer conditions was associated with overall decreased solar activity between 1962 and 1977. The dry period between 1962 and 1977 has previously been highlighted as a period with reconstructed decreased peat surface moisture content in the catchment mire and reduced lake water TOC concentrations in Lake Inre Harrsjön (Kokfelt et al., 2009a).The observation that summer precipitation and sunspot numbers co-vary after 1960 lends support to an interpretation towards lower summer runoff during solar minimum phases in the IR_-BSi record.

Other factors may however have influenced the variability in the record. Variation in winter snow/spring flood is likely a further significant contributor that can explain part of the recorded runoff variability. A strong influence of the NAO on the distribution of winter precipitation is, for example, of great importance to European glacier dynamics (Nesje and Dahl, 2003). At Abisko close to the studied lake, winter precipitation and temperature has previously been demonstrated to be weakly, but significantly and positively, correlated with the Arctic Oscillation (Kohler et al., 2006); an index highly correlated with the NAO (Wallace, 2000). The precipitation response to solar variations around Abisko is probably partly dependent on changes in the atmospheric circulation that affect the large-scale distribution of precipitation across Europe (Hurrell, 1995), with a negative state of the winter NAO occurring more frequently during solar minima phases. This is in correspondence with findings from other model and data studies (Shindell et al., 2001; Trouet et al., 2009).

Apart from the correlation with solar activity, the variability in our lake record was found to correlate positively with last millennium July–August–September (JAS) temperature variability of the Atlantic water entering the Norwegian Sea (Figure 2c). The transport of Atlantic water along the Norwegian coast is in turn affected by the stronger southwesterlies that prevail during a positive state of the NAO (Ingvaldsen, 2005 and references therein). Our IR-_BSi record from Lake Inre Harrsjön and the Atlantic water temperature reconstruction from the Norwegian Sea both show an overall decrease between c. ad 1200 and 1700. This trend is also found in the development of reconstructed Northern Hemisphere air temperatures (e.g. Moberg et al., 2005) potentially pointing to a long-term temperature influence also present in the IR-_BSi record. Our study thus indicates that Atlantic sea-water temperatures likely provide an important link between the solar record and the reconstructed runoff variations.

External factors other than solar variability may obviously influence precipitation. Contemporaneous periods of explosive volcanism and solar minima have made a clear separation of low frequent temperature responses difficult during some periods over the last 1000 years (Ammann et al., 2007) and recent research has introduced the view that the onset of the ‘Little Ice Age’ (LIA) cold summer conditions can have been triggered by explosive volcanism and maintained by sea-ice/ocean feedbacks (Miller et al., 2012). Strong volcanic forcing may also influence precipitation, but such effects are normally restricted to short-lived events (e.g. Lamoureux et al., 2001), whereas our record seem to mirror more low frequent variability (Figure 2). The possibility of volcanic impacts on precipitation in the region can and should not be excluded, but detection of such a signal would require a higher and preferably annually resolved and/or varved sediment record. Support that the runoff record from Lake Inre Harrsjön is at least partly forced by solar-induced changes in precipitation, is however found not only in the comparison between sunspot number and summer precipitation but also in the sediment record itself. Hence, the best dated of the IR_-BSi minima (constrained by three radiocarbon dates) in our lake record that coincide with reduced solar activity occurred c. ad 1230 and thus pre-date the explosive volcanism considered as a trigger for the LIA by Miller et al. (2012).

We conclude that the Sun has been an important driver of climate in northern Fennoscandia over the last 1000 years and that summer precipitation may be affected by variations in solar activity. The relationship between solar proxies, instrumental data from the last century, NAO, and Atlantic water temperatures off the Norwegian coast from the last millennium highlight links that could shed light on the mechanisms behind the Sun–climate relationship.

Footnotes

Acknowledgements

The authors are grateful to D Hammarlund and M Rundgren for help with collection of the lake sediment sequence, as part of a former project funded by the Swedish Research Council. The authors are further grateful to H Sejrup, SJ Lehman, N Reuss and Abisko Scientific Research Station (ANS) for providing data. I Snowball and two anonymous reviewers are thanked for constructive comments on the manuscript.

Funding

U Kokfelt was supported by RQ08, the Crafoord Foundation and the Carlsberg Foundation. R Muscheler is supported by the Royal Swedish Academy of Sciences through a grant financed by the Knut and Alice Wallenberg Foundation. The Swedish Research Council supported this study through a Linnaeus grant to Lund University (LUCCI). Thanks to the National Danish Research Foundation for funding the activities within the Center for Permafrost (CENPERM).

References

Ammann

Joos

Schimel

. (2007) Solar influence on climate during the past millennium: Results from transient simulations with the NCAR Climate System Model. PNAS 104(10): 3713–3718.

Bond

Kromer

Beer

. (2001) Persistent solar influence on North Atlantic climate during the Holocene. Science 294: 2130–2136.

Büntgen

Frank

Nievergelt

. (2006) Summer temperature variations in the European Alps, AD 755–2004. Journal of Climate 19: 5606–5623.

Callaghan

Bergholm

Christensen

. (2010) A new climate era in the sub-Arctic? Accelerating climate changes and multiple impacts. Geophysical Research Letters 37(14): 14705, doi:10.1029/2009GL042064.

Christensen

Johansson

Åkerman

. (2004) Thawing sub-arctic permafrost: Effects on vegetation and methane emissions. Geophysical Research Letters 31: L04501, doi:10.1029/2003GL018680.

Gimeno

Torre

Nieto

. (2003) Changes in the relationship NAO–Northern Hemisphere due to solar activity. Earth and Planetary Science Letters 206: 15–50.

Haigh

(2003) The effects of solar variability on the Earth’s climate. Philosophical Transactions of the Royal Society in London A 361: 95–111.

Haigh

Blackburn

Day

(2005) The response of tropospheric circulation to pertubations in lower-stratospheric temperature. Journal of Climate 18: 3672–3685.

Haltia-Hovi

Saarinen

Kukkonen

(2007) A 2000-year record of solar forcing on varved lake sediment in eastern Finland. Quaternary Science Reviews 26: 678–689.

10.

Hurrell

(1995) Decadal trends in the North Atlantic Oscillation: Regional temperatures and precipitation. Science 269: 676–679.

11.

Ingvaldsen

(2005) Width of the North Cape Current and position of the Polar Front in the western Barents Sea. Geophysical Research Letters 32: L16603.

12.

Johansson

Malmer

Crill

. (2006) Decadal vegetation changes in a northern peatland, greenhouse gas fluxes and net radiative forcing. Global Change Biology 12: 2352–2369.

13.

Kodera

(2002) Solar cycle modulations of the North Atlantic Oscillation: Implication in the spatial structure of the NAO. Geophysical Research Letters 29: L014557.

14.

Kohler

Brandt

Johansson

. (2006) A long-term snow depth record from Abisko northern Sweden, 1913–2004. Polar Research 25(2): 91–113.

15.

Kokfelt

Reuss

Struyf

. (2010) Wetland development, permafrost history and nutrient cycling inferred from late Holocene peat and lake sediment records in subarctic Sweden. Journal of Paleolimnology 44: 327–342.

16.

Kokfelt

Rosén

Schoning

. (2009a) Ecosystem responses to increased precipitation and permafrost decay in sub-arctic Sweden inferred from peat and lake sediments. Global Change Biology 15: 1652–1663.

17.

Kokfelt

Struyf

Randsalu

(2009b) Diatoms in peat – Dominant producers in a changing environment? Soil Biology & Biochemistry 41: 1764–1766.

18.

Labitzke

(2006) Solar variation and stratospheric response. Space Science Reviews 125: 247–260.

19.

Lamoureux

England

Sharp

. (2001) A varve record of increased ‘Little Ice Age’ rainfall associated with volcanic activity, Arctic Archipelago, Canada. The Holocene 11(2): 243–249.

20.

Lukianova

Alekseev

(2004) Long-term correlation between the NAO and solar activity. Solar Physics 224: 445–454.

21.

Magny

(2004) Holocene climate variability as reflected by mid-European lake-level fluctuations and its probable impact on prehistoric human settlements. Quaternary International 113: 65–79.

22.

Malmer

Wallén

(1996) Peat formation and mass balance in subarctic ombrotrophic peatlands around Abisko, northern Scandinavia. In: Karlsson

Callaghan

(eds) Plant ecology in the sub-arctic Swedish Lapland. Ecological Bulletins 45: 79–92.

23.

Malmer

Johansson

Olsrud

. (2005) Vegetation, climatic changes and net carbon sequestration in a North-Scandinavian subarctic mire over 30 years. Global Change Biology 12: 1895–1909.

24.

Matthes

Kuroda

Kodera

. (2006) Transfer of the solar signal from the stratosphere to the troposphere: Northern winter. Journal of Geophysical Research 111: D06108, doi:10.1029/2005JD006283.

25.

Miller

Geirsdottir

Zhong

. (2012) Abrupt onset of the Little Ice Age triggered by volcanism and sustained by sea-ice/ocean feedbacks. Geophysical Research Letters 39: L02708.

26.

Moberg

Sonechkin

Holmgren

. (2005) Highly variable Northern Hemisphere temperatures reconstructed from low- and high-resolution proxy data. Nature 433: 613–617.

27.

Muscheler

Joos

Beer

. (2007) Solar activity during the last 1000 yr inferred from radionuclide records. Quaternary Science Reviews 26: 82–97.

28.

Nesje

Dahl

(2003) The ‘Little Ice Age’: Only temperature? The Holocene 13(1): 139–145.

29.

Ogi

Tachibana

Yamazaki

(2003a) Impact of the wintertime North Atlantic Oscillation (NAO) on the summertime atmospheric circulation. Geophysical Research Letters 30: L017280.

30.

Ogi

Yamazaki

Tachibana

(2003b) Solar cycle modulation of the seasonal linkage of the North Atlantic Oscillation (NAO). Geophysical Research Letters 30: L018545.

31.

Sejrup

Lehman

Haflidason

. (2010) Response of Norwegian Sea temperature to solar forcing since 1000 AD. Journal of Geophysical Research 115: C12034.

32.

Shindell

Schmidt

Mann

. (2001) Solar forcing of regional climate change during the Maunder Minimum. Science 294: 2149–2152.

33.

Sonesson

(1972) Cryptogams. In: International Biological Programme Swedish Tundra Biome Project. Technical Report No. 9, April 1972. Swedish Natural Science Research Council Ecological Research Committee.

34.

Trouet

Esper

Graham

. (2009) Persistent positive North Atlantic Oscillation mode dominated the Medieval Climate Anomaly. Science 324: 78–80.

35.

Van Geel

Buurman

Waterbolk

(1996) Arcaeological and paleoecological indications for an abrupt climate change in The Netherlands and evidence for climatological teleconnections around 2650 BP. Journal of Quaternary Science 11: 451–460.

36.

Wallace

(2000) North Atlantic Oscillation/Annular Mode: Two paradigms – One phenomenon. Quarterly Journal of the Royal Meteorological Society 126: 791–805

37.

Wang

Cheng

Edwards

. (2005) The Holocene Asian Monsoon: Links to solar changes and North Atlantic climate. Science 308: 854–857.