Persistent Solar Influence on North Atlantic Climate During the Holocene
Tóm tắt
Từ khóa
Tài liệu tham khảo
G. C. Bond et al. in Mechanisms of Global Climate Change at Millennial Time Scales P. Clark R. Webb L. D. Keigwin Eds. (Geophysical Monograph Series 112 American Geophysical Union Washington DC 1999) pp. 35–58.
P. Reimer calculated a 14 C production rate for us from the INTCAL98 Δ 14 C tree-ring data set using a four-box carbon-cycling model (atmosphere oceanic mixed layer deep sea and the biosphere) from (40).
Vautard R., Ghil M., Physica D58, 95 (1992).
A Blackman-Tuckey cross-spectral analysis (11) was applied to the smoothed and detrended records of 14 C and 10 Be production rates and to the detrended HSG in MC52-VM29-191 (shown graphically in Web fig. 2). For HSG and 10 Be 70-year sampling interval 95% confidence interval 48 lags and 0.000557 cycles/year bandwidth were used. Coherency in a 300- to 500-year band is 0.75 and in a 900- to 1100-year band is 0.93. For HSG and 14 C 70-year sampling interval 95% confidence interval 69 lags and 0.00039 bandwidth were used. Coherency in both the 300- to 500-year and 900- to 1100-year band is 0.84. We also used singular spectrum analyses (14) to decompose the same records into millennial (1000 to 2000 years) and centennial (300 to 500 years) components and found statistically significant coherence between the millennial components of the records. Singular spectrum analyses (SSA) were performed with an embedding dimension of 30 or 20 to 25% of the time-series lengths. Only the leading four to six empirical orthogonal functions which cumulatively described 75 to 80% of the variance in the detrended and smoothed records were retained in these analyses. Standard Fourier analyses were used to aggregate individual reconstructed component time series into the millennial- and submillennial-frequency time series. Quasi-coherences were then computed between these SSA time series corresponding to the solar and marine proxies. For the marine and 14 C records the coherency is 0.53 and is significant at >90% with nine effective degrees of freedom; for marine and 10 Be records the coherency is 0.74 and is significant at >95% with seven effective degrees of freedom; and for 10 Be and 14 C records the coherency is 0.76 and is significant at >95% with seven effective degrees of freedom. Coherence between submillennial components of the HSG and 14 C records of 0.23 with 17 effective degrees of freedom is not significant at the >90% level.
A box diffusion carbon cycle model (41) implies that a Δ 14 C increase of 25‰ in 100 years (equivalent to a production-rate increase of about 0.8 atoms cm −2 s −1 during the large century-scale events shown in Fig. 3B) would require a 50% reduction in global ocean mixing. Coupled climate–carbon cycle models suggest that even for large millennial-scale climatic variations such as during the Younger Dryas Δ 14 C would change only 10 to 40% (42); similarly carbon reservoir models suggest that a 30% change in production rate over ∼200 years would require a two-thirds reduction in global wind speeds relative to the current average (43). There is no evidence of such large ocean/climate forcing in the North Atlantic during the Holocene (1); the “8200-year” cold event regarded by some as the most extreme climatic event of the Holocene (44) does not stand out conspicuously in either the Δ 14 C or 10 Be records relative to Holocene baselines (Fig. 3 A and B) nor does the event at about 10 300 years ago which appears to be associated with a modest decrease in NADW production (1). It has been further demonstrated from box models that over the entire Holocene centennial-scale Δ 14 C changes calculated from a 10 Be-based 14 C production rate agree well in both amplitude and phasing (generally within several percent) with the measured Δ 14 C variations (45–47). Because 10 Be and 14 C have significantly different geochemical behavior it is highly unlikely that climate forcing alone could have produced such similar changes in the two nuclides. Over the last 7000 years centennial-scale variations in ice core 10 Be flux appear to correlate well between Greenland and Antarctica (48) also consistent with production as the dominant control.
S. R. O'Brien et al. Science 270 1962 (1995).
S. Björck et al. Geology in press.
M. Cane A. Clement in Mechanisms of Global Climate Change at Millennial Time Scales P. Clark R. Webb L. D. Keigwin Eds. (Geophysical Monograph Series 112 American Geophysical Union Washington DC 1999) pp. 373–383.
For a cooling of 0.5° to 1°C accompanying the increases in drift ice (1) δ 18 O would increase between 0.1 and 0.2‰ (49). Because the mean decrease in planktic δ 18 O in MC52-VM29-191 (Fig. 2) is about 0.3‰ a 0.4 to 0.5‰ decrease in δ 18 O due to salinity is required during each increase in drift ice. On the basis of the δ 18 O of seawater from near the coring site in (50) 1‰ δ 18 O of seawater = about 1.2 salinity units. Hence a salinity decrease of about 0.5 to 0.6 salinity units is required to account for the observed decrease in planktic δ 18 O at MC52-VM29-191.
G. A. Schmidt G. R. Bigg E. J. Rohling. 1999. “Global Seawater Oxygen-18 Database” (www.giss.nasa.gov/data/o18data/).
U. Pflaumann et al. Paleoceanography in preparation.
S. Levitus R. Gelfeld NODC Inventory of Physical Oceanographic Profiles. Key to Oceanographic Records Documentation No. 18 (National Oceanographic Data Center Washington DC 1992).
J. McManus D. Oppo L. Keigwin J. Cullen G. Bond Quat. Res. in press.
H. Lamb Climate History and the Modern World (Routledge London New York ed. 2 1995).
L. Labeyrie et al. in Mechanisms of Global Climate Change at Millennial Time Scales P. Clark R. Webb L. D. Keigwin Eds. (Geophysical Monograph Series 112 American Geophysical Union Washington DC 1999) pp. 77–98.
J.-C. Duplessy L. Labeyrie A. Juillet-Leclerc J. Duprat in The Last Deglaciation: Absolute and Radiocarbon Chronologies E. Bard W. S. Broecker Eds. (NATO ASI Series I: Global Environmental Change Springer-Verlag Berlin 1992) vol. 2 pp. 201–217.
Updated from M. Weinelt M. Sarnthein E. Jansen H. Erlenkeuser H. Schulz in The Younger Dryas S. Troelstra J. van Hinte G. M. Ganssen Eds. (North-Holland Amsterdam 1995) pp. 109-115 [M. Weinelt personal communication
sea-surface temperatures in this publication were estimated with the GLAMP 2000 method (51)].
Isarin R., Bohncke S., Ouat. Res. 51, 158 (1999).
We thank J. Lynch-Stieglitz M. Cane R. Anderson G. Kukla J. Lean and D. Shindell for comments on the manuscript. This work was supported in part by grants from the National Science Foundation and from the National Oceanic and Atmospheric Administration. Support for the core curation facilities of the Lamont-Doherty Earth Observatory (LDEO) Deep-Sea Sample Repository is provided by the National Science Foundation (grant OCE00-02380) and the Office of Naval Research (grant N00014-96-I-0186). This is LDEO contribution 6272.