There seems to be lots of interest in solar issues and lots of controversy among specialists. For now, let’s simply look at the millennial proxies for solar activity. For now, I don’t want to discuss issues like cosmic ray modulation or that sort of stuff on this thread – all in good time. Put it on Unthreaded if you like, but save this thread to try to assess the proxies. Let’s start by at least canvassing what the proxy data is.
There are two classes of proxy data – Be10 from ice cores in Antarctica and Greenland; and àÅ½”¬?C14 from tree rings. There are several main protagonists in the field of solar proxies – on the one hand, Usoskin, Solanki et al who argue that 20th century solar activity is at unusual elevated levels; on the other hand, Muscheler et al who argue that 20th century solar activity is not anomalous; Ralph Bard is another important protagonist.
Here I’m going to present some of the primary solar reconstructions. I’m not going to comment extensively on the pro’s or con’s of any of the reconstructions as I’m still feeling my way through the data and methods. It’s a complicated area in itself and I’m doing other things as well.
Bard Reconstruction from South Pole
First, here is one of the first millennial length solar reconstructions – this one from Bard et al (Tellus 2000) – this figure from Bard 2006. This reconstruction was based on smoothed Be10 values from South Pole ice core data (inverted). In this case, both the data and reconstruction are archived.
Figure 1. Bard et al 2000 reconstruction.
Here is the South Pole Be10 data (inverted) together with a smooth (reproduced from ftp://ftp.ncdc.noaa.gov/pub/data/paleo/climate_forcing/solar_variability/bard_irradiance.txt ) . This has a five 9s correlation to the Bard et al reconstruction (also archived.) So the Bard et al 2000 reconstruction is simply a linear transformation of Be10 data.
Figure 2. Plot of archived Be-10 South Pole data from WDCP.
Usoskin et al Reconstructions
Usoskin and associates in a series of publications (including Usoskin et al PRL 2003; Usoskin et al Astr Astr 2004 ; Solanki et al Nature 2004 and others) provided two reconstruction variations, one using Be-10 and one using àÅ½”¬?C14. Here is a summary figure from Usoskin et al (PRL 2003) entitled: Millennium-Scale Sunspot Number Reconstruction: Evidence for an Unusually Active Sun since the 1940s.
The red series here (Antarctic) can be recognized as a variation of the South Pole Be-10 series used in Bard et al 2000. The green series (Greenland) is based on Dye-3 Be-10 data (shown in the next figure); the àÅ½”¬?C14 is reconstructed from Intcal data. It’s my understanding that Usoskin et al made a nonlinear model linking the proxy data to sunspot numbers. It’s dressed up in complicated terminology, but I’m not sure how much it rises above a fairly simple univariate relationship.
Usoskin et al PRL 2003 FIG. 2 (color). Time series of the sunspot number as reconstructed from 10Be concentrations in ice cores from Antarctica (red) and Greenland (green). The corresponding profiles are bounded by the actual reconstruction results (upper envelope to shaded areas) and by the reconstructed values corrected at low values of the SN (solid curves) by taking into account the residual level of solar activity in the limit of vanishing SN (see Fig. 1). The thick black curve shows the observed group sunspot number since 1610 and the thin blue curve gives the (scaled) 14C concentration in tree rings, corrected for the variation of the geomagnetic field . The horizontal bars with attached arrows indicate the times of great minima and maxima : Dalton minimum (Dm), Maunder minimum (Mm), Spo¨rer minimum (Sm), Wolf minimum (Wm), Oort minimum (Om), and medieval maximum (MM). The temporal lag of 14C with respect to the sunspot number is due to the long attenuation time for 14C .
The underlying Be-10 proxy series for Greenland (Beer et al 1990) is shown below as excerpted from Usoskin et al (Astr Astr 2004) (I have not been able to locate a digital version of the Dye 3 data.) The series are inverted in the solar reconstruction process. I discussed the Intcal data in an earlier post (url)
Usoskin et al 2004 Fig. 1. Raw and smoothed 10Be data. The lower curves (left axis) give the raw (thin curve) and the 1-2-1 filtered 8-year-averaged data from Antarctica (Bard et al. 1997). The upper curves (right axis) show the raw and the 11-year smoothed yearly data from Greenland (Beer et al. 1990).
Solanki et al (Nature 2004) had quite similar looking reconstructions – again the two Be-10 reconstructions and a àÅ½”¬?C14, both fitted to reconstruct sunspot number.
Solanki et al Nature 2004 Figure 2. Comparison between directly measured sunspot number (SN) and SN reconstructed from different cosmogenic isotopes. Plotted are SN reconstructed from D14C (blue), the 10-year averaged group sunspot number1 (GSN, red) since 1610 and the SN reconstruction14 from 10Be under the two extreme assumptions of local (green) and global (magenta, dashed) production, respectively. The slightly negative values of the reconstructed SN during the grand minima are an artefact; they are compatible with SN ¼ 0 within the uncertainty of these reconstructions as indicated by the error bars.
The C14-based solar reconstruction is archived at WDCP here, but not any of the other versions. Here is a plot of the archived data (matching the C14-based reconstruction above.) Sunspot data is digitally archived elsewhere. (This matches the purple series above ending ~1900 AD.)
Plot of archived data at WDCP matching C14 solar reconstruction of sunspot number.
In 2006, Usoskin et al GRL 2006 re-visited this reconstruction using a different geomagnetic series to normalize the data, obtaining similar results. (Jan 16, 2007) In response to a request yesterday, Dr Usoskin forwarded a digital version of the reconstruction in Usoskin et al GRL 2006, which is re-plotted below, showing both Holocene and millennium values. HE referred me to JàÆà⻲g Beer, the author of the Dye3 Be-10 series for that data, which I’ve requested. This series goes to 1995 on 10-year intervals, with recent values among the highest on record (the 2nd-most recent value is the highest in the entire series.)à’à As a potential contributor to the Holocene Optimum, this data version suggests a solar peak after the Holocene Optimum; it would be interesting to examine the basis of this reconstruction in detail at some point.
Usoskin et al GRL sunspot number reconstruction. Left – BP; right – detail denominated in years AD. The “record” is in 1985. Plotted from emailed data.
Solanki et al (Nature 2004) was criticized by Muscheler et al 2005 in a Nature Brief Communications Arising, with Muscheler et al presenting a different reconstruction, also presented the version below in a post by Muscheler at realclimate: (See links to Muscheler articles)
Muscheler et al reconstruction (from realclimate) Figure 1: Reconstructions of the (11-year smoothed) solar magnetic modulation based on 14C data from tree rings combined with instrumental measurements (Muscheler et al.,2005). The purple curve shows the results according to a normalization using balloon-borne measurements. The black and green curves show results using an alternative normalization and the northern hemisphere (black) and globally averaged (green) 14C data. The red curve is the group sunspot number.
As long as the differences between the 10Be records are not understood, conclusions based on only one of these records should be treated with caution. Atmospheric 14C concentrations, on the other hand, are much less sensitive to a climate influence during the last 1000 years and, therefore, can provide good estimates of the history of the sun. However, the disagreement between 14C-based solar activity and group sunspot number (Muscheler et al., 2005) should remind us that the variations of the solar activity are not yet completely understood.
Usoskin replied sharply to the Muscheler criticism arguing that they had improperly normalized their data:
We conclude that by basing their normalization procedure on inappropriate data, Muscheler et al. have heavily overestimated the solar modulation parameter before AD 1950, which was further exaggerated by the nonlinear relation between Qand .Unfortunately, when performing the normalizing procedure, they made use of an inappropriate data set, from which the long-term trend has been implicitly removed (see details in Solanki et al., 2005). This resulted in a distorted long-term trend in their reconstruction. When using the correct data set (called “alternative” by Muscheler et al., 2005), they obtained the results which are very similar to ours as the purple curve in Muscheler’s Fig.1 is close to our solar activity reconstruction from 14C (see Solanki et al., 2004). All other curves in Muscheler’s Figure are affected by this normalization flaw which is seen as a constant offset between the curves (all curves are parallel to each other before 1950). Actually, this large difference of about 300 MV between purple and green curves serves as their model uncertainties. Briefly, Muscheler, using a different (unpublished and unverified) method, obtained a result which differs from our results. We don’t think that just this difference provides a substantial basis to claim that our results are wrong.
Recent discussions by Bard QSR 2006 is online here and Muscheler et al QSR 2007 is online here. In addition to discussions covering the millennium, there are interesting issues regarding the Holocene, which I’ve not touched on here, but which recur in the references.
Update: New paper by MacCracken et al here
Bard 2003 ftp://ftp.ncdc.noaa.gov/pub/data/paleo/climate_forcing/solar_variability/bard_irradiance.txt
Bard EPSL 2006 http://www.college-de-france.fr/media/evo_cli/UPL13816_Bard06EPSL.pdf
Beer, J., A. Blinov, and G. Bonani, et al., Use of 10Be in polar ice to trace the 11-year cycle of solar activity, Nature, 347, 164– 166, 1990.
J Beer, Long-term Solar Variability derived form Cosmogenic Radionuclides http://www.hao.ucar.edu/scostep/scostep11_lectures/Beer.pdf
MacCracken et al here
Muscheler et al Nature 2005 How Unusual is Today;s Solar Activity? http://www.cgd.ucar.edu/ccr/raimund/publications/Muscheler_et_al_Nature2005.pdf
Muscheler at realclimate http://www.realclimate.org/index.php?p=180
Musceler QSR 2007.Solar activity during the last 1000 yr inferred from radionuclide records http://www.climate.unibe.ch/~joos/OUTGOING/publications/muscheler06qsr_solar.pdf
Solanki et al Nature 2004 http://cc.oulu.fi/~usoskin/personal/nature02995.pdf NAture SI http://www.nature.com/nature/journal/v431/n7012/extref/nature02995-s1.pdf
Solanki et al Response to Muscheler http://cc.oulu.fi/~usoskin/personal/sola_nature05.pdf
Usoskin, I. G., Mursula, K., Solanki, S., Sch¨ussler, M., & Kovaltsov, G. A. 2002a, J. Geophys. Res., 107(A11), SSH 13-1 http://cc.oulu.fi/~usoskin/personal/2002ja009343.pdf from 1600 on.
Usoskin et al AA2003 http://cc.oulu.fi/~usoskin/personal/AG_2003.pdf
Usoskin, I. G., Solanki, S. K., Schu¨ssler, M., Mursula, K. & Alanko, K. A millenium scale sunspot number reconstruction: evidence for an unusually active Sun since the 1940s. Phys. Rev. Lett. 91, 211101 (2003). http://cc.oulu.fi/~usoskin/personal/Sola2-PRL_published.pdf
Usoskin, I. G.,Mursula, K., Solanki, S. K., Schu¨ssler,M. & Alanko, K. Reconstruction of solar activity for the last millenium using 10Be data. Astron. Astrophys. 413, 745–751 (2004). http://cc.oulu.fi/~usoskin/personal/aah4688.pdf
Usoskin, I.G., S.K. Solanki, M. Korte, 2006. Solar activity reconstructed over the last 7000 years: The influence of geomagnetic field changes, Geophys. Res. Lett., 33(8), L08103.à’à http://cc.oulu.fi/~usoskin/personal/2006GL0259211_pub.pdf