AR4 ANALYSIS SERIES Supplementary Analysis Paper #2 Solar Changes and the Climate

1. Time Scales

The sun plays a role in our climate in direct and indirect ways. The sun changes in its activity on time scales that vary from 11, 22, 80, 180 years and more. A more active sun is brighter due to the dominance of faculae over cooler sunspots with the result that the irradiance emitted by the sun and received by the earth is higher during active solar periods than during quiet solar periods. The amount of change of the total solar irradiance based on satellite measurements since 1978 during the course of an 11 year cycle is about 0.1%. This was first discovered by Willson and Hudson (1991) from the results of the SMM/ACRIM1 experiment and later confirmed by Fröhlich and Lean (1998). This finding has caused many to conclude that the solar effect on climate is negligible. But many questions still remain about the actual mechanisms involved and the sun’s variance on century and longer timescales.

The irradiance reconstructions of Hoyt and Schatten (1993), Lean et al. (1995), Lean (2000), Lockwood and Stamper (1999) and Fligge and Solanki (2000) assumed the existence of a long-term variability component in addition to the known 11-year cycle, such that during the seventeenth century Maunder Minimum total irradiance was reduced in the range of 0.15% to 0.6% below contemporary solar minima.

The cumulative energy of even the most dramatic solar energetic events during a solar cycle is miniscule compared with the TSI. The largest flare during the past 30 years was barely identifiable as a small variation in TSI data. TSI is so many orders of magnitude greater in total energy transfer to the Earth that even tiny variations can cause climate swings like the 'Little Ice Age'. Special amplification mechanisms must be postulated to produce measurable climate forcings by high energy solar events like flares, solar wind and CME’s.

Wang and Lean (2005) used a solar reconstruction model that simulated the eruption, transport and accumulation of magnetic flux during the past 300 years using a flux transport model with variable meridional flow. They suggested a radically different picture of the long term variation of solar output, most notably an increase since 1700 of only 27% of the lower end of the previously-estimated range (0.037%). In the AR4 the IPCC has embraced this finding to support its claims that there is only a small solar influence on recent climate change.  This result contrasts sharply with other estimates mentioned above, as well as Lockwood and Stamper (1999) which showed how the total magnetic flux leaving the sun has increased by a factor of 2.3 since 1901. Moreover, as the AR4 itself states (Ch 2), long-term trends in geomagnetic activity and cosmogenic isotopes, together with the range of variability in Sun-like stars (Baliunas and Jastrow, 1990) suggested that the Sun is capable of a broader range of activity than witnessed during recent solar cycles.

In addition, the AR4 conceptualization of solar forcing does not account for the sun’s eruptional activity (flares, solar wind bursts from coronal mass ejections and solar wind bursts from coronal holes) which may have a magnifying effect on the basic TSI variances through indirect means. Labitzke (2001) and Shindell (2000) have shown how ultraviolet radiation, which changes as much as 6-8% even during the 11 year cycle, can produce significant changes in the stratosphere that propagate down into the mid troposphere. The work of Svensmark and Friis-Cristensen (1997), Bago and Butler (2000) Tinsley and Yu (2002) and Shaviv (2005) and many others have documented the possible effects of the solar cycle on cosmic rays, and through them the amount of low cloudiness. It may be that through these other indirect factors, solar variance is a much more important driver for climate change than it is currently assumed. It may be that solar irradiance measurements are useful simply as a surrogate for the total solar effect.

2. Correlations with Total Solar Irradiance

In recent years, satellite missions designed to measure changes in solar irradiance, though promising, have produced their own set of problems and conflicts. Fröhlich and Lean (1998) noted that the problem is that no one sensor collected data over the entire time period from 1979, “forcing a splicing of data from different instruments , each with their own accuracy and reliability issues, only some of which we are able to account for.” Their assessment suggested no increase in solar irradiance had occurred in the 1980s and 1990s.

There are three TSI composites available, denoted ACRIM, PMOD and IRMB, each originating from the same underlying data but differing based on analysis techniques (Fröhlich 2006). Willson (2003) finds a TSI trend of 0.04% per decade during solar cycles 21 -23. Further, he finds specific errors in the dataset used by Lean and Frohlich to bridge the ‘ACRIM Gap’ between the ACRIM1 and ACRIM2 satellite experiments (1989  - 1991). Lean and Frohlich’s results arose from modifying the published results from the Nimbus7/ERB, ERBS/ERBE and SMM/ACRIM1 experiments instead of making algorithm improvements and reprocessing raw satellite data. 

Lean and Frohlich added degradation corrections to the results of Nimbus7/ERB and ACRIM1 results, which had the effect of lowering their TSI results during the solar cycle 21 maximum, and conforming the TSI time series to the predictions of Lean’s solar proxy model. Their method is not consistent with the degradation analyses published by the ACRIM1 science team.  Frohlich and Lean chose overlapping ERBS/ERBE results to relate ACRIM1 and ACRIM2 results across the crucial ‘ACRIM Gap’. Willson has argued that the ERBS/ERBE results are inferior to those of  the Nimbus7/ERB in general and specifically during the ‘ACRIM Gap’ when uncorrected sensor degradation of the ERBS/ERBE results causes lower results after the ‘Gap’ and the absence of a trend in the Lean-Frohlich composite TSI time series (Willson & Mordvinov 2003).

Not surprisingly given the uncertainty on the decadal scale, studies vary on the importance of direct solar irradiance on the longer century time scale. Wang and Lean (2005) suggest that long term solar forcing is 70% smaller than earlier thought, with no significant effect in the last half century. Lockwood and Stamper (1999) estimated that changes in solar luminosity can account for 52% of the change in temperatures from 1910 to 1960 and 31% of the change from 1970 to 1999. Scafetta and West (2006) argued that total solar irradiance accounted for up to 50% of the warming since 1900 and 25-35% since 1980. The authors noted the recent departures may result “from spurious non-climatic contamination of the surface observations such as heat-island and land-use effects [Pielke et al., 2002; Kalnay and Cai, 2003]”.

Figure 1. Running mean of USHCN mean annual temperature versus Total Solar Irradiance, 1900—2000.

Figure 2: From Soon (2005). Top: correlation between solar output and Arctic air temperature anomalies; Bottom: relationship with CO₂ is much weaker.

Kärner (2002) studied the time series properties of daily total solar irradiance and daily average tropospheric and stratospheric temperature anomalies. He showed that average temperature anomalies exhibit a temporal evolution characterized by antipersistency, in which the variance expands as the observed sample length increases on all time scales, but at a diminishing total rate. CO₂ forcing is not antipersistent, instead it has a steadily increasing trend, implying persistency. But Kärner showed that total solar irradiance is antipersistent, implying a discriminating hypothesis: the dominant forcing mechanism will endow the atmospheric temperature data with its time series property. Since the temperature series is antipersistent this implies that solar forcing dominates. The test supported this finding on all available time scales, from daily to decadal. He concluded that:

“The revealed antipersistence in the lower tropospheric temperature increments does not support the science of global warming developed by IPCC [1996]. Negative long-range correlation of the increments during last 22 years means that negative feedback has been dominating in the Earth climate system during that period. The result is opposite to suggestion of Mitchell (1989) about domination of a positive cumulative feedback after a forced temperature change. Dominating negative feedback also shows that the period for CO2 induced climate change has not started during the last 22 years. Increasing concentration of greenhouse gases in the Earth atmosphere appeared to produce too weak forcing in order to dominate in the Earth climate system.” (Kärner 2002)

3. Warming Due To Ultraviolet Effects Through Ozone Chemistry

Though solar irradiance varies slightly over the 11 year cycle, radiation at longer ultraviolet (UV) wavelengths are known to increase by several percent with still larger changes (factor of two or more) at extremely short UV and X-ray wavelengths (Baldwin and Dunkerton 2004). Palamara (1998) reports that during a solar flare, extreme ultraviolet can increase by a factor of 10 (Foukal 1998).

Ozone in the stratosphere absorbs this excess energy and converts it to heat, which has been shown to propagate downward and affect the general circulation in the troposphere. Shindell et al.(1999) used a climate model that included ozone chemistry to reproduce this warming during high flux (high UV) years. Labitzke and Van Loon (1988) and later Labitzke in numerous papers has shown that high flux (which correlates very well with UV which she notes changes 6-8% over the 11 year cycle) produces a warming in low and middle latitudes in winter in the stratosphere and then penetrates down into the middle troposphere.

Figure 3. Solar cycle 23, strong high flux second maximum around January 2002.

Figure 4. Labitzke correlation with solar flux (correlates well with UV) for stratospehere heights. She found high heights and warming at solar maxima extending down into the middle troposphere. Shindell (1999) model showed similar warming at high flux times. Bottom is the actual January to February 500 mb height anomalies for 2001 during very high flux second solar max,.

The warming that took place with the high flux from September 2001 to April 2002 caused the northern winter polar vortex to shrink resulting in an extremely warm winter in low and middle latitudes and the southern summer vortex to both contract and for a time even break into two centers for the first time ever observed. This disrupted the flow patterns and may have contributed to the brief summer breakup of the Larsen ice sheet observed at that time.

4. Geomagnetic activity, weather and climate

As early as 1976, Bucha speculated on the variations of geomagnetic activity, weather and climate. In recent years Bochnicek et al (1999), Bucha and Bucha (1998) have shown statistically significant correlations between geomagnetic activity and the atmospheric winter circulation patterns in high and mid-latitudes as controlled to a large degree by the Northern and Southern Annular modes (NAM and SAM) and modulated by the Quasi-Biennial Oscillation (QBO). They have found the tendency for the modes to be cold during the east QBO at solar minimum and west QBO at solar maximum and warm at the west phase during the solar minima and east at solar max.  This relates to the strength of the stratospheric vortex which Baldwin and Dunkerton (2004) show control the tendencies in the middle and lower atmosphere for the phase of the Arctic Oscillation (AO) and North Atlantic Oscillation (NAO). Since the QBO alternates east and west approximately each year, this would suggest a tendency for winters to alternate cold and warm near solar max or min but would not argue for long term changes.

5. Helio- and Geomagnetic activity, Solar Winds, Cosmic Rays and Clouds

A key aspect of the sun’s effect on climate may well be the indirect effect on the flux of Galactic Cosmic Rays (GCR) into the atmosphere. The hypothesis is that cosmic rays have a cloud enhancing effect through ionization of cloud nuclei. As the sun’s output increases the solar wind-induced atmospheric magnetic field shields the atmosphere from GCR flux. Consequently the increased solar irradiance is accompanied by reduced cloud cover, amplifying the climatic effect. Likewise when solar output declines, increased GCR flux enters the atmosphere, increasing cloudiness (and thus planetary albedo) and adding to the cooling effect associated with the diminished solar energy

In an excellent treatise on the geomagnetic solar factors, Palamara (1998) noted how Forbush was the first to conclude that there was a relationship between geomagnetic activity and cosmic ray decreases (called the Forbush decrease). Ney (1959) proposed a chain of events whereby solar activity influences atmospheric temperatures via cosmic rays and ionization, with the greatest effects in Polar Regions. Dickinson (1975) proposed that cosmic rays could modulate the formation of sulphate aerosols which could serve as cloud nuclei.  Tinsley and coauthors in a series of papers, proposed instead that cloud cover changes could relate to changes in atmospheric electricity brought about by ionization (see discussion in Tinsley and Yu 2004). These theories were points of contention among researchers concerning the mechanisms proposed. There was little evidence to support any of them until Svensmark and Friis-Christensen (1997) found changes of 3 to 4% in total cloud cover during the solar cycle 21.

This paper was also quickly challenged. Among the challenges, Kristjansson and Kristianson (2000) and Jorgensen and Hansen (2000) disputed the theoretical mechanisms linking cosmic rays to clouds, the latter arguing the changes in clouds might be explained by the El-Nino Southern Oscillation (ENSO) or volcanic eruptions. Kernthaler et al. (1999) repeated Svensmark’s work but included the Polar Regions, where it was though the effects would be greatest because that is where the cosmic ray attenuation was greatest.  They found that by including the polar region, the correlations were weakened.  Friis Chrisetnsen (2000) reported this latter work was based on data derived with instrument calibration issues and that with the adjusted cloud data, Kernthaler’s work could not be reproduced. Though they acknowledged some effect on cloudiness could be attributed to ENSO, they could not rule out the cosmic ray influence.

Svensmark’s work received support from papers by Tinsley and Yu (2002), and Palle-Bago and Butler (2001). The latter showed that low clouds in all global regions changed with the 11 year cycle in inverse relation to the solar activity extending over a longer period than the original Svensmark 1997 study. Usoskin et al., (2004) found a significant correlation between the annual cosmic ray flux and the amount of low clouds for the past 20 years. They found that the time evolution of the low cloud amount can be decomposed into a long-term trend and inter-annual variations, the latter depicting a clear 11-year cycle. They also found that the relative inter-annual variability in low cloud amount increases polewards and exhibits a highly significant one-to-one relation with inter-annual variations in the ionization over the latitude range 20–55°S and 10–70°N. This latitudinal dependence gives strong support for the hypothesis that the cosmic ray induced ionization modulates cloud properties.

Figure 5. The cosmic ray versus smoothed sunspot data and low cloud cover for various latitude bands according to Palle-Bago and Butler (2001).