Discussion About the Possible Effects of the Solar Activity upon the Radiation Balance
| Scientific Paper | |
|---|---|
| Title | Discussion About the Possible Effects of the Solar Activity upon the Radiation Balance |
| Read in full | Link to paper |
| Author(s) | Jan Olof Jonson |
| Keywords | Galactic cosmic ray, greenhouse effect, solar radiation, solar cycle |
| Published | 2010 |
| No. of pages | 38 |
Read the full paper here
Abstract
This paper is written as a bachelor thesis work at Stockholm University. It is a preliminary manuscript, based upon a more extensive version, shown on this website earlier. Please note the copyright rules with respect to the figures. The paper has now, after that a substantial reduction of the content has been undertaken, been accepted as a BS thesis at Stockholm's University (April 27, 2012). The BS thesis paper is available through the author.
In searching for the reasons behind the rising temperature a broad scope of potential triggering factors is currently investigated by the scientific community. Among those are the effects of extraterrestrial origin. For the time span of the last one-and-a-half century it has been shown that there is a negative correlation between the solar activity and the temperature in the Northern Hemisphere. However, beginning with the 1990's, the overall temperature rise has increased to the extent that the scientific community has felt the need to search for new models capable of explaining this new phenomenon. As a physical explanation, variations in the irradiance from the sun have also been considered, but the effects have appeared to be too small to offer a complete explanation of the observed temperature rise.
Secondary effects of the solar activity have also attained increasing interest. It has among others been assumed that the Galactic cosmic ray flux affects aerosol formation, as decreased solar activity would allow for a deeper intrusion of cosmic rays into the Earth's atmosphere, which in turn is predicted to lead to an increase of the amount of condensation nuclei. Historic records further show that increased cloudiness namely corresponds to a decrease in the solar constant. Higher amount of aerosols leads to higher planetary albedo, and, accordingly to a lower temperature.
The effects of varying cosmic rays have been estimated to be of the same order as the radiative forcing of the increase of carbon dioxide since 1750. Given our knowledge today it is still difficult to judge which the main forcing effects behind the increased temperature are. The results that have been attained tend to corroborate the assumption that Galactic cosmic rays have the effect on temperature, as proposed above. However, there is also a partial ambiguity of the results, which points to the need for further investigation of the field. How each proposed variable affects cloudiness and temperature must further be explored and a serious effort is needed to attain the 'final formula'.
Overview
This 38-page thesis is a critical literature review rather than a report of original measurement. Jonson sets himself a deliberately narrow task: to survey the published evidence for extraterrestrial forcing of the Earth's radiation balance — direct changes in solar irradiance, and the indirect route by which solar magnetic activity modulates the galactic cosmic ray (GCR) flux, which in turn is proposed to modulate cloud cover. He states plainly that the thesis "does not deal with the greenhouse effect at all"; its job is to weigh the solar/GCR literature on its own terms, not to adjudicate between it and the carbon-dioxide account.
The departure from the mainstream framing is one of emphasis rather than of physics. Where the IPCC synthesis treats solar forcing as a small and largely settled residual, Jonson organises the whole review around the possibility that an indirect, amplified solar channel exists — a chain running from solar wind and heliospheric magnetic field, through GCR shielding, through atmospheric ionisation, through nucleation of aerosol into cloud condensation nuclei, to cloud albedo and finally to surface temperature. He gives the proponents (Svensmark, Friis-Christensen, the CERN CLOUD proposal, Harrison and Stephenson, Yu and Turco, Tinsley and Yu) and the critics (Kristjánsson and Kristiansen, Sun and Bradley) separate chapters and then attempts a balanced verdict. That verdict is genuinely equivocal: the GCR effect "seems to have been sufficiently corroborated" in outline, but "the partly lacking coincidence between theory and experimental evidence" leaves the case unfinished.
The argument
The direct channel and why it is judged insufficient
Jonson first establishes the magnitudes. Solar luminosity has risen roughly 30% since the formation of the solar system, but at a constant rate that amounts to about one millionth over the last 200 years — negligible. Over a solar cycle the output varies about 0.1%, giving a change at the top of the atmosphere of order 0.3 W/m2 against a global mean total of 342 W/m2. Sunspots are ~1700 K cooler than the 6000 K photosphere, but faculae are ~1000 K hotter and cover a much larger fraction of the disk, so the net flux actually rises at sunspot maximum. Cosmic rays, by contrast, carry only about one billionth of the solar radiative input directly, and their modulation over a cycle is of order ten percent.
He reviews Thejll and Lassen's solar-cycle-length (SCL) study, which finds correlations of order 0.8–0.9 between smoothed cycle length and Northern Hemisphere land air temperature from 1861 onward — and then loses that correlation in the 1990s, a failure the authors cannot explain and tentatively attribute to human activity. He reviews Lean, Beer and Bradley's reconstruction of irradiance back to 1610 from sunspot darkening, facular brightening and the 14C and 10Be cosmogenic records, correlation 0.86 for 1610–1800, with about half of the 0.55 °C warming since 1860 ascribed to direct solar forcing — but only 0.11 °C of the 0.36 °C warming since 1970. Jonson's summary is that the direct effect is "unquestionable, but small, and insufficient".
Cloud physics and the altitude criterion
Because the whole indirect argument turns on clouds, Jonson devotes a section to microphysics. He gives the size hierarchy — ultrafine condensation nuclei with radius ≥ 3 nm, condensation nuclei ≥ 10 nm, cloud condensation nuclei (CCN) ≥ 80 nm, aerosol ~0.2–1 µm — and works through the Köhler curve for droplet growth, noting that once the critical supersaturation ratio S* is passed growth continues independently of any subsequent fall in supersaturation.
Following Hartmann's Global Physical Climatology, he reproduces the top-of-atmosphere balance
- RTOA = S0(1 − αp)/4 − F↑(∞)
and the change caused by inserting a cloud layer, ΔRTOA = −S0Δαp/4 + F↑clear(∞) − σTzct4. Setting ΔRTOA = 0 and using Tz = Ts − Γzct gives the crossover altitude at which a cloud stops cooling and starts warming: about 4 km for vanishing albedo change, about 10 km for an albedo change of 0.5. The physically important consequence, which the rest of the thesis relies on, is that low and middle clouds cool, high cirrus warms — so any GCR effect must act on low cloud to produce cooling.
The proposed GCR–cloud mechanism
The mechanism Jonson assembles from Tinsley and Yu, Yu and Turco, and Masarik and Beer runs as follows. GCR primaries — 90% protons, 9% alpha particles, energies around 1 GeV — initiate hadronic cascades whose secondary muons deposit ionisation deep in the troposphere. Ionised O2 and N2 react with H2SO4, H2O, NH3 and organic vapours; the resulting "core terminal ions" accumulate ligands and grow as charged clusters. The key physical claim is that the collision kernel for a charged embryo is larger than for a neutral one, so ion-mediated nucleation (IMN) outruns binary and ternary homogeneous nucleation — which, as Yu and Turco note, underpredict observed ultrafine aerosol by a factor of ten. A second route, electroscavenging, uses image charges on falling droplets to sweep charged aerosol across streamlines faster than neutral aerosol.
Quantitatively, ionisation rates run ~20–30 ion-pairs cm−3s−1 in the upper troposphere and ~2 at ground level. Yu and Turco simulate that a 25% change in background ionisation — a typical solar-cycle range — produces about a 16% maximum change in the number of particles larger than 3 nm under marine boundary layer conditions.
The observational case
Svensmark and Friis-Christensen (1997) reported a 3–4% variation in global cloud cover over a solar cycle tracking the GCR flux, using ISCCP-C2 and later D2 satellite data and Climax neutron monitor counts. Svensmark (1998) added the decisive-looking detail that cloud cover follows the GCR curve rather than the 10.7 cm solar radio flux, from which it lags by 1.5–2 years. The CERN CLOUD proposal (2000) estimated the correlated area fraction of the Earth at 29.6%, with a 0.01% probability of arising by chance, found the correlation vanishing above 3 km altitude, and derived a net GCR radiative forcing of 1.2 W/m2 over the last century against an observed 0.6 °C rise. Svensmark's own forcing estimate gives ~0.1 °C from irradiance versus 0.3–0.5 °C from the cloud channel.
Jonson treats Harrison and Stephenson as the strongest independent support, precisely because it uses a different observable: UK surface measurements of the diffuse fraction (DF) of solar radiation, 1951–2000, against Climax neutron counts. On high-neutron days the diffuse fraction rises by 2% and the odds of an overcast day (DF > 0.9) by 19%, significant at the 0.1% level; the measured temperature sensitivity is about −0.2 K per 0.01 change in DF.
The case against
Kristjánsson and Kristiansen find no physical mechanism for GCR-enhanced low-cloud growth, arguing that GCR ionisation peaks at 10–20 km and falls off rapidly below. Kristjánsson et al. separately find solar irradiance correlating with low cloud better than GCR does, with coefficients of −0.6 to −0.8 — though with only 198 monthly points, reduced by Quenouille's method to an effective 4–13 independent points, forcing them to use Ebisuzaki's non-parametric significance test. Sun and Bradley attack Svensmark for restricting the analysis to oceans: over the contiguous US, China east of 110°E and the USSR south of 60°N the coincidence is very weak, the Atlantic correlation weakens badly under extended ISCCP data, and 1953–1995 ship observations show none at all.
Jonson's own contribution in the discussion chapter is to note a specific data problem: the CLOUD proposal's figure 3 and Kristjánsson and Kristiansen's figure 9 show different cloud-cover curves from 1992 onward, with an upward jump in the former plausibly caused by the C2-to-D2 dataset change in 1990 (new satellites, instrumentation and retrieval algorithms), delayed two years by the 12-month running mean. He concludes that "the divergence of the GCR curve and the cloud curve from around 1992 constitutes an obstacle to the GCR proponents." He also offers a defence of the missing continental correlation: land dust carries its own radioactivity and CCN counts over land are far higher, so a small GCR increment would be buried in the noise.
Assessment
The genuine merit of the thesis is its refusal to be partisan. Jonson names the mechanism that would have to work, states the magnitude it would have to reach, and then reports the numbers that do not fit — including, notably, numbers that damage the side he is evidently more interested in. The identification of the C2/D2 dataset transition as a candidate artefact behind the post-1992 divergence is a real piece of critical reading rather than a rhetorical move, and the altitude criterion derived from Hartmann is the right physical constraint to impose: it forces the debate onto low cloud specifically, which is where the observational disagreement is sharpest. His summary of Harrison and Stephenson is also well chosen, since the diffuse-fraction record is methodologically independent of the satellite cloud retrievals that both camps otherwise fight over.
The difficulties are of two kinds. The first is that a review can only be as decisive as its sources, and Jonson's sources do not settle the question — so the thesis ends where it began, with "much research is necessary". The correlations it reports are mostly one solar cycle long, which he himself flags as the fatal weakness: a decade of data cannot separate a GCR signal from any other decadal forcing. The second is that several links in the chain are asserted rather than closed. Ion-mediated nucleation is shown to produce more ultrafine particles, but the step from 3 nm particles to 80 nm CCN — a factor of roughly ten in diameter, and the step that actually matters for cloud albedo — is described as needing "growth" without a quantitative growth rate under realistic conditions. The land/ocean asymmetry is explained twice, once by radioactive dust noise and once by high background CCN, but neither explanation is turned into a prediction that could be tested against the continental data that falsified the correlation.
There is also a tension the thesis does not resolve. Kristjánsson finds solar irradiance correlating with low cloud better than GCR; Svensmark's lag argument uses the 10.7 cm flux to argue the opposite. Both cannot be right, and Jonson reports them side by side without adjudicating. Finally, the framing question — whether a GCR channel could substitute for greenhouse forcing rather than merely add to it — is deliberately excluded by the thesis's own scope, yet the 1.2 W/m2 century-scale figure quoted from the CLOUD proposal is precisely a claim about that substitution, and it is quoted without the caveat that Lockwood's own later work found the heliospheric flux trend reversing after about 1985 while temperatures continued to rise. On the thesis's own terms, however, the work is sound: it does not overclaim, it reports its opponents accurately, and its conclusion — that the direct effect is real but small, and the indirect effect plausible but unproven — is what the evidence surveyed actually supports. The CERN CLOUD chamber experiments that Jonson anticipates in his "Outlook" have since been performed, and their results (nucleation enhancement real, but too small at CCN sizes to account for the claimed cloud modulation) fall broadly on the cautious side of his own summary.