Solar irradiance (or the flux of the Sun’s output directed toward Earth) provides virtually 100% of the energy received and absorbed by the Earth system. So it shouldn’t be any surprise that if you open up an elementary climate change textbook and search for suitable candidates that can actually cause the global climate to vary in time, changes in sunlight will be on the list (along with volcanic eruptions, land use changes, anthropogenic greenhouse gas emissions, etc).
When one actually targets specific past climate change “events,” however, changes in sunlight rarely enter the picture in any appreciable way. It is not self-evident that this should be the case, but we seem to be rather fortunate in orbiting a well-behaved star that only exhibits changes in output on the order of a tenth of a percent or so on decadal-to-centennial timescales. See the record below, for instance. Various choices for solar forcing used by groups in the recent (CMIP5/PMIP3) generation of climate simulations for the last millennium can be found in Figure 8 of Schmidt et al., 2011. Note the vertical scales. We’ll come back to some of the features in the plot below.
In fairness, records of solar activity do not go back very far, but this stability is the intuition we have gained from the evidence that does exist over the the ~10,000 year Holocene epoch (from sunspot counts or cosmogenic isotope proxy data- the most common isotopes for paleo-solar activity used are 10Be in ice cores and 14C in tree rings. This review paper is informative on mechanisms of solar influences on climate and some detail on inferring past solar variability). There’s no obvious reason (at this to this person who studies climates and not stars) that solar output couldn’t have exhibited much larger, transient fluctuations in the very deep past. But this hypothesis is not necessary and is insufficient to understand any “first-order” details in the never-ending jigsaw puzzle of paleoclimatology.
There’s little doubt that on much longer timescales, the Sun is slowly getting brighter (by ~7-8% or so every billion years) and its luminosity will continue to increase as it evolves on the “main-sequence” stage of its life toward becoming a red giant. On this timescale, however, Earth has a clever thermostat that is capable of outpacing solar changes and keeping the Earth’s climate in a relatively narrow temperature range. “Narrow” is rather loose in this context, but I have in mind the stability range of liquid water over virtually all of Earth’s history. When silicate-rich rocks and minerals from the Earth’s interior are uplifted and exposed to surface, they dissolve due to chemical weathering. The slow processes that react CO2 with silicate rocks and bind it up in carbonate rocks acts as a long-term sink for CO2, and it turns out that this weathering rate is dependent on the climate (via precipitation and runoff). This gives rise to a negative correlation between temperature and the eventual CO2 concentration on geologic timescales (This doesn’t mean Earth’s temperature should relax to some constant value on the million year timescale, since the thermostat can have different “settings” depending on volcanic outgassing rates or the availability of rocks that can be weathered, which depends on things like mountains and continental distribution). The slow nature of this thermostat, however, does little to ensure that Earth’s climate cannot vary considerably on shorter timescales such a glacial-interglacial fluctuations (on this timescale, CO2 concentration seems to be a positive feedback to temperature by virtue of the oceans rather than the rocks). But the point is that the Sun varies by very small amounts, or slowly enough for the thermostat to kick in.
As the opening figure shows, however, the Sun can vary by some amount over decades, and for climate studies we need to quantify this. There is a rather robust quasi- 11 year sunspot cycle for example, which is also quite readily expressed in proxies of solar activity. By “solar activity” we mean variations in surface features and those related to the structure and magnitude of the solar magnetic field, such as coronas or solar flares. Much like a “flipped” Earth, the Sun has a “radiative layer” (like Earth’s stratosphere) and a convective layer (as in Earth’s troposphere) whereby energy developed in the core of the Sun via nuclear reactions is transferred across the radiative zone by radiative transfer, but at some distance from the core is then transferred to the stellar surface by convection (hot gas rising and cool gas sinking). At Sun-like temperatures and densities, heavy elements like silicon and iron lose an electron (thus allowing free electrons roaming around in the solar material, or plasma) making for a good conductor of electrical currents. Magnetic field lines thus follow plasma flows, which can be distorted or “wound up” by the differential rotation of the star (i.e., the equator rotates faster than the poles); solar activity appears to be directly connected with the properties of this field. Magnetic fields traverse the convection zone before they reach the photosphere to form the observed solar active regions, and seem to act as “heat blocks” for the convection. Sunspots themselves are actually “dark” and colder than the surrounding solar surface, but large bright prominences called “faculae” are also associated with higher solar activity (higher sunspots) and observations show that faculae increase irradiance considerably more than the sunspots decrease irradiance from the Sun. So high sunspot counts tend to correlate with higher total solar irradiance, as in Figure 1 (panels a and f).
One often sees very exciting headlines about sunspots and solar activity in the context of Earth’s climate. The canonical example of a situation where reductions in solar output may have influenced climate is during the so-called “Little Ice Age (LIA),” which is not well defined historically but loosely encompasses the time period from 1400-1700 or so. It is coincident with certain minimums in solar activity (e.g., the Maunder Minimum beginning in the mid 1600’s). As with any discussion of pre-industrial climate change during the Last Millennium, one must bear in mind that this LIA is much smaller in amplitude than, say, the Last Glacial Maximum (over 20,000 years ago) when ice sheets covered much more of North America and Eurasia. There may have been global changes of only a few tenths of a degree, preceded by a Medieval Climate Anomaly (MCA) and with perhaps much larger radiative changes from volcanic eruptions than from the Sun. Efforts to model last millennium climate with multiple and single forcings such as this CESM LME project generally suggest these eruptions are more important (and probably require more robust records) than any other forcing during this timeframe. There is also internal variability at play, with complex spatio-temporal expression of climate change across different regions, demanding careful interpretation of what we mean by “MCA” or “LIA.”
Astrophysicists do not fully understand sunspot cycles and predictions of solar activity have come with mixed degrees of success. One such prediction of near-future activity has generated lots of headlines drawing analogy to the LIA and predicting a looming “mini ice age.” This prediction comes in the form of reduced solar cycle amplitude and possibly irradiance, based on a statistical model of magnetic activity using past data. One can suspend judgment on the robustness of their method and ask questions about the climate impact, even if “Maunder Minimum” type solar levels are borne out in the near future. The typical solar cycle amplitude is on the order of ~1 W/m2 and the difference from Maunder Minimum to present is around 1.5 W/m2 or so. Figure 2 allows you to eyeball some of the secular trends. These changes in total solar irradiance can be converted into a radiative forcing by dividing by a spherical geometric term and multiplying by the absorbed component (~70%) of the incoming radiation so 1.5*0.7/4 ~ 0.26 W/m2. This radiative forcing is the “common currency” by which we can compare solar trends to other active forcing mechanisms for a sense of the relative scale. Given the prediction of 40-80% reduction in solar cycle amplitude, the implied reduction in forcing is around 0.1 W/m2 or so, similar to just a few years of CO2 buildup. Thus, one should not worry about a “little ice age version 2” given this prediction. It is, at best, a second-order fine tuning knob on the problem of contemporary climate change.
So the big picture, allowing one to place the prediction and media headlines in context, is that the small amplitude of the solar forcing does not compete with the historical or expected anthropogenic forcing on the climate system. The other point, much like the one that crops up with geo-engineering the planet by sulfate aerosols, is that the length of a few solar cycles are much shorter than the timescale over which CO2 remains perturbed in Earth’s atmosphere. The CO2 remains elevated until the very slow weathering processes described previously take out the long tail of elevated CO2 that the oceans are unable to, due to the exhaustion of the carbonate ion that (like a Tums tablet to treat your stomach acid) acts as an anti-CO2 agent when carbon invades the ocean. This disparate timescale problem of CO2 drawdown is fascinating, but it follows that if one thinks that short-term mitigation strategies or hoping for a small reduction in solar activity will buy time for tackling the CO2 problem, you’re only ensuring much larger peak warming in the future.