Solar activity not a key cause of climate change, study shows

Climate change has not been strongly influenced by variations in heat from the sun, a new scientific study shows.

The findings overturn a widely held scientific view that lengthy periods of warm and cold weather in the past might have been caused by periodic fluctuations in solar activity.

Research examining the causes of climate change in the northern hemisphere over the past 1000 years has shown that until the year 1800, the key driver of periodic changes in climate was volcanic eruptions. These tend to prevent sunlight reaching the Earth, causing cool, drier weather. Since 1900, greenhouse gases have been the primary cause of climate change.

The findings show that periods of low sun activity should not be expected to have a large impact on temperatures on Earth, and are expected to improve scientists’ understanding and help climate forecasting.

Scientists at the University of Edinburgh carried out the study using records of past temperatures constructed with data from tree rings and other historical sources. They compared this data record with computer-based models of past climate, featuring both significant and minor changes in the sun.

Solar activity not a key cause of climate change, study showsThey found that their model of weak changes in the sun gave the best correlation with temperature records, indicating that solar activity has had a minimal impact on temperature in the past millennium.

The study, published in Nature GeoScience, was supported by the Natural Environment Research Council.

Dr Andrew Schurer, of the University of Edinburgh’s School of GeoSciences, said: “Until now, the influence of the sun on past climate has been poorly understood. We hope that our new discoveries will help improve our understanding of how temperatures have changed over the past few centuries, and improve predictions for how they might develop in future. Links between the sun and anomalously cold winters in the UK are still being explored.”

Solar activity not a key cause of climate change, study shows

The material in this press release comes from the originating research organization. Content may be edited for style and length. Have a question? Let us know.


One email, each morning, with our latest posts. From medical research to space news. Environment to energy. Technology to physics.

Thank you for subscribing.

Something went wrong.

16 thoughts on “Solar activity not a key cause of climate change, study shows”

  1. fnredbortz I hope you have a good weekend with your children. After I finish up a few projects I think I will take a look at testing my hypothesis.

    I agree that taking the solar energy as a constant makes sense in the short term. Starting with a static model the coefficients of solar and green house gases can be assigned. Then varying green house gasses provides a 100 year forecast. It is the solar coefficient that I suspect is hard to determine. Effects of green house gases depend on the solar energy as they function by trapping solar energy, so determination of solar contribution to temperature is important to predict future temperatures.

    Huyer’s observation of orbital variations gives us the relationship between orbital change and temperature change. He also observed that we are at a minimum in the orbital cycle, an ice age. This means the temperature difference between normal ice age and today needs to be accounted for. If due to human activity, almost certainly, then increased green house gasses account for a large temperature increase. This should make for a good starting point as a static model. Orbital variations are slow and unlikely to have any effect over 100 year projections.

    Obviously there is debate about effect of much shorter Maunder minimum associated solar cycle on current environment. I will look into studies that shed light on measurement of solar activity for this. It has a period in a few hundreds of years so may matter to a 100 year prediction. If the Maunder minimums can be predicted accurately then they will likely make a difference in a model. Not just the solar output change observed in the past. But rather that change multiplied by the increase in greenhouse gases in the present. That multiplier effect may be very important in the 100 year predictions of temperature. Only running the simulations will tell.

  2. One more thought on this. In my middle-grade book Seven Wonders of Exploration Technology (Twenty-First Century Books, 2010,, Chapter 2 is entitled “Exploring Earth’s Climate.” It describes what goes into climate models and how they are used.

    I write the following
    (Copyright 2010, Alfred B. Bortz, All Rights reserved):

    Scientists use climate models to study how Earth’s weather changes as conditions in the atmosphere change. They can even make predictions about changes that haven’t happened yet. The programs need these facts and figures about Earth:
    • A map of its landforms, oceans, lakes, and streams;
    • The energy reaching it from the Sun every second;
    • The gases in its atmosphere;
    • The tilt of its axis, which causes seasonal changes;
    • The length of a day and a year;
    • Weather or climate conditions at the start of the prediction period; and
    • Other details that affect the prediction of future weather, such as dust or pollution in the atmosphere.

    Putting that in the context of this discussion, depending on the length of time a model covers, those are all variable, including the position of the continents. If you are studying only a century or so, you need to consider known solar variations as they follow an 11-year cycle.

    If you are studying thousands of years, short-term solar variations are not important, but changes in the shape of Earth’s orbit and variations in its tilt are. If you are studying millions of years, then continental drift needs to be included.

    The place where many discussions go off track is trying to apply conditions that are appropriate for one time-scale of an “experiment” (or scenario) to a simulation at another time scale. Climate change deniers are particularly adept at pulling that trick. They conflate time changes on geological time scales with effects that might be happening over the coming century.

    Even though my book and chapter are aimed at 6th-8th grade, I hope they provide enough insight that my readers can recognize when someone is trying to fool them with a specious argument.

    So Lee and I aren’t disagreeing about the value and validity of models. We are only discussing whether certain quantities need to be treated as constants or variables in those models. You can treat them all as variables, but there is a computational cost. I’d rather put my computational resources into looking at variations that matter on the timescale of my “experiment.”

    And now I will take my leave of this discussion. My grandchildren are here for the weekend!

  3. A point of disagreement, that may be useful to other readers:

    Lee writes: “My contention that things like orbital variations and Maunder minimums need to be included is because they are part of the contribution to global temperature.”

    I disagree that orbital variations need to be included if you are comparing different scenarios at the same time. All you need is a proper set of parameters to describe the insolation. It’s a waste of computer resources to calculate something that isn’t going to vary from one scenario to another. (Such as a 21st century comparison of the effects of CO2 content.)

    As for Maunder minimums, at this point, it is still hypothetical that a near absence of sunspots causes cooling. So building that into the model is assuming an effect that may not exist. It might make an interesting way to compare scenarios to actual historical data, if that’s what you want to study.

    I’m all for making models as complete as necessary, but if we want them to produce useful results, we shouldn’t load them down with extra computation. Models are like experiments. It’s usually better to control all the parameters that you are not studying if at all possible.

    Models have the advantage that you can have an absolutely stable sun if you want to study the contribution of atmospheric concentrations, *all else being equal*.

  4. I see I touched a nerve. Rest assured, I am not a climate change denier. There is plenty of evidence to show ocean warming for example. I have personally observed shifts of warm water fish like Bluefish to more northerly waters. The increase in Cod parasites typical of warm water is another. Permafrost thawing and new methane seeps from methane hydrate ocean deposits are other widely published examples.

    My contention that things like orbital variations and Maunder minimums need to be included is because they are part of the contribution to global temperature. Their contribution to global temperature is at a LOW. This means CO2 may have MORE heating effect to account for the warm temperatures we have now. They support even more aggressive action to deal with CO2 and other greenhouse gases, not less.

    I think Dr. Huybers sums it up nicely: “Another important aspect to consider is that the orbital configuration we now have is almost exactly where it was 20,000 years ago, during the Last Glacial Maximum, but this time we’re near a glacial minimum,” he said. “If you think about what the difference is between then and now, it’s not the orbital configuration, it’s the CO2. I think that’s important to keep in mind, because it shows that glacial changes are not a simple function of the orbital configuration.”

    Let us not forget that models must be tested by comparison with known data. As an engineer working on control systems I have used Mathlab models time and again. Past data is the only thing we have to validate a model, be it climate or other physical system. If a model can not show itself consistent with observed data then the model is incomplete. In the case of climate discussions an inability to account for observed past temperatures will be used as a weapon against climate change. In the case of engineering people die when a system does not behave as predicted. Oil refinery explosions for example.

    Yes, people need to be educated about how a model works in order for us to persuade them that the future is indeed going to be hotter. Unfortunately very few will be capable of understanding. What they understand are predictions that either come true or fail to happen. This is why models must be accurate, it is the only tool to persuade people to take warming seriously.

    • Actually, you didn’t touch a nerve. I could tell we are both on the same side of the issue.

      But I am probably more sensitive to the kinds of things that climate change deniers pick up on, and I am therefore more careful about my language. (I’m a writer of science for children. See for details.)

      We need to address those deceptions, not repeat the language of deception, such as the so-called “pause,” and stating that a model is “incomplete” as if it could ever be complete.

      My comments are addressed to the casual readers of this thread who could easily use that deceptive language to conclude the wrong thing about climate science and climate models.

      So please consider my comments to be illuminations, not argumentative. I hope this helps put our conversation–which is a good one–in the proper perspective.

  5. Lee also notes that “Models that failed to account for the recent warming pause do not help matters either.”

    That “pause” is not a pause. Most people who discuss it usually base their argument by starting their trend line in 1998, which was an anomalous peak year for reasons that are now understood.

    In fact, it is not yet clear the rate of increase of air temperature is slower–we don’t have a long enough time-frame to draw statistically sound conclusions. But even if the air temperature is decreasing more slowly, the ocean temperature is not.

    Again, the issue is how much detail you put into the model and for what purpose. Although Lee and I agree on the importance of this issue, I am calling attention to his phraseology, which inadvertently echoes that of the deniers.

  6. A very quick observation: The obsession with increased global temperature is doing a great disservice to all. Warming is NOT required to show that elevated CO2 levels are affecting the climate. (CO2 levels due to humans is undisputed.) Things like a new Maunder minimum merely gives a short reprieve. Educating the public on this to account for a pause in warming is difficult but needs to be done. Models that failed to account for the recent warming pause do not help matters either. The public sees such failure as proof that CO2 warming does not exist.

  7. Another thing that models need to account for is the history of rapid climate change. Ice cores show that a large fraction of temperature change happened in a decade. A 15 degree jump in 10 years would devastate humanity if it happened now. And it could. Non-linear response to temperature change is already identified. For example release of methane from thawing permafrost and methane hydrate. Methane being a far stronger green house gas than CO2, thus setting up a positive feedback to cause rapid temperature increases. See

    We need models that account for this historical record. Otherwise we will be unprepared for sudden change And the models will be inaccurate.

  8. Fredbortz is absolutely correct that climate change is superimposed on top of natural change. It is important to account for the natural change to judge the effects of things like CO2 emissions. For example Huybers made the off hand comment “Another important aspect to consider is that the orbital configuration we now have is almost exactly where it was 20,000 years ago, during the Last Glacial Maximum, but this time we’re near a glacial minimum,” It is entirely possible that accounting for orbital variation, solar output, etc will show human activity contributes MORE to current temperatures than currently thought.

    I point out these contributors to climate change not to support or attack CO2 model of climate change. I merely am pointing out that there is much more involved than just human CO2. The models must be able to account for cyclic climate changes going back thousands of years, otherwise they are useless going forward.

  9. Any climate model that fails to include the proven effects of orbital variations is incomplete. Orbital variations are proven to be very important for major climate change, like ice ages or the ending of ice ages. It is not clear if this study included that or not. Interestingly Huybers points out that warming due to CO2 appears to be preventing a new ice age from happening. Maybe global warming isn’t so bad after all. Human activity appears to be countering normal climate change that would kill millions of people.

    Researchers have found that orbital variations of the earth account for differences in crust thickness at mid ocean ridges. Places where new crust is formed as plates separate. What makes this relevant is that it is climate change by orbital variations that affects crust thickness. This is through the differences in weight of the oceans above. During an ice age ocean levels can be reduced by about 200 meters. See

    This gives additional evidence in support of orbital variations driving major climate change, ie. ice ages. Already investigated by researchers like J.D.Hayes, Milankovitch, Huybers et al. See,

    The cyclical nature of climate change has lead to success of a rather odd approach. Empirical evidence is used to identify underlying cycles. Effectively a Fourier transform on the data to extract the individual oscillations. This then correlates with planetary movements. An introduction to Scafetta’s work can be found here:

    • Lee is correct when he states “Any climate model that fails to include the proven effects of orbital variations is incomplete.”

      However any model is, by definition, incomplete. The purpose of a model is not to cover all possible situations, but rather to enable a scientist to look at a specific subset of scenarios. You need to choose what features to include in the model based on the questions you intend to address with it.

      If you want to model what the climate would be like with various levels of CO2 production rates in the rest of this century, you don’t have to consider orbital variations. You just use the orbital parameters that exist today (or will exist in 2050) since the change over 100 years is negligible.

      One of the biggest problems with the political discussions surrounding climate change is that people don’t understand the way models are constructed and used. By leaving out features that don’t matter in the scenarios of interest, you can devote more computing power and data to the features that do matter.

      Yet those who want to muddle the discussion will say that the model leaves out x, y, or z, and that makes it unreliable. That’s pure hogwash but it fools people who don’t understand how the model is designed to work.

  10. It would be good to remember that volcanic events result in relatively short term changes in climate. If I recall correctly from 1 year to 10’s of years. Solar output, as measured by sun spots, has a cycle in hundreds of years and an 11 year cycle. It is this longer period change that can not be very well tested with a model over only 1000 years. It is not stated if this study even included the long term cycle.

    The recent low activity in 11 year peak has been recorded only once before by humans. That was 1645 to 1710. It corresponds to the Maunder Minimum, aka “little ice age”. What is most interesting is that it is the long term changes in solar output that are associated with climate change. The 11 year cycle does not have much effect, as far as I know. Which would be in agreement with the study.

    Temperature data is very noisy, it must be filtered to reveal any long term changes. If you do not then then any correlation to long term solar changes will not be found. Consider the time frame of various ice ages as an example.

  11. The late Devendra Lal did some very good work on solar output. Way past the short 1,000 year period used by this study. In particular he showed how isotopes can be used to measure solar output for the past 35,000 years. He was able to show that solar activity is associated with known climate changes. This is especially important as humans have been able to record solar activity via sunspots for only a few hundred years. Do a search for “Devendra Lal solar” to find some of his work.

    • We need to be very careful in this discussion. Whatever natural cycles are doing, the human production of greenhouse gases is superimposed on it. Thus Dr. Lal’s work, no matter where it leads, does not suggest ignoring the large human contribution to changes in climate that has been clearly identified.

      That is what I was discussing in my earlier Science Blog post. (

      That post asks how we should respond to the possibility that changes in solar activity may be leading to a cooler period. Part of my discussion there is as follows:

      My answer is a resounding “No,” and here’s why:

      (1) Correlation does not mean causality. The Maunder minimum may have been responsible for seven cool decades in the 17th-18th century, but there is no consensus on the physics of how changes in solar activity (and consequent changes in Earth’s ionosphere) could have resulted in global cooling. It may well have been coincidence. But even granting that it was causal does not change my position, as the subsequent points explain. [A quick look at Dr. Lal’s presentation suggests a possible mechanism for causality. Further research may confirm, clarify, or dispute his work, but that is not relevant to the points that follow.]

      (2) We are not certain that we are on the verge of another Maunder minimum. Even if we are, it will likely only cool the Earth for a few decades while the effects of increased greenhouse gases in the atmosphere will continue beyond that.

      (3) What does the increased level of greenhouse gases imply? Analysis of the large body of climate data has led to a strong consensus among climate experts that human activity which has increased atmospheric CO2, primarily the burning of fossil fuels, is responsible for most, if not all, of the global warming that has occurred in the last century. Even if sunspot activity temporarily mitigates the increase in global temperature, we still need to be concerned about the projected increases in the next few centuries.

  12. Two points:

    1. This article dovetails with my Science Blog article, first written in 2010 and re-posted with minor changes in 2011. (

    2. I dispute Oliver K. Manuel’s comment. The sun’s makeup was definitively shown to be largely hydrogen and helium in a breakthrough thesis by Cecelia Payne (later Payne-Gaposchkin) in 1925, not 1945. Her work conflicted with the earlier “iron sun” theory of her thesis professor Henry Norris Russell and others, who accepted that her spectral data was consistent with her hypothesis but insisted she add a disclaimer in her thesis that there may be some other factor at work.

    In 1929, Russell acknowledged that Payne was correct [Henry Norris Russell (1929). “On the Composition of the Sun’s Atmosphere”. Astrophysical Journal 70: 11–82. Bibcode:1929ApJ….70…11R. doi:10.1086/143197], and became a strong backer of her conclusion without the disclaimer.

    See Bortz, Fred, Physics: Decade by Decade, Twentieth-Century Science Series, Facts on File, 2007, pp. 63-65, which includes the following excerpt: “As soon as [Russell] saw other research that bolstered Payne’s original conclusions, he became a powerful advocate of her work, which ultimately led to a deep understanding of the life cycle of stars and the origin of the elements.” (

    P.S.: Oliver Manuel will defend the “iron sun” theory until he dies, and he may dispute my statements about it here, too. I have learned that there is no value in engaging with him on this topic, so I will not respond to him further. This is not a personal attack, just a statement of fact based on previous interaction.

  13. Oh what a tangled web we weave,
    When first we practice to deceive!

    The current inability of humans to deal with nature effectively is the direct result of decisions made to deceive the public in 1945-46:

    1. The internal composition of the Sun was changed from mostly iron (Fe) in 1945 to mostly hydrogen (H) in 1946.

    2. Textbooks replaced Nobel Laureate Aston’s rigorously valid “nuclear packing fraction” with von Weizsacker’s convincing but deceptive “nuclear binding energy” after the Second World War.

    Best wishes for the New Year,
    Oliver K. Manuel
    Former NASA Principal
    Investigator for Apollo

Comments are closed.