Data Suggest That Solar Wind Impacts Global Temperature

By Ed Caryl

On May 17th, Willis Eschenbach posted “Sailing on the Solar Wind” on WUWT. In it he used a graph of the daily solar wind data since 1963. I wish to thank Willis for his effort in tabulating the daily data AND for making it available for the rest of us.

That graph had some shapes that were hauntingly familiar. Here it is, repeated, but with the vertical axis reversed. Less solar wind is warming, more is cooling.

Figure 1 is a plot of the daily solar wind data since 1963, courtesy of Willis Eschenbach, here.

I saw two large white-space notches in this plot, one at 1996-7, and another at 2008-9, where the solar wind decreased for long periods. These are just before the two recent El Niños of 1998 and 2010. I downloaded Willis’s data and converted it to monthly averages so that it could be compared to the various available temperature data. Here (Figure 2) is that data with a three-month centered average applied to both the solar wind data and the Hadley Center HADSST3 sea-surface temperature.


Figure 2 is a plot of the solar wind (inverted) and sea surface temperature since 1963.

Note the similarity. The solar wind acts similarly to the cosmic ray impact described in Cosmic Rays and the Impact On Climate. A decrease in solar wind occurs just prior to an increase in temperature. This probably happens because the solar wind acts in a similar fashion as CME’s in producing Forbush decreases in Neutron flux. Both produce an increase in cosmic ray production of cloud nuclei. But why the delay? Several temperature data bases were investigated and compared, using one-month delay steps and computing the R-squared value in Excel for each comparison. Figure 3 resulted:


Figure 3 is a plot of the response to a solar wind change (reduction) on various temperature databases. All temperature data were obtained from and de-trended.

Two questions are suggested by this data. Where is the initial temperature change taking place? And why is there a delay? The answers seem to be: 1) The initial warming is in the tropics, specifically in the central Pacific in the El Niño region. In that area, there is no delay, the warming occurs immediately and for the following 3 or 4 months. (The green trace.) The other sea surface temperature data base, HADSST3, and its 3-month average, also reflect the initial immediate warming, but also have a peak at 8 months after, probably for those areas away from the tropics that are warmed later, either by moving currents, or by moving atmosphere.

The land data, HADCRUT4, shows a peak at 8 months and again at 11 months. The reason becomes a little clearer when the hemispheric components are examined. The southern hemisphere reacts quickly then drops off after 8 months, probably because the oceans react faster than the land. The northern hemisphere reacts at 8 months and 11 months because the northern hemisphere has greater land mass.

I recognize that this relationship is speculative. But the relationships seen in Figure 3 are consistent and devoid of “noise”. Delays outside of the range shown have consistently very low R-squared values. It is suggested that this research area bears further study.


13 responses to “Data Suggest That Solar Wind Impacts Global Temperature”

  1. John F. Hultquist

    It might be that I am looking at Figure 2 and expecting to see a closer correspondence than is there. The dates are not clear but near 1964 -65 the SST drops while the SW goes up to ~400. About 1975 the curves go in opposite directions. Somewhat the same in ~1984. 2003 and 2011-12 also.
    I do see “the similarity” for much of the record but the episodes indicated above need to be explained if this relationship is to be useful.
    Or I have totally missed what the figure is supposed to do.

    Also, in the next to last paragraph you say “… probably because the oceans react faster than the land.” This seems not to fit with conventional statements of land and water reaction to energy inputs. Again, I may have missed the point.

  2. Ed Caryl

    There are obviously many things going on. The sun drives things on many levels: solar wind, CMEs, radiation at all wavelengths, and magnetic fields. Then there are all the various ocean cycles that interact with those drivers. Then throw in volcanic activity. No one thing drives temperature.

    The ocean surfaces absorb solar radiation and transfer that heat to the atmosphere. In the tropics this heat is the heat of vaporization of seawater. That reaction is very fast. Heat transfer to the deeper layers by currents and wind-driven mixing, is much slower. That may be where some of the delay occurs.

  3. Walter H. Schneider

    Ed, are you arguing against Willis Eschenbach’s conclusion?

    He stated, “….So please note that in this case, I am NOT saying that solar wind has no effect on the climate. It may well have such an effect, although the lack of any 11-year signal in temperature datasets argues strongly against it.

    What I am saying is that the manifold flaws and problems in the study of Zhou, Tinsley, and Huang 2014 mean that they are a long ways from demonstrating that such a purported effect actually exists.”

    1. Ed Caryl

      Walter, not really. there doesn’t seem to be any 11-year or 22-year pattern in the solar wind. That pattern is in the magnetic field that is embedded in the wind, and it doen’t appear to have much if any effect.

  4. TedM

    Interesting post Ed. The lack of correlation at a couple of points may be because of the different effects of a positive or a negative Bz at the magnetopause. This is something I don’t think Willis understood. At least that’s how I read his post.

    Also important to remember is that in the natural environment there are not just single causes and effects.

    1. Ed Caryl

      Ted, the earth’s climate is like an elephant. Each input is a body part. The solar wind is like the left hind leg, perhaps just the knee. Sunspots might be the other hind leg. Methane is the farts. CO2 is the breath. Water is the bloodstream. The ocean is the intestines. Rain forests are the lungs. Things on the inside we know little about. Things on the outside have not all been identified. Right now, this is about all we know about the elephant.

  5. Peter Azlac

    As you so rightly say the impact of the solar wind on climate is complex because it can vary with solar cycles but also in the short term with solar flares. But overall the impact on the Earth’s climate involves the sites in the stratosphere of formation and destruction of ozone by the varying UV flux – in total and wavelength – and the resulting heating in different places in the stratosphere as well as troposphere. The solar wind impacts ozone levels via proton flow affecting NOx , OH radicals that destroy ozone, especially in places where the Earth’s geomagnetic field is strong at the Poles but also along leakage lines, for example to the Pacific Warm Pool. The result is effects on surface pressure via changes in intensity of the Polar and Equatorial Jets – one obvious result is the strength and timing of EL Nino events and whether they are normal or El Nino Mokadi .

    A good review of these impacts – as far as they are currently known – is given in a review paper by Joanna Haigh of Imperial College London:

  6. Verity Jones

    Very interesting Ed, but I’m not getting exactly what you plotted in Figure 3:
    “Several temperature data bases were investigated and compared, using one-month delay steps and computing the R-squared value in Excel for each comparison. “ What exactly is plotted on the Y axis?

    1. Ed Caryl

      Verity, solar wind against each of those temperature data sets listed in the legend.

      1. Verity Jones

        Ed, I take it that you have taken two sets of data such as in Fig. 2, where the offset is 3 months, and you plot one against the other and get a correlation. Then you plot that as one point on the Fig 3. chart – an R-squared value. Is that right?
        R-squared values of less than 0.1?

        1. Ed Caryl

          Figure 2 has no time offset. The apparent delay is 8 months, as can be seen in the R-squared values.

  7. Ed Caryl

    For monthly data the correlation is very low. It gets much better if you average and us yearly data but the time resolution is lost. Compare the HADSST3 and the 3-month average of the same data. That is to be expected with this noisy data. But the r-squared values are very consistent and not noisy at all, so I believe that what I’m seeing is real.

  8. Paul Vaughan

    Reinterpreting ERSST EOFs 1-4
    (Summary: Earth Climate = Sun + Moon)

    It’s a ruthlessly concise technical document that includes an alert about Mann’s clever trick to redefine the Atlantic to match Trenberth’s deep ocean “missing heat” narrative.

    Best Regards

By continuing to use the site, you agree to the use of cookies. more information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this. More information at our Data Privacy Policy