by Javier Vinós & Andy May
“Probably no subfield of meteorology has had as much effort devoted to it as the effects of solar variability on weather and climate. And none has had as little to show for the research labor.” Helmut E. Landsberg (1982)
For those that prefer it, Christian Freuer has translated this post into German here.
The sun has been correctly identified as the source of climate since the dawn of human intelligence, and consequently the sun was worshipped in many ancient cultures. Large sunspots are visible with the naked eye when the sun is low on the horizon and partially obscured by dust or smoke. Several myths and iconography suggest sunspots were known to ancient cultures from America, Africa, and Asia, however, the first written mention of a sunspot comes from Theophrastus’ De Signis Tespestatum c. 325 BC. This first written record of solar variability was already linked to a climatic effect, since Theophrastus mentions it is related to rainfall. Theophrastus is considered the father of botany and was a student of Aristotle. He succeeded Aristotle as the head of the Lyceum when Aristotle, teacher of Alexander the Great, had to flee Athens due to anti-Macedonian sentiment. Theophrastus’ mention in passing of sunspots must have referred to common knowledge from the past, since he lived during the Greek grand solar minimum of 390–310 BC (Usoskin 2017) and it is very unlikely that anybody at that time could have seen a sunspot with their naked eyes. Most naked-eye sunspot observations known to us come from China, where records have been found starting from 165 BC. The oldest known drawing of actual sunspots is from the Chronicon ex chronicis by John of Worcester, dated in the manuscript to December 1128, during the Medieval grand maximum in solar activity.
Aurorae are an atmospheric light phenomenon that results from the interaction of solar wind and the geomagnetic field, usually produced between 10-20° from the geomagnetic poles. Aurora is the Roman goddess of the dawn, sibling of Sol and Luna. Every morning she would open the gates of heaven for the sun to rise and then race across the early morning sky in her chariot to announce a new day. The name aurora borealis was given to the atmospheric phenomenon by Galileo in 1619, indicating its northern (boreal) direction. Aurorae are occasionally seen in mid-latitudes and rarely in low latitudes when a geomagnetic storm temporarily enlarges the auroral oval. Aurorae have been observed since antiquity. The first records of aurora appear to be three Assyrian clay tablets c. 660 BC (Hayakawa et al. 2019). The prophet Ezekiel also recorded an aurora c. 593 BC in the Bible, and Aristotle wrote about aurorae in his treatise Meteorologica in 340 BC. However, aurorae were not associated with solar activity until the arrival of modern science. Anders Celsius was the first to propose that aurorae were linked to the Earth’s magnetic field in 1733, but the solar link had to wait until the Carrington event of 1859, when the solar flare detected by Richard Carrington and Richard Hodgson that caused a great geomagnetic perturbation was followed by the most intense and lowest latitude reaching aurora in recorded history. Aurorae historical records since antiquity are used, together with naked-eye historic sunspot records, to study past solar activity.
The invention of the telescope in 1608 was soon followed by multiple sunspot telescopic observations. The first records correspond to Thomas Harriot in 1610, and the first publication to Johannes Fabricius in 1611 (Vázquez & Vaquero 2009). Galileo Galilei and Christoph Scheiner carried out systematic sunspot observations in 1612, both realizing they were never far from the solar equator and rotated with the sun. Telescopic sunspot observations arrived just in time to register the Maunder grand solar minimum (GSM) from about 1645 to 1715. No other GSM has taken place since, as the Dalton Minimum was not a GSM (Usoskin 2017).
At least one variable star was known to the ancient Egyptians three millennia ago. There is now evidence that the Cairo Calendar, dated 1244–1163 BC, records, as lucky and unlucky days, the period of the eclipsing binary star Algol, associated with the Egyptian god Horus (Jetsu & Porceddu 2015). Algol was the second variable star described by modern astronomers in 1669. It was preceded by the discovery that Mira was pulsating with an 11-month period by Johannes Holwarda in 1638. The number of known variable stars grew slowly until c. 1850 when it accelerated, and particularly since the introduction of astrophotography in the 1880s. The 2017 General Catalog of Variable Stars (version GCVS 5.1) contains data for 52,011 variable stars. The sun is currently considered a variable star with a very slight variation of one milli-magnitude. Over 80% of sun-like stars display a variability like the sun (Connolly et al. 2021).
Fig. 1.1 shows solar activity since 1600 after Usoskin et al. 2021 (Fig. 8), and Wu et al. 2018 for the 20th century, in open solar flux Weber units. Some landmarks in the study of the sun-climate effect are indicated. In 1608 the telescope was invented. In 1801, Herschel developed the sunspots to climate hypothesis. In 1843, Heinrich Schwabe discovered the solar cycle. In 1968, Roger Bray discovered the 2500-year solar activity cycle associated with a 2500-year climate cycle. In 1974, Colin Hines proposes a sun-climate mechanism mediated by planetary waves. In 1976, John Eddy wrote a landmark article on the Maunder Minimum. In 1986, Karin Labitzke discovered the first solid sun-climate effect in the polar atmosphere during winters. In 1996, Joanna Haigh proposed the “top-down” sun-climate mechanism. In 2022 we propose the “Winter Gatekeeper” sun-climate mechanism hypothesis.
1.2 William Herschel, Heinrich Schwabe, and the early sun-climate frenzy
With the advent of the telescope and the interest in sunspots there came speculation that changes in sunspot number and changes in weather were related, as Theophrastus suggested in 325 BC. Italian Jesuit astronomer Giambattista Riccioli and Mexican astronomer José Antonio Alzate made the same suggestion in 1651 and 1784 respectively.
Musician, composer, mathematician, astronomer, and the best telescope builder of his time, the discoverer of Uranus and infrared radiation, William Herschel, was the first to propose that the sun was a variable star and sunspots reflected changes in solar activity that influenced climate. In an article presented in 1801 at the Royal Society he said:
“I am now much inclined to believe that … [abundant sunspots], may lead us to expect a copious emission of heat, and therefore mild seasons. And that on the contrary, … the absence of … [sunspots], will denote a spare emission of heat, may induce us to expect severe seasons”(Herschel 1801)
It is interesting that Herschel was also the first to correctly relate more sunspots to higher solar emissions, unlike all previous observers and nearly all that followed him until the 20th century. He then proceeded to relate the price of wheat since 1650, obtained from Adam Smith’s The Wealth of Nations to early sunspot counts, finding a correspondence. Herschel’s proposed correspondence is likely incorrect. He himself warned that the criterion was probably not real since the price of commodities is also regulated by their demand, but the lack of temperature records left him with no other method. Figure 1.2 shows that grain production is a better choice, as expected. The great mortality from the 1317 famine and the Black Death from 1346 that killed one third of the European population resulted in decreased demand that kept grain prices low despite the fall in production during the Spörer Minimum from c. 1400-1500.
Herschel sun-climate proposal was met with derision. Lord Brougham scoffed and called it “a grand absurdity” and went on to say that “since the publication of Gulliver’s voyage to Laputa, nothing so ridiculous has ever been offered to the world” (Edinburgh Review 1803).
In Fig. 1.2, plotted are: a) Solar activity reconstruction that shows the Wolf, Spörer, Maunder, and Dalton solar minima. After Wu et al. (2018). The quadratic regression (thin line) follows the long-term change in solar activity. Plot b) shows the wheat price in Dutch guilders per 100 kg (inverted), for France (continuous line), England (dashed line) and Germany (dotted line). After Lamb (1995). Plot c) shows three main crops of grain net yield per acre in England, with annual data (thin line), and a long-term trend (thick line). After Campbell & Ó Gráda (2011). Plot d) shows Northern Hemisphere population growth in percent. After Zhang et al. (2011). Boxes at the bottom identify the periods considered to be the 14th and 17th century crisis by historians. Vertical bars (ACE, abrupt climate event) are periods of climate deterioration. Fig. 1.2 is after Vinós (2022).
Herschel’s detailed solar observations might have revealed the 11-yr solar cycle, except they took place during the Dalton Minimum. That discovery had to await Heinrich Schwabe who was looking for a hypothetical planet inside the orbit of Mercury, called Vulcan, proposed by many astronomers at the time. For 17 years (only one and half periods!) he made detailed solar observations trying to distinguish a transit of Vulcan among the sunspots. He published his solar observations every year, and in 1843 he reported:
“From my earlier observations, which I have reported every year in this journal, it appears that there is a certain periodicity in the appearance of sunspots and this theory seems more and more probable from the results of this year. … If one compares the number of groups with the number of days when no spots are visible, one will find that sunspots have a period of about 10 years, and that for five years of this period they appear so frequently that during that time there are very few or no days when no spots at all are visible”(Schwabe 1843)
Schwabe’s idea attracted little attention until the inclusion of his sunspot data in Alexander von Humboldt’s 1851 monumental work Kosmos. Then four astronomers, including Rudolf Wolf, director of the Bern observatory, noticed that periodic changes in the small daily fluctuations in the geomagnetic field corresponded in period and epoch, with the sunspot cycle described by Schwabe. Rudolf Wolf then began a systematic study of solar variations giving rise to the sunspot record. Heinrich Schwabe, despite not finding Vulcan (it is in the 40 Eridani triple star system, according to Star Trek), was awarded the Gold Medal of the Royal Astronomical Society in 1857.
The discovery of the solar cycle triggered a frenzy to find 11-year periodicities in any weather record. In the 1860s only three articles were published about the sun-climate connection. The next decade they were more than a hundred (Hoyt & Schatten 1997), and in the following decades they continued to multiply. In 1958 the American Meteorological Society listed 1278 articles on solar-weather relationships in its bibliography. Most of the sun-climate studies between 1870 and 1920 agreed that there was a negative correlation between sunspots and temperatures in most locations where a good correlation could be found.
From this period the studies by Wladimir Köppen stand out. Köppen established a climate classification system still in use and made substantial contributions to several branches of science. He was one of the foremost climate scientists of his time and, with his son-in-law Alfred Wegener, lent crucial support to the Milankovitch theory. Köppen’s sun-climate studies were rigorous. His 1873 article about the eleven-year period of temperature was the most comprehensive of that time. He updated the article in 1914. He concluded that the sun does indeed cause a periodicity in the Earth’s surface temperature, not only in many locations but also on a hemispheric and global scale. At that time, the small changes in CO2 levels could not have been the cause of the climate changes detected by Köppen.
At the turn of the century the consensus was that there was no doubt that the 11 and 22-year solar cycles negatively correlated with surface temperature at most locations and even hemispherically and globally, showing a positive correlation with precipitation at low and high latitudes and a negative one at mid-latitudes. In 1903 Nordmann stated:
“The mean terrestrial temperature exhibits a period sensibly equal to that of the solar spots; the effect of spots is to diminish the mean terrestrial temperature, that is to say, the curve which represents the variations of this is parallel to the inverse curve of the frequency of solar spots.”(Hoyt & Schatten 1997).
1.3 The solar constant and the subfield discredit
The amount of solar energy arriving to the Earth, or total solar irradiance (TSI), started to be measured with the invention of the pyrheliometer by Claude Poillet in 1837. The accuracy of the data during the 19th century was poor due to the unreliability of the early instruments and the lack of proper standardization in the early measurements. Despite these shortcomings, astrophysicists of the time noticed that variations in TSI are very small, giving birth to the concept of the solar constant. In 1878 Samuel Langley invented the bolometer and in 1890 he became director of the Smithsonian Astrophysical Observatory. With the help of Charles Abbot, who succeeded him in 1906, they set up a program to determine solar constant variations with stations located at mountain peaks in the US and Chile. Data from 1923 to 1954 showed small variations associated with the solar cycle of 0.02–0.25% and a controversial increase of 0.2% during the 31-year period. Charles Abbot was convinced of the sun-climate connection, and after the end of the program he wrote in the first issue of the now named Solar Energy Journal that: “As solar radiation and the weather appear to be affected by identical periods of variation, it is therefore likely that weather changes are produced by solar variation.” He then proceeded to compare solar forecasts for precipitation and temperature at St. Louis and Peoria with actual weather data from 1854 to 1939, as evidence of his thesis (Abbot 1957). Nevertheless, he acknowledged the difficulty of attributing significant weather changes to such small variations in the solar constant, recognizing that the general opinion was against his hypothesis.
The consensus of the time had changed from supporting the sun-climate connection at the beginning of the 20th century to rejecting it by mid-century. The change was because of a better determination of the solar constant, which at the time could still accommodate significant variability, but by something that happened around 1920. All those statistically significant correlations that the best scientists of the time, like Wladimir Köppen, had found in about seven decades of weather data (c. 1840–1910), started to fail around 1920, or even worse they inverted, something that nobody could explain. The situation became very confusing, some authors claiming positive correlation, others negative correlation, and others no correlation at all. Elaborate hypotheses were proposed, indicating a troubled paradigm, and the whole subfield fell into disrepute. By 1950 the study of sun-weather relationships was considered by many to be an undignified pursuit for a meteorologist (Hoyt & Schatten 1997), and this state was recognized by Abbot in his 1957 article.
Fig. 1.3 shows the timing of the sign reversal in several sun-climate correlations clustered between 1920–30 (Fig. 1.3). The temperature correlation inverted, as well as many other meteorological variables, like precipitation, winds, the preferential location of the Icelandic low, and the strength of the Indian monsoon. Even though the sign reversals could not be explained, they still suggested a sun-climate relationship, just not one based on TSI changes, since the relationship between solar emissions and sunspot activity does not invert. This important conclusion escaped most climate researchers at the time, and it escapes many of them today. In the early 20th century, climate shifts were unknown, but it is known today that c. 1924 a regime change took place in the Pacific from a cool Pacific Decadal Oscillation to a warm one (Mantua & Hare 2002). This shift took place right after the 1923 solar minimum and resulted in global warming (the early 20th century warming), despite solar activity being below average until 1934.
According to Hoyt and Schatten (1997) the sun-temperature correlations have changed sign several times during the past 400 years and have been negative between about 1600-1720 and 1800-1920, and positive between about 1720-1800 and 1920 and the present. The 1800 reversal is illustrated in figure 1.4.
1.4 Roger Bray, John Eddy and the 1970s revival
While the sun-climate subfield was falling into disrepute, the seeds for its renaissance were being planted. Andrew Douglass was an astronomer that had been fired by Percival Lowell in 1901 for his skepticism about the artificial nature of Martian canals. Through his entire career Douglass was convinced of the sun-climate effect, and in 1904 he noticed a correlation between tree-ring widths in Arizona, related to precipitation conditions, and sunspots. By pursuing this relationship, he developed the new subfield of dendrochronology over the following 40 years, the only precise method for dating ancient structures until the advent of radiocarbon dating. Douglass studied the annual rings of trees in relation to climate and solar activity and was the discoverer of the centennial solar cycle (which he named the triple-triple solar cycle). He did not find it in sunspot records, but in its climatic effect on sequoia ring-growth (Douglass 2019). It is the only instance of a solar cycle first identified in the paleoclimate record.
Willard Libby developed radiocarbon dating in the late 1940s. For the method to be accurate it was essential to know how the atmospheric 14C/12C ratio had changed over time. Scientists had to build a calibration curve (IntCal) from precisely dated tree rings by Douglass’ dendrochronological method, to transform radiocarbon ratios into calendar ages. Hans Suess in California and Minze Stuiver in Arizona were among those leading the effort. In 1961, Stuiver was the first to suggest that atmospheric 14C variations lasting a few centuries or less were due to solar modulation of cosmic ray 14C production in the upper atmosphere (Stuiver & Quay 1980). Suddenly the sun appeared to be more variable over long periods of time than recent solar constant measurements indicated.
This finding opened the door to using the recently reconstructed 14C changes to study solar variability and its relationship to climatic changes in the distant past. Starting in 1963, and based on his glaciological and botanical studies, Roger Bray proposed that there was a close relation between solar activity and climate during the past centuries and millennia. In 1968 he identified the solar and climate 2500-year cycle that has been recently named after him (Vinós 2016). This long solar cycle is the most important, in terms of climatic effect, during the Holocene. The most recent low in Bray’s solar cycle, during 1388–1834 (Bray 1968), coincides with the Little Ice Age (LIA) discovered by François Matthes in 1939. Roger Bray was the first to propose that the LIA had a solar cause. During the 1960s and 1970s Roger Bray published 14 articles in Nature and Science linking solar variability and volcanic activity to climate change, but since he was a botanist researching climate independently from New Zealand and removed from other sun-climate scientists, he was unjustly not credited for his findings. His cycle was given the absurd “Hallstatt” name by Paul Damon and Charles Sonnet (Damon & Sonnet 1991), despite their being aware of Bray’s work.
In 1974 Robert Currie published a study of 226 weather stations throughout the world, that in 1993 was updated using 1,200 U.S. stations. Using new statistical methods just developed, he found both a 10.5 solar and an 18.9 lunar signal in many of them but decided that local effects could mask the regional signal at some stations. Interestingly, he detected that stations east of the Rocky Mountains displayed a positive correlation between solar activity and temperature, whereas stations west of the Rocky Mountains displayed a negative one (Currie 1993), an effect not unlike the signal-reversal in correlation observed in the 1920s. In 1980 Currie detected an 11-yr sunspot cycle signal in Earth rotation. He was not the first to do so after the invention of the atomic clock, but since solar effects on Earth are so controversial (probably due to the lack of an accepted mechanism) the Sun-Earth rotation effect has been “discovered” independently multiple times, the last time in 2010, and it continues to be ignored.
However, the first viable mechanism for the sun-climate effect was suggested by Colin Hines in 1974. A year earlier Wilcox et al. (1973) discovered that the solar magnetic field sector structure affected the average area of low-pressure troughs during the winter in the Northern Hemisphere at an altitude of 300 mb (roughly 30,000 feet or 9,100 meters). Hines (1974) was skeptical of any sun-climate effect but suggested that planetary waves subjected to variable reflection in the upper atmosphere, may induce variable interference patterns in the lower atmosphere. These could constitute a possible candidate for the effect if it were real.
A widely circulated article by Joe King (1975) did much to popularize the renewed interest in sun-climate relationships by presenting a great variety of evidence and concluded that:
“the accumulated evidence is so compelling that it is no longer possible to deny the existence of strong connections between the weather and radiation changes.”King, 1975
It prepared the scene for John Eddy’s landmark article in Science a year later. Eddy brought to light the forgotten finding by Gustav Spörer and Edward Maunder that during the 1645-1715 period the sun behaved in a very unusual way and displayed very few sunspots. Eddy, very interested in the history of astronomy, supported their finding with naked-eye sunspot observations, auroral observations, eclipse observations, and 14C data (Eddy 1976). The Science article on the Maunder Minimum became hugely popular. Eddy followed up with several articles on the sun-climate relationship over the past 7500 years (Fig. 1.5).
George Siscoe optimistically reviewed the 1970s golden decade of sun-climate research (Siscoe 1978), citing three major advances. Those of Wilcox and Eddy, and the studies linking drought in the North American Southwest to the 22-yr Hale solar magnetic cycle. But while the sun-climate subfield was again bursting with activity, advanced with well attended meetings, given its interdisciplinary and controversial nature, it was still criticized. Barrie Pittock published a critical look at 140 sun-climate articles (Pittock 1978) and concluded “that despite a massive literature on the subject, there is at present little or no convincing evidence of statistically significant or practically useful correlations between sunspot cycles and the weather or climate.”
1.5 1980s Global Warming and the second sun-climate demise
In the 1980s, the sun-climate renaissance of the 1960s-70s was drained of energy by improvements in solar constant measurements. Cavity radiometers were first equipped in the Earth Radiation Budget experiment onboard Nimbus 7 satellite in November 1978. The Active Cavity Radiometer Irradiance Monitor (ACRIM) experiment started with the Solar Maximum Mission in February 1980. For the first time solar constant values reached the precision of two decimal places of a percent. The decrease in the solar constant from the 1980 solar cycle maximum to the 1986 minimum was determined to be 0.15%, or 2 W/m2. But the yearly change had already been determined by 1982 to be of only 0.02%. The great majority of researchers believed that only changes in total energy could affect climate, to them a change of ± 0.07% could not produce significant effects. Those defending the idea that small solar changes could act on atmospheric instabilities that amplified their climatic effect were at odds to explain how they could do so in an inherently unstable atmosphere.
In 1980 Nastrom and Belmont appeared to have identified how the sun-climate effect worked. Using radiosonde data for 174 stations in the Northern Hemisphere for the period 1949-1973 they found that tropospheric winds displayed a clear solar signal. They found that wind speed and temperature were responding to the solar cycle, and the effect was maximal near the tropopause during winter (Nastrom & Belmont 1980). In 1983 the same authors declared their result statistically insignificant after more tests (Venne et al. 1983). Also in 1980 Minze Stuiver, the authority in radiocarbon dating that had initiated past solar activity reconstructions in 1961, published an influential article in Nature comparing the new detailed 14C variability reconstruction and several long climate records (Stuiver 1980). He concluded that low solar activity periods like the Maunder Minimum had taken place several times during the past 6,000 years and that a relationship between climatic series and the 14C derived record of solar activity for the past millennium could not be established.
Minze’s article demolished John Eddy’s work. Two years later Eddy certified the death of sun-climate studies as mainstream science:
“Spacecraft measurements have established that the total radiative output of the Sun varies at the 0.1–0.3% level. … Such changes can be expected to perturb the terrestrial surface temperature by a fraction of a degree centigrade and probable evidence of this solar-induced signal has been found. The effect, though important in terms of understanding the climate system, is too small to be significant in practical weather or climate predictions”(Eddy et al. 1982)
As the old proverb says: “fool me once, shame on you, fool me twice, shame on me.” Sun-climate researchers had been burned twice, in the 1920s and 1980s; it should not happen again. The subfield fell into absolute disrepute. Nothing with the words “solar” and “climate” in the same phrase was to be taken seriously again. The timing was perfect for the CO2 hypothesis of climate change, as global warming started for the second time in the 20th century and this time it could be blamed solely on CO2 changes. It couldn’t be the sun and whoever suggested it faced ridicule and an insignificant career. It has reached a point where even clear solar effects on Earth’s rotation or on El Niño/Southern Oscillation are meticulously ignored.
The length of each solar cycle must be taken into consideration when comparing solar activity. Top of Fig. 1.6, the sum of the sunspots for every year in the cycle is divided by the number of years in the cycle, and the 1700-2020 sunspot average is subtracted from the result. The result is displayed as a bar graph with bar width proportional to the duration of the cycle. Line is the linear regression trendline. The bottom of Fig. 1.6 is the yearly sunspot international number from WDC-SILSO. The 1934–2008 period is the 75-year period with highest solar activity in at least 700 years, as we know solar activity was very low during the LIA after c. 1270. This period is named the modern solar maximum. The close correspondence in time between the highest solar activity 75-year period and the highest global warming 75-year period (1925-2000) in 700 years is unlikely to be a coincidence and deserves a thorough investigation that is not taking place (From Vinós 2022).
The turnaround was complete for those who wished to continue their careers. Wilcox, Svalgaard, and Scherrer published in 1976 “On the reality of a sun-weather effect” (Wilcox et al. 1976). They were close to solving the problem. They were in the right part of the planet (the Northern Hemisphere extratropics), at the right location (the upper troposphere-lower stratosphere), at the right time of the year (during winter), looking at the right variable (pressure), and seeing a clear effect. That the sun-climate effect is stronger in Arctic latitudes during the winter is further confirmation of what could be deduced from the reversal of sun-weather correlations: The sun cannot affect Arctic winter climate through changes in TSI because there is no solar irradiation during the polar night, and in the Arctic, it doesn’t matter how small or large the TSI changes are for a sun that doesn’t shine. The existence of a different mechanism is required.
Colin Hines (1974) had already identified the sun-climate mechanism based on the Wilcox et al. results; it was the differential propagation and reflection of planetary waves due to changes in zonal wind speed. The same changes identified by Nastrom and Belmont and later discarded. But Wilcox et al. walked away. They forfeited their chance to find proof of a sun-climate effect in this 200-year-old climatological quest. That honor would go to a woman more interested in science than her career or reputation. Wilcox, et al. co-author Leif Svalgaard has dedicated his latest years to vigorously refuting any suggestion that solar variability may have contributed to modern global warming, and to tirelessly promoting a controversial change to the sunspot record that better supports his views.
1.6 Karin Labitzke and the unacclaimed first solid proof of a solar signal
In 1982 the National Research Council published a monograph on “Solar Variability, Weather, and Climate.” It reads like a death certificate of the subfield. It includes articles by James Holton and Barrie Pittock among others and was under the chairmanship of John Eddy. James Holton, one of the foremost experts in the atmosphere, analyzed the possible physical mechanisms for a sun-climate effect through a dynamic coupling between the stratosphere and the troposphere in a negative light (Holton 1982). When considering Hines’ (1974) mechanism, Holton conceded that changes in stratospheric flow related to solar variability might alter the reflection/absorption of planetary waves and through wave interference produce effects in the troposphere. To him, this mechanism provided a possible link between solar variability and tropospheric weather and climate, which could be significant despite the huge energy difference between the solar input and the climate response. However, he concluded that the mechanism was speculative.
Two years earlier Holton and Tan (1980) had published a seminal article showing that equatorial stratospheric winds, despite circling the Earth at high altitude above the equator, modulated global circulation. These winds are known as the Quasi-Biennial Oscillation (QBO) because they alternate between easterly and westerly direction with a quasi-periodicity of slightly over two years. The effect of the QBO on Northern Hemispheric circulation discovered by Holton and Tan was to alter the mean geopotential (pressure) at the pole during the winter via planetary waves. This finding should have raised all kind of questions about possible solar activity involvement, since it was clearly related to the Wilcox et al. findings and involved Colin Hines’ planetary wave mechanism, but at the time a solar explanation was (and still is) unacceptable to most academics. During winter, strong westerly winds circle the polar region trapping a cold-air low-pressure center, forming a strong polar vortex. The north polar vortex modulation by the QBO is so important that it received the “Holton-Tan effect” name. Interestingly, the north polar geopotential modulation by the QBO was only significant during the winter season when the mean zonal wind is westerly and vertically propagating planetary waves are present. Holton and Tan had to introduce the planetary waves condition because at certain times the correlation broke down.
Karin Labitzke (1987) noticed that the polar vortex–QBO correlation broke down sometimes during the westerly phase of the QBO, but only when solar activity was near its cyclical maximum. She decided to segregate the data on stratospheric polar temperatures according to QBO phase. The very low correlation between solar activity and polar temperatures, when all data are considered, becomes very high using the segregated data (Fig. 1.7). After 186 years Labitzke had solved the quest initiated by William Herschel in 1801. In a follow-up article with Harry van Loon (Labitzke & van Loon 1988) they extended the study on the solar effect on winter atmospheric pressure and temperature to the Northern Hemisphere troposphere. The main conclusion from this work is that the signal of the QBO in the extratropical stratosphere is strengthened in solar minima and weakened in solar maxima. That the QBO orientation flips the solar effect from one sign to its opposite is not unlike other correlation sign-reversals in the sun-climate effect, and a third indication that the effect cannot be caused just by changes in TSI.
In Peixoto and Oort’s (1992) indispensable Physics of Climate manual Labitzke and van Loon’s findings were properly appraised. After declaring it the most convincing statistical evidence of a solar-weather relationship found, they continue:
“Even at the earth’s surface, the correlations between solar activity and sea level pressure or surface temperature … are unusually high and appear to explain an important fraction of the total interannual variability in the winter circulation”(Peixoto & Oort 1992)
However, Labitzke’s ending of the 186-year quest for a sun-climate effect, initiated by William Herschel in 1801, could not have come at a more inconvenient time. Global warming was already blamed on CO2 and the scientific dogma was completely against her finding, since all sun-climate studies had been discredited. James Holton said:
“Superficially, I can’t find anything wrong with it, but there is absolutely no physical basis, and that bothers me. These people have the highest correlation I’ve seen, but if I were a betting man, I would bet against it.”(Kerr, 1987)
She had found a clear and indisputable effect of solar activity on climate. It could not be disputed, but it could be ignored. And it was going to be ignored as an oddity with little practical effect, and no place in modern climate understanding.
In Fig. 1.7, the following are plotted: A) Lack of correlation between winter North Pole stratosphere temperature and solar activity (10.7 cm solar flux) when data for all years are considered. B) shows a clear positive correlation when only QBO west phase years are considered. Not shown is the clear negative correlation when only QBO east phase years are considered (Fig. 1.7 is from Kerr 1987).
The scientific consensus about an important sun-climate effect went from being against, prior to 1850, to being in favor between 1860s–1920s, negative from the 1920s–1960s, positive again in the 1960s–1970s, and then negative since the 1980s. It only demonstrates that scientific consensus has no place in science. Scientist’s opinions are not science. Only evidence constitutes science. Even though evidence can be dismissed or ignored, it remains, waiting for the time when it will be properly appraised.
The fall in disgrace of the sun-climate subfield during the early 1980s discouraged further research into how solar variability affects climate. The strengthening of the politically supported CO2 hypothesis of climate change turned the subfield into a scientific dead zone. Only a handful of researchers decided to pay the steep reputational and career price of pursuing this research interest. The difficult sun-climate relationship research subject experienced scant progress between 1870 and 1980, considering the amount of research labor invested. Turning it into a disreputable field after 1980 has resulted in slower progress despite the rapid advances in climatology in the decades since.
Sun-climate researchers must accept not getting recognition for their findings, having more difficulties in publishing their results in good journals, not getting good students, and being lowly considered by their peers. Due to that, they have difficulties advancing their careers, and the rate of career failure among young scientists entering the subfield is high. The exception are sun-climate researchers that work under the premise that solar variability may have not significantly contributed to modern warming. Those are well-considered, highly cited, contribute to IPCC reports, and sometimes produce low solar variability inputs for climate models.
Despite these difficulties, as models attempt to reproduce real phenomena, and reanalysis is fed real climate data, the sun-climate effect keeps appearing, defying whack-a-mole attempts to keep it buried. In 1996 Joanna Haigh showed in a landmark article in Science that changes in atmospheric circulation, reproduced only weakly in models, had a clear solar origin (Haigh 1996). Tropical ozone changes appeared to be critical for the effect in the model. It soon became clear that the changes in TSI during the solar cycle did not involve enough energy to explain the observed climate effects. So, it was assumed that some sort of amplifying mechanism was responsible. Even though climate models did not include the stratosphere until recently, Haigh developed what is known as the “Top-down mechanism” for amplifying the solar effect on climate (Fig. 1.8).
In Fig. 1.8 solar UV radiation acts on the ozone layer in the stratosphere increasing its temperature (T) and the amount of ozone (O3). The change in temperature alters the latitudinal temperature gradient, and through thermal wind balance affects the zonal-mean zonal winds (ΔU). The change in zonal winds alters the properties of the atmosphere for propagating planetary waves. The effect creates an anomalous divergence (>0) of the Eliassen-Palm flux (F) proportional to the eddy potential vorticity, changing the deposition of momentum and kinetic energy. The strength of the polar vortex (not shown) depends on those changes, driving changes in the Arctic Oscillation (AO), North Atlantic Oscillation (NAO), and the Hadley and Walker circulations. Thick interrupted arrows indicate coupling. The figure is after Gray et al. 2010.
The orthodox view of the sun-climate effect at present can be summed up in Judith Lean’s 2017 review. The 0.1% increase in total irradiance between solar minimum and maximum is associated with an increase of 0.1 °C in Earth’s global surface temperature. There are dynamic processes that alter the regional response both at the surface and in the atmosphere. The effect of a grand solar minimum, like the Maunder Minimum, is likely less than a few tenths of a °C of global cooling (Lean 2017). Over the following five parts in this series of articles we will explain the recently proposed Winter Gatekeeper hypothesis of sun-climate effect (Vinós 2022). It involves some very complex climate phenomena, which explains why it escaped discovery for 220 years. In the next part we will see that the orthodox IPCC sanctioned climate change view ignores the effects of solar variability on at least five very important climate-related phenomena that essentially refute it. It is hoped that the time has arrived for another reversal in the sun-climate consensus.
Note: This is the first of a six-part series on the effect of solar variability on climate change. Javier’s previous 13-part series on climate change was posted between 2016 and 2018 and can be read at judithcurry.com by introducing “Nature Unbound” in the search box. It generated over 4,000 comments and was the basis of his September 2022 book, Climate of the Past, Present and Future. A Scientific Debate, 2nd ed., where part of the material in this series is included.
The bibliography can be downloaded here.
A list of abbreviations used can be downloaded here.
This post was first published, in slightly modified form, on Climate Etc.
45 thoughts on “The Sun-Climate Effect: The Winter Gatekeeper Hypothesis (I). The search for a solar signal”
Good job providing the background information. Thank you
Thanks for a great review of solar studies.
Re: this hypothesis: “The effect of the QBO on Northern Hemispheric circulation discovered by Holton and Tan was to alter the mean geopotential (pressure) at the pole during the winter via planetary waves. This finding should have raised all kind of questions about possible solar activity involvement, since it was clearly related to the Wilcox et al. findings and involved Colin Hines’ planetary wave mechanism, but at the time a solar explanation was (and still is) unacceptable to most academics. During winter, strong westerly winds circle the polar region trapping a cold-air low-pressure center, forming a strong polar vortex. The north polar vortex modulation by the QBO is so important that it received the “Holton-Tan effect” name. Interestingly, the north polar geopotential modulation by the QBO was only significant during the winter season when the mean zonal wind is westerly and vertically propagating planetary waves are present. Holton and Tan had to introduce the planetary waves condition because at certain times the correlation broke down.”
On the basis of the strength of the energetics at the pole, reflecting the physical forces involved, causing the atmosphere to spin in the same direction as the Earth but faster, its sensible to guess that the causation is likely the other way round. The strength of the circulation is greatest in the stratosphere, that is also the bulk of the atmospheric column in winter when the influence of ozone in reversing the lapse rate is seen above the 400 hPa pressure level and traces of ozone are present at 200 hPa. So, 80% of the atmospheric column in higher latitudes is affected by electromagnetic propulsion of molecules that carry a charge, ozone being a case in point.
The best illustration of the strength of these dynamics, and how it can change over time is seen over Antarctica where surface pressure has been in decline for seventy years.
I believe that the fluid dynamics notion that the movement of the atmosphere is driven by the temperature differential between the equator and the poles is incorrect. It gives rise to the notion that so called ‘planetary waves’ are influential. This promotes the idea that effect is cause.
It’s more fruitful to consider the flux in the ‘annular modes’ and pose the question: What drives the reversal in the ratio between surface pressure on the margins of Antarctica and the mid latitude high pressure zone. Answer: The flux in polar cyclone activity. What causes polar cyclones? They emanate from the mid to lower stratosphere and may or may not propagate to the surface..
I’m confused by your comment, and you may need to explain your ideas better. This entire series is about the meridional transport of energy from the equator to the poles and the variability in the process. Polar (or near polar) cyclones are part of that process and play a key role, especially in the Northern Hemisphere. We will also discuss our ideas about how the Northern and Southern Hemisphere surface pressures change between the pole and the mid-latitudes. Atmospheric (including planetary) waves are very important and influential in our opinion, we disagree on that point, as well as the importance of the LTG (latitudinal temperature gradient). The remaining posts will outline our ideas, this discussion may be a bit early. I’ve forwarded your comment to Javier, so we may have more later.
I’ll have another go at explaining what I think I have worked out by looking at the history of the atmosphere as it is revealed in reanalysis data.
It seems to me that we are conscious of the same phenomena that change the rate of transport of energy from the equator to the pole. It’s just a question of the original modes of causation. Observation of the annular modes phenomenon, that relates to a fall in pressure in high latitudes (compensating rise everywhere else including Northern Hemisphere and the Arctic) is that a flooding of the Antarctic polar cap region with ozone rich air (called a sudden stratospheric warming) air that is normally on the equatorial side of the polar vortex of very cold descending air, is associated with, and the root cause of the fall in surface pressure in high latitudes. This comes about as there is a stalling of descent of very cold ozone deficient air from the mesosphere. It’s a top down phenomenon. The change proceeds from the highest altitudes to the lowest. First at 1hPa, and many days later at the tropopause. So rule out any near surface mode of causation.
The circulation of air in the stratosphere over the pole in winter is vigorously west to east at both poles. In summer this circulation stalls and is weakly east to west. This is a seasonal (summer) stratospheric warming. It occurs after the ‘Final warming’. The period of stalling over Antarctica is just a few months whereas it is six months in the Arctic. In the past, essentially prior to 1976, when surface pressure was higher over Antarctica, the period of the weak east to west flow was two months or less. The relative weakening of the vorticity of the winter circulation over Antarctica, in the shoulder months of October -November and March April, that happened post 1976, is related to a reduction in surface pressure over Antarctica and its surrounds as ozone proliferated giving rise to an abrupt warming of the entire stratosphere (at all latitudes) in the late 1970s. This was initiated over Antarctica. The consequent loss of surface pressure (loss of atmospheric mass) affected all latitudes pole-wards of about 45° South. Pressure increased elsewhere, including over the equator and most of all in the mid latitudes of the southern hemisphere as atmospheric mass moved away from high southern latitudes.
The resulting increased pressure differential enhanced the westerly flow in the mid latitudes, speeding the transfer of energy from equator towards Antarctica. This changes the climate of the mid latitudes, but only weakly.
The biggest consequence of the change in air flow is via its influence on cloud cover in the mid latitudes of the southern hemisphere due to 1. Higher surface pressure and 2. The descent of ozone from the stratosphere that warms and dries the air, reducing cloud cover. And remember that the mid latitudes of the southern hemisphere are mainly ocean.
Take into account the heating power of ozone and its instantaneous impact on the temperature of the air mass containing that ozone.
So far as the quotient of ozone in the stratosphere is concerned, its likely determined more by chemical interaction with NOX species from the troposphere and the mesosphere rather than the quotient of short wave radiation that disassociates the oxygen molecule. That occurs chiefly in the ionosphere rather than the stratosphere. Ozone builds up in the stratosphere because its a relatively safe place. Low sun angles reduce the density of ionizing UVB that destroys ozone and NOX is less present than in either the mesosphere or the troposphere.
The stratosphere has been gradually cooling since the abrupt warming of 1976-82. Its monitored here: https://www.cpc.ncep.noaa.gov/products/stratosphere/temperature/
I am looking forward with great interest in reading your paper.
It is nice to see your ideas. We definitely agree that a major climate shift occurred in 1976 and we discuss that in Part 4 at length. I think we disagree on the impact of changes in ozone, but from your comment I can’t tell. The asymmetric ozone field is very important in delivering the effects of changing solar output through atmospheric waves, but we see it as a component, not a driver. Let’s discuss this again after you have read Part 2, which will be out next week.
A few more comments.
Winter extratropical circulation responds to solar activity because changes in Earth’s length of day (LOD) say so. The energy to alter zonal circulation at the extra-troposphere is provided by planetary waves. This has been demonstrated thoroughly through the studies on sudden stratospheric warmings. Much of your comment refers to stratospheric zonal circulation, not meridional circulation, which is the main vehicle for poleward energy transport. Much more in the next few posts.
Andy, I see it the other way round. Change in length of day is due to change in atmospheric drag. Change in the drag relates to the balance between east to west and west to east flows.
Erl, I think you misread the sentence, we agree LOD varies due to atmospheric drag.
And furthermore, I would maintain that the meridional flow is wholly dependent on stratospheric dynamics that are strongest in the winter hemisphere. Labitzke found her solar effect in winter.
It’s the extent of the ionization of the atmosphere that determines the impact of the solar wind on the east to west circulation that occurs in winter. Speeding up or slowing down of that circulation depends on the impact of the interplanetary magnetic field on the Earth’s own magnetic field, present in its atmosphere. In winter the air moves in the same direction as the Earth, but faster. The atmosphere assists the Earth’s rotation.
We agree that the impact of meridional flow is higher in winter. The details of how we think small changes in solar UV affect the troposphere are covered in part 2. We can pick this up after you have read that, it is too complicated to put in a comment. You are on the same path as we are, but we might differ in some details, hard to say. Suffice it to say that ozone affects planetary waves.