Sunlight, the ‘fire from heaven,’ is source of almost all life on Earth—or at least, the energy it relies upon. In Part One of “Fire From Heaven,” we saw how scientists sought to measure that fire, and how succeeding generations attained better and better values for what was long termed the ‘solar constant’—roughly, the amount of energy per square meter reaching the top of Earth’s atmosphere.
Later they instead examined ‘TSI’—‘Total Solar Irradiance.’ This term was introduced in recognition of the small but not necessarily insignificant variations that occur in solar output. The ‘solar constant’ isn’t constant—not quite.
We also noted that French physicist Claude Pouillet, who in the 1830s made the best early estimate of the solar constant, also measured energy flows by night. He found that even then, with the sun screened by the bulk of the Earth, energy flows to the Earth from the sky.
This downward flow takes place, not in the shorter wavelengths that characterize visible light, but in the longer ones that we call ‘infrared radiation,’ and that used to be termed 'radiant heat.' This ‘flux’ has come to be termed DLR (for "downwelling longwave radiation"), or sometimes DLW—the “L-W” meaning "Long-Wave."
But though Pouillet was the first to examine DLR in the large context of Earth’s heat budget, he was not the first to make experimental measurements of radiation at night. Nor was he first to draw shrewd analytic conclusion, nor to publish his results. Those honors go to a medical doctor and amateur scientist, William Charles Wells (May 24, 1757-September 17, 1817.)
Wells was born in Charleston, South Carolina, but lived most of his life in England, where he made a modest living as a physician and devoted much of his time to scientific investigation. As one might expect, most of his scientific papers related to medicine--but his most famous study was quite different. As a younger contemporary wrote:
He began an inquiry into the nature of dew, and published ' An Essay on Dew' in 1814. He demonstrated, after a series of well arranged observations made in the garden in Surrey of his friend James Dunsmure, that dew is the result of a preceding cold in the substances on which it appears, and that the cold which produces dew is itself produced by the radiation of heat from those bodies upon which dew is deposited. For this, the first exact explanation of the phenomena of dew, he was awarded the Rumford medal of the Royal Society.
Reading his Essay reveals Wells to have been a careful observer, a resourceful experimenter, and an analyst who reasoned closely and paid attention to detail. (“An Essay Upon Dew” is available free via Google Books.)
It must be remembered that in Wells’ day, dew was variously thought to rise from the Earth, or to literally fall from the skies, like mist. It was even believed to cause putrefaction, to produce cold, or to be a threat to health. These beliefs go back to antiquity—Wells cites Aristotle several times, for instance. In this context it’s remarkable how essentially modern Wells’ thought was--despite his eighteenth-century writing style.
Wells did his work with relatively modest technological means—basically, a selection of thermometers and a few additional ‘props’—mostly either coverings or supports for the thermometers.
(Although it's worth remembering in this context that affordable, reliable measurement of temperature was a relatively recent phenomenon in his day—Fahrenheit’s scale, the first widely-accepted system to feature fixed reference points, dates to 1724. Prior to that, thermometers essentially measured only relative temperature. Wells' collection of thermometers was not so typical of his day.)
Wells measured nighttime temperatures in various places and under various physical positions, and observed the temperature changes relative to the appearance of dew and to meteorological conditions. He found that clear nights were especially favorable to the production of dew, and that on those nights temperatures tended to be lower at the surface of the garden’s grass than either at the ground or in the air.
Particularly interesting for our purposes were observed changes in sky conditions:
. . .I have frequently seen, during nights that were generally clear, a thermometer lying on the grassplat rise several degrees, upon the zenith being occupied only a few minutes by a cloud. On the other hand, upon two nights I observed a very great degree of cold to occur on the ground, in addition to that of the atmosphere, during short intervals of clearness of sky, between very cloudy states of it.
It’s a startling observation for many, even today—it may seem strange that clouds affect night-time ground temperatures, when there is no sunlight to be blocked by cloud shadows. It may even seem as if this involves some kind of spooky ‘action at a distance.’
But Wells was not puzzled. He realized that there were two factors at work. Firstly, the ground radiates energy to the sky:
. . . bodies situated on or near to the surface of the earth become, under certain circumstances, colder than the neighbouring air, by radiating more heat to the heavens than they receive in every way. . .
Secondly, the sky radiates—or at least, the clouds in the sky radiate—energy back to earth:
Dense clouds near the earth must possess the same heat as the lower atmosphere, and will therefore send to the earth, as much, or nearly as much heat as they receive from it by radiation.
Moreover, this radiation from sky to earth is variable—clouds radiate more than clear sky, and different clouds radiate differently:
. . . similarly dense clouds, if very high, though they equally intercept the communication of the earth with the sky, yet being, from their elevated situation, colder than the earth, will radiate to it less heat than they receive from it, and may, consequently, admit of bodies on its surface becoming several degrees colder than the air. In the first part of the Essay an example was given of a body on the ground becoming at night 5º colder than the air, though the whole sky was thickly covered with clouds.
(Wells was quite aware of what we now call the lapse rate—the decline of temperature with increasing altitude; observers such as Horace De Saussure, whom we met in Part One, had written about this. Wells also knew about inversions, in which warmer air overlies colder air; for example, he discusses measurements of inversions made by scientifically-inclined balloonists.)
Wells’ observations, then, disclose two opposing radiative fluxes. At night the earth radiates to the sky; this flux came to be termed ‘nocturnal radiation’--though as Wells himself noted it is not strictly nocturnal:
Radiation of heat by the earth to the heavens must exist at all times; but, if the sun be at some height above the horizon, the degree of which is hitherto undetermined, and probably varies according to the season, and several other circumstances, the heat emitted by it to the earth will overbalance, even in places shaded from its direct beams, that which the earth radiates upwards.
There is also an opposing flux from sky to earth—later termed ‘sky radiation.’ Like its counterpart, sky radiation occurs around the clock. But when (as would remain the case for decades) the instruments available to study it were limited to variations on the alcohol thermometer, measuring daytime fluxes other than the overwhelming direct solar ‘fire from heaven’ was not very practical.
Though the Rumford Medal gave Wells’ work a forceful endorsement, his ideas also attracted opposition from some—even, at times, opposition with a political agenda. A nineteenth-century English commentator, writing of the eminent Italian physicist Macedonio Melloni (1801-1853), explains some of this context:
[In Naples], in the autumn of 1846, he conducted his researches on the nocturnal cooling of bodies. His memoirs on this subject were read before the Royal Academy of Naples, in February and March, 1847. Mr. Tomlinson, in giving an account of these experiments, says: "The most curious point connected with this inquiry is its origin. It will hardly be believed that the Austrian and Bourbon governments, in their dread of novelty, would not allow the true theory of dew to be taught. Melloni, in order to show that the laws of terrestrial radiation are the same in Italy as in countries where there is more political liberty, undertook these researches.”
Of course, “the true theory of dew” refers to the work of Wells.
Melloni’s confirmation of Wells was carried out with instrumentation similar to the original study, but he had previously pioneered another approach:
He had already assisted his friend and countryman, Nobili, in perfecting the idea of a highly sensitive thermometric instrument, based on the thermo-electric pile, and recording its indications by means of a galvanometer. By means of this instrument and his own marvellous experimental skill, it was proved that rays of heat present phenomena as complex as those of light.
Melloni died of typhoid at a comparatively young age, with important work still undone. Nevertheless, his work on heat had already earned him a Rumford medal, and the nickname “the Newton of heat.”
A Gallery of Thermopiles
- Differential Thermopile
Thermopiles by Melloni and others, from a collection at Kenyon University.
This use of electric technology anticipates research themes of Samuel Langley, whose quest for the ‘solar constant’ was outlined in Part One. (Langley, like Wells and Melloni--and for that matter, like John Tyndall, who famously first discovered specific greenhouse gases--would in due course receive a Rumford Medal of his own.)
In particular, Langley’s electric instrument, the bolometer, vastly increased the sensitivity with which radiation could be observed. Both Langley and his disciple Charles Greeley Abbott focused primarily upon the ‘fire by day,’ not the ‘cloud by night.’ But Abbott, like Langley, was nothing if not systematic, and in 1912 and 1913 collaborated with physicist Anders Knut Angstrom to study nocturnal radiation. This collaboration would result in a lengthy 1918 paper by Angstrom, “A Study Of The Radiation Of The Atmosphere.”
Angstrom--not to be confused with his more famous grandfather, Anders J. Angstrom, the physicist after whom the "Angstrom" is named--had developed an electric instrument of his own, conceptually a type of actinometer which he usually termed a ‘pyrgeometer.’ (This term has come to be standard for modern instruments intended to measure DLR.)
Citing Wells, Melloni and Pouillet as pioneers, Angstrom credited the Swiss researcher Maurer as being the first to attempt measurement of nocturnal radiation in 1886. (Presumably this means direct measurement, as opposed to less-direct thermometric studies.) Angstrom also noted related work by Perntner (1888), Homen (1897), Exner (1903), K. Angstrom (1903) and Lo Surdo (1908.)
Abbott and Angstrom undertook two campaigns of observation: a preliminary season in Algeria (1912) and a more extensive one at Mount Whitney in 1913.
Both featured synchronous observations at different altitudes, including in the latter case an extension by means of automated instruments flown by balloon to 1500 meters—the balloon flights were made both from the Lone Pine site at the foot of the mountain and from the summit.
Angstrom presents eleven “principal conclusions.” They are too lengthy to quote in full here, but interested readers can find the paper online via Google Books.
In general, the conclusions deal with the influence of temperature and humidity upon what Angstrom terms “effective radiation.” This term is introduced to replace “nocturnal radiation,” since, as noted above, it occurs during the day, too:
VIII. There are indications that the radiation during the daytime is subject to the same laws that hold for the radiation during the night-time.
Less equivocally, Angstrom roughly quantified the effect of cloud altitude Wells had noted:
X. The effect of clouds is very variable. Low and dense cloud banks cut down the outgoing effective radiation of a blackened surface to about 0.015 calorie per cm.1 per minute; in the case of high and thin clouds the radiation is reduced by only 10 to 20 per cent.
A conclusion of great moment:
IX. An increase in altitude causes a decrease or an increase in the value of the effective radiation of a blackened body toward the sky, dependent upon the value of the temperature gradient and of the humidity gradient of the atmosphere. At about 3,000 meters altitude of the radiating body the effective radiation generally has a maximum. An increase of the humidity or a decrease of the temperature gradient of the atmosphere tends to shift this maximum to higher altitudes.
This very nearly amounts to a statement of today’s consensus view of the essential mechanism of the greenhouse effect: that greenhouse gases (specifically water vapor, in Angstrom’s conclusion) shift the effective radiating altitude higher as their concentrations increase.
The unstated corollary is that since higher altitudes are generally colder—the concept of the ‘lapse rate’ noted above—radiative intensity is thereby decreased.
Angstrom develops this idea further later on in the paper:
These considerations have given a value of the radiation from a perfectly dry atmosphere, and at the same time they lead to an approximate estimate of the radiation of the upper atmosphere, which is probably chiefly due to carbon dioxide and a variable amount of ozone. The observations indicate a relatively high value for the radiation of the upper layers—almost 50 per cent of the radiation of a black body at the prevailing temperature of the place of observation. Hence the importance of the upper atmosphere for the heat economy of the earth is obvious. . .
I think it very probable that relatively small changes in the amount of carbon dioxide or ozone in the atmosphere, may have considerable effect on the temperature conditions of the earth. This hypothesis was first advanced by Arrhenius, that the glacial period may have been produced by a temporary decrease in the amount of carbon dioxide in the air. Even if this hypothesis was at first founded upon assumptions for the absorption of carbon dioxide which are not strictly correct, it is still an open question whether an examination of the "protecting" influence of the higher atmospheric layers upon lower ones may not show that a decrease of the carbon dioxide will have important consequences, owing to the resulting decrease in the radiation of the upper layers and the increased temperature gradient at the earth's surface. The problem is identical with that of finding the position of the effective layer in regard to the earth's radiation out to space.
It’s an interesting passage for the son of Knut Angstrom to write—the elder Angstrom had strenuously argued against Arrhenius’ ideas. Also interesting is the fact that certain modern writers claim that Arrhenius was ‘falsified’ by contemporaries, but fail to note either Anders Angstrom’s data or his conclusions regarding carbon dioxide in the atmosphere.
Another aspect of this paper foreshadows later developments. Anders Angstrom was a physicist, like most of the researchers we have mentioned, from Greeley back to Pouillet.
But in 1904 another physicist, the Norwegian Vilhelm Bjerknes, had put forth the idea that his “primitive equations”—a set of differential equations providing for a reasonable mathematical description atmospheric flow—could provide the foundation for numerically-based weather prediction methods that could potentially be much more accurate than existing methods.
Increasingly, sophisticated study of the atmosphere would acquire a practical side, as the still-young but aspiring field of meteorology incorporated the mathematical tools provided by physics.
As has been emphasized on several occasions, our observations indicate that the atmospheric radiation in the lower layers of the atmosphere is dependent chiefly on two variables: temperature and humidity. Hence it is obvious that if we know the temperature and the integral humidity as functions of the altitude, we can calculate the radiation of the atmosphere at different altitudes, provided that the relation between radiation, temperature, and humidity is also known. It has been the object of my previous investigations to find this relation; hence, if the temperature and humidity at the earth's surface are known, together with the temperature gradient and the humidity gradient, I can from these data calculate the radiation at different altitudes.
It's not clear whether Angstrom saw such calculations as a practical matter when he wrote these words, but the practical applications would become increasingly prominent over the ensuing years.
Turning toward the meteorological side of atmospheric studies, we may consider the career of William Henry Dines (August 5, 1855-December 26, 1926.) Contemporaries called him a “giant,” and in many ways he exemplifies the development of meteorology in the first decades of the twentieth century, although in others he was a 'Victorian gentleman.'
Trained as a mathematician—he attained the coveted title of ‘wrangler’ at Cambridge, from whence he received a B.A in 1881—Dines was the son of a prosperous builder who was also a Fellow of the Royal Meteorological Society.
As talented in practical matters as mathematical ones, the younger Dines invented and built numerous meteorological instruments, and mounted several notable observational campaigns, including efforts to make systematic observations at high altitudes broadly similar to the Mt. Whitney campaign of 1913. For the last, Dines designed light and inexpensive autorecording instruments which could be lofted by kite or balloon, organized the observations, and analyzed the data.
From the 1920s on his attention turned toward the study of radiation in the atmosphere. He devised yet another instrument, the "ether differential radiometer." ('Ether' refers to the chemical formerly used as an anesthetic, not to the purported pre-Einsteinian medium for electromagnetic radiation.)
A 1920 paper described it thus:
It consists essentially of a sensitive differential thermometer, upon the bulbs of which radiation from any part of the sky, or radiation from a full radiator at a given temperature, can be directed. . .
The differential thermometer is formed of two ordinary glass test tubes each containing a few drops of ether. The tubes communicate with each other by a “U” tube of about 0.7 mm. bore containing ether to form a pressure gauge. As the change of vapour pressure of ether is large for small changes of temperature the thermometer is very sensitive to small changes in the radiation falling on either bulb. It suffices to build up the thermometer with rubber stoppers to the test tubes. A completely sealed-up pair of bulbs nearly exhausted of air would no doubt be better, but would be very fragile and liable to break; but the rubber stoppers answer quite well and the ether will serve for months without renewing.
The pragmatic, practical Dines is very much in evidence here!
Working with his son, L.H.G. Dines, he used his radiometer to make a study of sky radiation at his home in Benson, Oxfordshire—often referred to as “the Benson observatory.” The younger Dines describes the result thus:
The paper consists chiefly of four tables of monthly mean values of radiation from the sky observed at Benson, Oxfordshire, during the five years 1922-1926. The radiation is dealt with under two heads (1) Luminous rays, (2) Dark heat rays of wavelength exceeding about 2 micrometers. Each is measured under conditions of (1) Clear skies, (2) Completely overcast skies. The conditions under which the observations were made are stated and a few of the more salient features of the tables are briefly discussed.
The paper appeared in 1929 in the Quarterly Journal of the Royal Meteorological Society; the lead author is given as “The late W.H. Dines, F.R.S.” It proved a fitting memorial, directly inspiring or informing work by Sir David Brunt (1930) and Guy Callendar (1938)—and even work as late as Swinbank’s “Long-wave Radiation From Clear Skies” (1963.)
Guy Callendar’s The Artificial Production of Carbon Dioxide and Its Influence On Temperature was undoubtedly the most consequential of these.
Drawing not only upon Dines and Dines (1929), but upon Brunt (1930) and Angstrom (1918), it began Callendar’s ultimately successful campaign to update the ideas of Svante Arrhenius to which Anders Angstrom had referred, and to bring them once again into the scientific mainstream.
Perhaps partly due to the influence of Dines and Dines, it focused strongly upon the role of ‘sky radiation’ in the greenhouse effect. As Callendar’s abstract puts it:
The radiation absorption coefficients of carbon dioxide and water vapour are used to show the effect of carbon dioxide on “sky radiation.” From this the increase in mean temperature, due to the artificial production of carbon dioxide, is estimated to be at the rate of 0.003ºC. per year at the present time.
A particularly interesting feature of Callendar (1938) is that it was published with a “Discussion”—one which reads rather like a transcript of a Doctoral dissertation Defense. (This ritual ‘trial by fire’ forms a normal part of the Doctoral process today.) Callendar apparently faced a committee of leading meteorologists.
To begin with, there was L.H.G. Dines, whom we have already encountered, and Sir David Brunt, who Guy may well have known through mutual connections with the Imperial College. (At that time it was still simply 'Professor Brunt'--he would only be knighted for his contributions to meteorology in 1949.)
Also present were Mr. J.H. Coste, Drs. C.E.P. Brooks and F.J.W. Whipple, and Sir George Simpson, who was perhaps the most famous of all--he had worked in places as diverse as India and Antarctica, and had acquired tragico-romantic luster as a surviving member of the Scott expedition.
All in all, the committee must have been highly intimidating: Whipple had only recently completed his term as President of the Royal Meteorological Society; Simpson's presidency was just a couple of years in the future; and Brunt would, as it turned out, succeed Simpson.
Their questions seem a mix of admiration, condescension toward an outsider (albeit an outsider who was the son of one of England’s leading physicists), and proper scientific skepticism:
Sir George Simpson expressed his admiration of the amount of work which Mr. Callendar had put into his paper. It was excellent work. . . [But Simpson] thought it was not sufficiently realized by non-meteorologists who came for the first time to help the Society in its study, that it was impossible to solve the problem of the temperature distribution in the atmosphere by working out the radiation. The atmosphere was not in a state of radiative equilibrium. . . temperature distribution in the atmosphere was determined almost entirely by the movement of the air up and down. . . One could not, therefore, calculate the effect of changing any one factor. . .
Callendar replied that:
. . . he realized the extreme complexity of the temperature control at any particular region of the earth’s surface, and also that radiative equilibrium was not actually established, but if any substance is added to the atmosphere which delays the transfer of low temperature radiation, without interfering with the arrival or distribution of the heat supply, some rise of temperature appears to be inevitable in those parts which are furthest from outer space.
At the same time in America, a young physicist was working on meteorological problems. His name was Walter M. Elsasser, and he had left a Europe increasingly shadowed by Adolf Hitler. Elsasser (March 20, 1904-October 14, 1991) was employed at the Meteorology Department at CalTech. Although he had no background in meteorology—his work in Europe, which had twice laid foundations for Nobel-winning studies by others, had been in sub-atomic physics—he obtained a modest position addressing the problem of atmospheric radiation:
The American Meteorological Society had for years been urging the American Physical Society to do research on infrared radiation in view of its importance for the atmosphere, but nothing was done. The far infrared spectra of the atmosphere have a very complicated structure that required an understanding of quantum mechanics. . .
Walter spent the years 1937-41 analyzing the properties of far-infrared atmospheric radiation from first principles. After lengthy calculations, he ended up with tables and graphs that a practitioner could use to find the cooling and heating of the atmosphere when he knew the distribution of temperature and moisture. . .
Walter then set out to measure the far-infrared transmission of the atmosphere along paths up to 300 meters in length. This was done on an athletic field next to CalTech with equipment he built. . .
Elsasser published “Heat Transfer by Infrared Radiation in the Atmosphere” in 1942. It essentially carried out the project Angstrom had proposed: accurate calculation of heating and cooling due to radiation in the atmosphere. (Interestingly, Elsasser’s work was reviewed by none other than C.G. Abbott, but sadly this writer has not yet been able to access that review.) The work today is much-cited in the scientific literature, boasting 248 citations in Google Scholar as of this writing. Refinements, extensions and validations of his work were proposed well into the 1960s.
The interest was two-fold: practically, Elsasser had provided a way for forecasters to quickly calculate radiative cooling or heating in specific situations where temperature and humidity were known: this was to be done by a graphic ‘calculator’ printed on the back cover of the original book. In the reissue from the early 1960s the values were also presented in tabular form.
On the theoretical side, Elsasser’s results could be—and were—compared with other similar calculations. The goal was to refined and improve understanding of the radiative behavior of the atmosphere.
As Kondratiev and Neilisk wrote in 1960:
One of the most reliable and widespread methods of calculating thermal radiation fluxes in the atmosphere is the usage of radiation charts.
As nowadays this is a number of radiation charts based on various principles, their comparison and study is of considerable interest. . . We have examined seven radiation charts by the following authors: F.A. Brooks, A.A. Dmitriev, W.M. Elsasser, R. Mugge & F. Moller, F.N. Shekhter, G. Yamamoto.
On this theoretical front, too, the use of computers was increasingly important. This had been pioneered by Canadian-American physicist Gilbert Plass in the 1950’s.
(Plass, who worked for Lockheed on the practical problem of using infrared radiation to guide missiles, “moonlighted” on studies of the radiative effects of atmospheric CO2, and was the first to use computers to perform the lengthy calculations necessary to achieve a really accurate quantitative picture of radiative transfers throughout the atmosphere.) Elsasser’s work naturally informed an important strand of this burgeoning research.
But physics always comes back to measurement, and measurement always relies on technology. And already in 1954, another milestone was reached in the measurement of the “cloud by night.” Sidney Stern and Frederick Schwartzmann published a study in the American Meteorological Society Journal describing their:
. . .long-wave radiometer . . . whose response is essentially independent of air flow and ambient temperature changes. By shielding the sensing element with a hemispherical “window,” the integrated value of the back radiation from the sky can be measured.