The Pre-Discovery Stratosphere
Serendipity led to the discovery of the stratosphere in 1900. An innovative program of unmanned balloon soundings of the upper atmosphere by the French physicist Teisserenc de Bort unexpectedly revealed a large region with constant temperature that he named the "stratosphere". Since that time, the stratosphere has often been probed by balloons and airplanes as well as remotely from the ground and from Earth satellites.
Clouds in the conventional sense are found to be almost wholly lacking in the stratosphere. When a great volcano erupts, however, very fine debris can be lofted far into the upper atmosphere. Heavy ash particles fall out quickly, but the light sulfur gases combine with traces of water vapor to form sulfuric acid aerosols. This so-called "dry fog" can linger in the stratosphere for several years.
In 1883 the famous volcano Krakatau blew itself to pieces and heavily polluted the upper atmosphere with sulfuric acid aerosols. These tiny particles produced remarkable optical effects easily seen by ground observers around the world, even with the naked eye. Accurate inferences about high-altitude atmospheric winds and about the thickness of the supposed "dust cloud" were soon drawn by following the progressive changes of the striking visual phenomena. Today, we recognize that the observers were making pre-discovery observations of the stratosphere under the influence of a large volcanic eruption.
What exactly are the main diagnostic optical effects? There are a number of them, but four that can be easily detected without special instruments include a clear-sky dimming of the Sun and stars, red or purple twilight glows, reddish haloes around the Sun, and dark total eclipses of the Moon. Dimming of the sky's luminaries occurs because aerosols scatter away incoming sunlight and starlight. Ordinary twilight glows are enhanced by the additional reflection of sunlight from the stratospheric aerosol layer. Reddish haloes are simply anomalous diffraction patterns in the sky surrounding the Sun, and have been eponymously dubbed "Bishop's Rings". During a total eclipse of the Moon, sunlight passing close to the Earth is refracted by the atmosphere into the shadow cone, rendering the Moon's face faintly luminous and ruddy because the red rays are refracted more than the others. But if volcanic aerosols fill the stratosphere, the passing rays are mostly screened out of the shadow cone and so the Moon appears dark. If the eruption is large enough, the Moon can completely disappear from view.
Even without understanding the causes and implications of these strange optical effects, people during past centuries may have noticed and chronicled the phenomena. In fact, this witnessing of events actually happened. A thorough search of the surviving Western literature from ancient and medieval times by climatologist Richard Stothers (NASA/GISS) has unearthed strong evidence for a number of dry fogs since the earliest documented period around 600 BC. He finds that explosive volcanism was, on average, about as frequent a millennium ago as it is today.
The historical method of dry fog detection has picked up the same extremely large volcanic eruptions that the glaciologists' ice core method has, at its limit of maximum capability. Cores drilled in the Earth's ice-covered polar regions contain a record of annual layers of frozen snow accumulation. These layers are contaminated with various impurities, such as sulfuric acid, that have settled out of the atmosphere and been preserved in the ice. It would now appear that, compared with the ice core method, the historical method shows at least equal sensitivity and much better time resolution for ancient and medieval eruptions. The very accurate historical dates can therefore be used to improve the calibration of the acid horizons detected in the ice cores.
A good bet for the future is that volcanic debris in the pre-discovery stratosphere will be exposed more and more by the modern marriage of history and glaciology.
Stothers, R.B. 2002. Cloudy and clear stratospheres before AD 1000 inferred from written sources. J. Geophys. Res. 107, no. D23, AAC-17.