temperature did not continue to decrease with altitude but became constant or even increased above a height termed the tropopause. By the end of World War II, data from operational weather balloons, which could in the best case reach altitudes of 30 km, provided a picture of stratospheric temperatures and winds up to this altitude but with low spatial and temporal resolution. After World War II, rocket soundings at about a dozen locations extended these measurements to 65 km and occasionally higher. These data were enough to delineate the global variation of the vertical temperature and wind distribution but yielded little information on features with smaller temporal or horizontal scales and provided little information on the southern hemisphere’s atmosphere. Only the roughest picture of motions at the longest horizontal scales was available, and many studies were constrained by data availability to limited altitude and geographic regions. Craig (1965) presents a good discussion of knowledge and speculation at that time.
Similarly, a limited number of ground-based instruments in the pre-satellite era, mainly UV spectrophotometers, could provide only a rudimentary view of the vertical, latitudinal, and seasonal variations of the ozone distribution, with large uncertainties (Goody 1954). One particular puzzle was the nature of the atmospheric motions that transport ozone from high altitudes in the tropics, where it is produced by solar UV radiation and atmospheric chemistry, to low altitudes in polar regions and midlatitudes, where processes destroying it dominate (Craig 1965).
It was recognized that the transport of ozone and other gases, as well as heat and momentum, was important. Initially, such questions were addressed in terms of conventional fluid dynamics (Hunt and Manabe 1968). However, these early estimates were so uncertain that it was often impossible even to determine whether the ozone transport was northward or southward (NRC 1979).
The understanding of atmospheric dynamics was revolutionized beginning in 1969, when satellite instruments were launched to measure temperature and ozone in the stratosphere. The first downward- or nadir-looking temperature sounders on Nimbus 3 (Box 5.2) demonstrated that remote sounding techniques could provide global observations of atmospheric temperatures from the surface to mesospheric altitudes. Although the soundings from these and subsequent nadir-looking instruments had low vertical resolution, they were sufficient to allow the heights of atmospheric pressure surfaces to be calculated. The balance between the slopes of these heights and Earth’s rotation allowed scientists to make accurate calculations of stratospheric winds (Smith and Bailey 1985). These winds could be separated into winds in the east-west direction (along parallels of latitude) and into north-south wavelike perturbations. A review of early results clearly showed that the stratosphere and mesosphere were dynamically very active, with large (planetary)-scale waves propagating from the troposphere into the stratosphere and mesosphere in the winter hemisphere (Hirota and Barnett 1977).
The advent of horizon-viewing sounders provided temperature measurements with higher vertical resolution. These more detailed pictures revealed a wider range of atmospheric motions, including waves in the tropics with short vertical wavelengths, and provided more detail on planetary wave activities. The accurate and densely spaced measurements of temperature and the derived estimates of wind speed made it possible to study global transport in more detail. To avoid the uncertainties associated with conventional fluid dynamics, theoreticians recast their equations to essentially follow air parcels, leading ultimately to simple but accurate approximations. These showed how planetary waves, propagating up from the troposphere, could interact with the eastward wind motions, and thereby change the mean vertical and poleward circulation (Matsuno 1971, Andrews and McIntyre 1976). This explained the phenomenon of “sudden stratospheric warming,” in which temperatures in the polar stratosphere at an altitude of 30 km can increase by 30°C or more in a few days.
One particular scientific achievement should be noted. Brewer (1949) and Dobson (1956) had independently postulated a mean north-south overturning circulation in the stratosphere, in which air rises from the troposphere into the stratosphere in the tropics and then travels to high latitudes (in both hemispheres) where it returns to the troposphere. Observations of distributions of methane, nitrous oxide, ozone, and water vapor (all from the limb sounders) were used to test and validate these (then-novel) ideas and theoretical approaches to the calculation of net transport of these gases (Andrews et al. 1987). A related triumph was the observation of the tropical “tape recorder” (Mote et al. 1996), which significantly advanced and confirmed scientific understanding of stratospheric dynamics and motions (Box 5.3).
Remote sounding also provided information on the composition of the stratosphere. The first instruments to shed light on the global distribution of stratospheric ozone were the backscattered ultraviolet (BUV) on Nimbus 4 and the solar backscattered ultraviolet (SBUV) and the Total Ozone Mapping Spectrometer (TOMS) on Nimbus 7, which provided global measurements of the total ozone in a vertical column. In addition, the Limb Radiance Inversion Radiometer (LRIR), the Limb Infrared Monitor of the Stratosphere (LIMS), and the Stratospheric and Mesospheric Sounder (SAMS) on Nimbus 6 and 7 retrieved the distributions of ozone, water vapor, nitrogen dioxide, nitric acid, nitrous oxide, and methane. The power of these measurements is shown by the observations of nitrogen dioxide and nitric acid, present in concentrations on the order of only 10 parts per billion by volume.
The advances in our understanding of the cause and dynamics of the Antarctic ozone hole exemplify the productive interactions of satellite observations with in situ and ground-based observations and numerical models. As