Humankind lives at the bottom of the sea of air, and climate change is perceived by us mainly as a change in the overall conditions of this sea's lower layers. The atmosphere lies at the heart of decade-to-century climate change: It filters the sun's rays as they reach the surface of the earth and as they are reflected again into outer space, and it is the principal medium of exchange of heat, water, trace-gases, and momentum between the other components of the climate system—oceans, land surface, snow, ice masses, and the biosphere. The atmosphere and ocean are intimately coupled within the climate system, and are governed by similar physical laws. But the atmosphere has been explored in greater detail—in terms of available observations and of existing models—than any other component of the climate system. Thus it is natural to review the results of this exploration first, as we begin our examination of climate variability.
Understanding of natural phenomena proceeds through a sequence of observations, experiments, and models. Given the complexity of the climate system, laboratory experiments can reproduce only very incompletely the system's major aspects, and have not been included in the present volume. Atmospheric observations have led, in past centuries, to very simple, purely descriptive models of atmospheric motions. In the second half of this century, advanced computer models of the atmosphere have simultaneously benefited from an increase in the number and quality of observations, and stimulated vast field programs designed to verify model results and yield the new details necessary for improving the models. The separate treatment of observations and models in this chapter and the next is, therefore, only a matter of expository convenience.
The oldest instrumental records of atmospheric temperature—and, to some extent, precipitation—extend about 300 years into the past. The coverage and density of these measurements have grown more or less continuously, with a dramatic increase occurring in the 1940s and 1950s. This permits us to make a fairly informed assessment of past interannual variability, but we have considerably less confidence about interdecadal changes and only little or indirect information on the century-to-century time scale.
The Atmospheric Observations section below starts with a paper by H.F. Diaz and R.S. Bradley that addresses head-on the question of how different the climate of this century has been from those of previous ones. Proceeding from temperatures (which tend to be more uniform in space and time) to precipitation (which is considerably less so), S.E. Nicholson looks at the socioeconomically critical issue of African rainfall variability on interannual and decadal time scales. J. Shukla provides complementary insight on the initiation and persistence of drought in the Sahel.
The role of snow cover in the radiation balance at the surface makes it an important player in climate variability; this role is discussed by J.E. Walsh. Variability and trends of both liquid and solid precipitation over North America are reviewed by P.Ya. Groisman and D.R. Easterling. Returning to temperature, a careful study of the difference between trends in daily temperature maxima and minima is presented by T.R. Karl and his colleagues. C.D. Keeling and T.P. Whorf then analyze the decadal oscillations in global temperatures and in atmospheric carbon dioxide. These oscillations are at the heart of understanding natural variability on this time scale; they are also covered later in this section in the essay introducing atmospheric modeling.