Section 1.1: The Components of Climate Research
7
the surface pressure in the Eastern and Western Pacific. It was first noted
by Hildebrandsson (1897), in the time series of surface pressure between
Sidney and Buenos Aires. It is evident in many places, but traditionally the
observation taken at Darwin (in the north of Australia) and the island of
Tahiti are used to define it. Unfortunately, the observation taken by Captain
Bligh and his Bounty are not available, but the re cord is sufficiently long
that the Oscillation can be established beyond any doubt. It is obviously not
an oscillation, i.e. a sinusoidal wave, rather a more irregular quasi-periodic
oscillation, but the out of phase relation between Tahiti and Darwin is clear.
When the pressure in Tahiti tends to rise, then it falls in Darwin and vi ce
versa. Sir Gilbert's statistical analysis of the observation, in this case a simple
correlation analysis, was clearly pointing to some unknown physical process
that linked together the Western and Eastern tropical Pacific (Philander,
1989; Troup, 1965). Similarly susprising is the North Atlantic Oscillation,
described by Walker as a correspondence in pressure between the Azores and
Iceland. The surface pressure in the two sites shows a strong tendency to
anticorrelate, with high pressure over Iceland corresponding to low press ure
over the Azores.
More modern studies have shown (van Loon and Rogers, 1978) that these
local indices are part of larger structures, big circulation patterns stretching
over the entire Pacific, in one case, and over the Atlantic and West European
sector in the other. At the beginning of the 1980's it was clear that much
larger patterns existed in the atmosphere, as will be discussed in Chapter 12.
1.2 The Components of Climate Research
The availability of large scale computers made it possible to overcome the
absence of crucial experiments. If crucial experiments could not be done
because the Earth was reluctant to be modified according to the scientists'
whim, the scientists would just do without the Earth and simulate a new
Earth using the same equation and physical systems that are thought to
be active in the real Earth. The simulated Earth, a very large ensemble of
mathematical equations, can be computed with computers and experiments
can be designed to verify hypothesis on the simulated Earth. Numerical experiments became the best approximation to the crucial experiments of the
other physical disciplines. A new mode of scientific investigation has been
added to theory and experiments: numerical experiments. Numerical simulations blend some features of experimental techniques and the abstraction
of theoretical investigation. In latest years, numerical simulations have come
to play apredominant role in climate research.
The appearance of large scale computers also made adefinite advance on
the way to better define the present state of the climate by allowing the statistical treatment of large amounts of data. Statisticians could elaborate more
and more sophisticated techniques to extract information from a large en-
7
the surface pressure in the Eastern and Western Pacific. It was first noted
by Hildebrandsson (1897), in the time series of surface pressure between
Sidney and Buenos Aires. It is evident in many places, but traditionally the
observation taken at Darwin (in the north of Australia) and the island of
Tahiti are used to define it. Unfortunately, the observation taken by Captain
Bligh and his Bounty are not available, but the re cord is sufficiently long
that the Oscillation can be established beyond any doubt. It is obviously not
an oscillation, i.e. a sinusoidal wave, rather a more irregular quasi-periodic
oscillation, but the out of phase relation between Tahiti and Darwin is clear.
When the pressure in Tahiti tends to rise, then it falls in Darwin and vi ce
versa. Sir Gilbert's statistical analysis of the observation, in this case a simple
correlation analysis, was clearly pointing to some unknown physical process
that linked together the Western and Eastern tropical Pacific (Philander,
1989; Troup, 1965). Similarly susprising is the North Atlantic Oscillation,
described by Walker as a correspondence in pressure between the Azores and
Iceland. The surface pressure in the two sites shows a strong tendency to
anticorrelate, with high pressure over Iceland corresponding to low press ure
over the Azores.
More modern studies have shown (van Loon and Rogers, 1978) that these
local indices are part of larger structures, big circulation patterns stretching
over the entire Pacific, in one case, and over the Atlantic and West European
sector in the other. At the beginning of the 1980's it was clear that much
larger patterns existed in the atmosphere, as will be discussed in Chapter 12.
1.2 The Components of Climate Research
The availability of large scale computers made it possible to overcome the
absence of crucial experiments. If crucial experiments could not be done
because the Earth was reluctant to be modified according to the scientists'
whim, the scientists would just do without the Earth and simulate a new
Earth using the same equation and physical systems that are thought to
be active in the real Earth. The simulated Earth, a very large ensemble of
mathematical equations, can be computed with computers and experiments
can be designed to verify hypothesis on the simulated Earth. Numerical experiments became the best approximation to the crucial experiments of the
other physical disciplines. A new mode of scientific investigation has been
added to theory and experiments: numerical experiments. Numerical simulations blend some features of experimental techniques and the abstraction
of theoretical investigation. In latest years, numerical simulations have come
to play apredominant role in climate research.
The appearance of large scale computers also made adefinite advance on
the way to better define the present state of the climate by allowing the statistical treatment of large amounts of data. Statisticians could elaborate more
and more sophisticated techniques to extract information from a large en-