Middle-latitude climates - The Global Climate System: Patterns, Processes, and Teleconnections - page 104

workshop on ongoing reanalysis of the climate system held in Boulder,
Colorado, USA. Amongst other things a white paper was presented outlining
a US national programme on three new reanalysis initiatives: R1979 -
reanalysis for the satellite era using TOVS/SSU (TIROS Operational
Vertical Sounder/Stratospheric Sounding Unit) data and aimed at the best
possible spatial analyses rather than the best temporal analyses; R1950 -
reanalysis for the era of upper air observations; R1850 - reanalysis for the era
of surface observations, which would provide globally consistent gridded
analyses of the key surface fields. The final workshop report can be found at
www.joss.ucar.edu/joss_psg/meetings/climatesystem as can the text of the
white paper.
4.3.3 Summary
This essay has attempted to describe and explain one of the most recent and
exciting developments in the field of climatology. It has briefly traced the
subject from the early days of simple instrumental measurements, through
the analysis of the data that they produced to the current situation where the
demands of the current computers require data that is evenly spread spatially
across and above the Earth's surface. As a consequence of the computer
revolution in meteorology and climatology in the second half of the twentieth
century, the uneven station distribution that had developed in the previous
150 years, almost entirely on 26 percent of the land surface (very few being
sited on Antarctica or Greenland), was deemed unsuitable. In addition,
because of changes in location and the local environment, station data were
not always comparable. The first step was to statistically manipulate the data
so that local inconsistencies were ironed out and this led to homogenized
data. These techniques were developed to such an extent that data sets
covering the whole globe were produced. The final step was to again mani-
pulate the data so that their uneven distribution over the Earth's surface was
removed and they were replaced, via various interpolation techniques, by
gridded values in both horizontal and vertical planes. This was and is the aim
of the various reanalysis projects. But what should always be remembered is
that data are only as good as the original measurement and ground truth will
always be required to check the outputs of the computers in the reanalysis
projects.
4.4 Using reanalysis
The enormous value of reanalysis is illustrated in the number of research
papers that use the data. The wide variety of applicability and utility of reanalysis
can be aptly demonstrated by looking at but a few example from among the