The 1990s: the warmest decade of the twentieth century?
ACCORDING to the
surface temperature record which extends back to 1859, the
1990s was the warmest decade of the twentieth century. The global warming
industry goes on to claim that the 1990s was also the warmest decade of the last
millennium by reference to climate proxies (tree rings; ice cores etc). This
section attempts to confirm the relatively humble claim that the 1990s was the warmest decade of the twentieth century!
THE METHODOLOGY used here is certainly inferior to that
used to construct the surface temperature record which is
determined by dividing the globe into 2592 5o latitude x 5o
longitude grid-boxes and calculating a monthly temperature anomaly from the
1961-90 average for each box by averaging anomaly data from a number of
climatological stations/marine data sources within that grid-box (data
here).
Such an approach is such a herculean task that nobody is ever
likely to volunteer to attempt to replicate the construction of the surface
temperature record to try and confirm its accuracy.
THIS METHODOLOGY uses only land data available from the
Goddard Institute whereas the surface temperature record uses both land and
marine data. For each selected region, ten climatological station temperature
records were selected on the basis of two criteria:
1) long unbroken record spanning the twentieth century (where
possible) and much longer in some cases
2)
reasonable geographical spread over region
Neither
cross validation of temperature records using data from nearby stations was
performed, nor corrections to data for urban warming or any other possible
errors were attempted. However, long records with large gaps or long records
with data fluctuating more than what is expected for a temperature record were
avoided and better records sought.
THIS METHODOLOGY
is unsophisticated but by mainly using long temperature records of 100 years or
more, the real and significant long term variations exhibited by these
records provide a factual background against which to compare the temperature
data of the 1990s. A flaw of the surface temperature record which is perhaps not
widely appreciated is the low level of population of the majority of the
2592 grid-boxes for much of the record. The record commences in 1859 when
only 15% of the 2592 grid boxes were populated. By 1900, the percentage of
populated grid-boxes had risen to 40, by 1950, 70 and even in 2001, only 80%
of the grid-boxes were populated leaving 20% of the globe unrepresented by the
"global" temperature record. Grid-boxes with no data are effectively given a
monthly anomaly of 0.0o C which means the record
may be skewed by the anomalies of grid-boxes with data. Regional variations in
surface temperature anomalies have been documented (Parker, Jones, Folland &
Bevan, 1994) so it is a real possibility that regional anomalies recorded in the
minority of populated grid-boxes in say the 1930s and 1940s were
unrepresentative of the globe as a whole resulting in a "global" surface
temperature record which is unreliable. The methodology used in this exercise
will of course have the same problem of scarcity of data in some
regions of the world which the surface temperature record has. However, the
purpose of the exercise is to ascertain for each selected region, whether or not
the 1990s was the warmest decade of the twentieth century. Regions with no long
temperature records will be excluded.
THE
TEMPERATURE DATA from the ten selected stations in each region were used to
construct annual mean temperature anomalies from the 1961-90 average for each
station which were then presented as time series in graphs. Time series of
mean temperature for each season (winter; spring; summer and autumn) were also
presented graphically for each station. The annual temperature anomaly data for
the ten stations were then averaged and a time series of this average was
compared with that of the relevant hemispheric average (northern or southern) in
a graph. In these comparisons, the much greater variability shown by the
regional ten station average relative to the hemispheric average is owing to the
inclusion of marine data in the latter. Sea surface temperatures vary much less
than temperatures over land because of the much greater heat capacity of oceans
which take much longer to heat up or cool down. Decadal averages of the
"average" anomaly of the ten stations was also calculated to determine whether
or not the 1990s was the warmest decade of the twentieth century in the region. All ten warmest years in the surface temperature record 1859-2001 have occurred since 1987. A Table listing the ten warmest years 1900-2001 in each region was also constructed to see if the ten warmest years in these regions occurred post 1987 or whether warm years were scattered throughout the twentieth century.
THIS is an ongoing exercise and further regions will be added
when time and enthusiasm permit. The Table below shows the regions where this
analysis has been completed and indicates whether or not the 1990s was the
warmest decade of the twentieth century in those regions according to the
methodology described above. Click on the region hyperlink to view the results
of the analysis in each region.
Table showing results of analysis to determine whether or not
the 1990s was the warmest decade of the twentieth century
* All ten warmest years 1900 to 2001 occurred since 1987 in the northern hemisphere surface temperature record
References
Parker, D.E., Jones, P.D., Folland, C.K. and Bevan, A. 1994: Interdecadal changes of surface temperature since the late nineteenth century. Journal of Geophysical Research, 99, 14373-14399.
Freedom Flights Virgin Atlantic Lastminute.com
Wegelo
Thomas Cook
