Environmental Engineering Reference
In-Depth Information
are not involved in the calculations.
The first index is the state of El Niño-Southern Oscillation in the tropical South
Pacific. Strong El Niños in this area favour upper-level westerly winds over the Atlantic
which typically reduce hurricane activity. The La Niña state favours hurricane activity, as
in 2000. The state of the Quasi-biennial Oscillation in the stratosphere is considered, as
there are usually more hurricanes when the equatorial stratospheric winds are from a
westerly direction than when they are from the east. The state of surface pressure in the
north-east Atlantic is included, as when this ridge is anomalously weak during the prior
autumn and spring periods the trade winds in this area are weaker. This means there is
reduced upwelling of cold water in the Canary Islands area and warmer sea surface
temperatures - which favour hurricane formation. Similarly the sea surface temperature
anomalies in three areas of the north, tropical and south Atlantic are included to give a
measure of the warmth of the ocean; warmer seas favour more hurricanes. Spring and
early summer sea-level pressure anomalies and zonal wind anomalies over the eastern
Caribbean have an impact through low pressure and easterly anomalies, indicating
enhanced seasonal activity, whilst positive values imply suppressed hurricane activity.
Surprisingly rainfall in west Africa shows some relationship with hurricanes. When
rainfall in the western Sahel in the previous August-September period is above average
and when August-November Gulf of Guinea rainfall of the previous year is also above
average, there are more strong hurricanes. Finally note is taken of the west-to-east surface
pressure and temperature gradients across West Africa between February and May, as
strong gradients are associated with greater hurricane frequency.
Using these indices as predictors based on the period 1950-97, statistically based
forecasts are made. Analogues are also examined for those years with similar precursor
climatic signals to the forecast year. If the analogue year was followed by a year of
increased hurricane activity it is assumed that the forecast year will also experience more
storms than average. So far the predictions have worked reasonably well, though like
many statistical models the significance of the variables can vary. For example, the
significance of the Sahel rainfall factor has declined since 1995. Further work is
continuing to improve the performance of the model. It can be found on
http://tropical.atmos.colostate.edu/forecasts/index.html.
droughts, as the result of a reduction in the number of rain-generating systems, usually
affect a large area and take many months to develop, though our techniques of longterm
forecasting are not good at predicting when the drought will finish. Tropical and
temperate-latitude cyclones can be predicted reasonably well, so that we know
approximately the areas they are likely to affect. On a smaller scale, the warnings of
tornado formation are announced for a large area, but precisely where the funnel clouds
will touch down on the surface is not known. It is probably impossible to forecast such
conditions for more than a few minutes ahead. Flash floods from a single thunderstorm
are in a similar category. We have to accept them as one of the micro-scale features of
our atmosphere that occasionally may cause devastation over a small area. The chance of
any one site being affected by them is very small.
