Flood Forecasting and Early Warning: An Example from the UK Environment Agency - Reducing Disaster: Early Warning Systems for Climate Change - page 197

generation (Research & Development Programme - a joint initiative by Department
of Environment, Flood and Rural Affairs (Defra) and EA) (Defra 2004 ).
However, our focus in this chapter has been the formal fl ood warnings which the
Environment Agency have been disseminating in England for a signifi cant amount
of time. In recent times, developments in science and technology have allowed
improvements to their warning and forecasting system. As ever, there are further
advances that could be made. Forecasts are by no means accurate yet. There is
always a desire to improve. The Environment Agency currently assesses the fore-
cast accuracy using statistical measures to identify bias, systematic errors and areas
for improvement. The results would suggest that modelling technology, allowing
prediction into the future should still be considered in development, with the quality
and calibration of the models requiring further improvement. Similarly, fl ood warn-
ings are by no means, at all times completely accurate, or applicable to all, or given
with suffi cient notice.
Further research and development is underway to try to better understand uncer-
tainty in forecasts and to integrate climate information amongst other advances.
However on the whole, the fl ood warning service in England should already be
considered an excellent response mechanism to reduce the impact of fl ooding in
England. This was shown during Exercise Watermark, a large-scale test of all
aspects of England's preparedness for fl ooding which was deemed largely a success
(EA 2011b ).
Understanding of localised fl ooding mechanisms has been gathered, compiled
and shared by staff over the years. In fact, staff at the Environment Agency are a key
factor in the whole process. Their training, knowledge and dedication are critical at
all stages of forecasting and warning processes, involved in all aspects from inter-
preting models and tailoring warnings.
However a large amount of money has also been invested, particularly to improve
systems. Historical and real-time fl ood information now feeds into systems devel-
oped by leading companies in their own fi elds, allowing models to be run, data be
displayed and manipulated and messages distributed effectively. By having separate
systems for specifi c parts of the warning and forecasting service, each section can
improve simultaneously and separately, allowing faster and better improvements.
The result of all factors is that excellent fl ood warning and fl ood forecasting
systems and processes have been developed. These are resilient and adaptable but,
most importantly, effective. The systems allow large volumes of messages to be
disseminated in a timely manner, with suffi cient lead time, directly infl uencing the
behaviour and actions of communities at risk of fl ooding.
More importantly perhaps, this 'model' of a fl ood warning and forecasting pro-
cess is essentially replicable and could be transposed to another location. If scaled
appropriately and given suffi cient investment, this early warning system structure
and operations could be effective in other locations as well as England.