|
| |
Examples
of Ensemble Forecast Presentations |
 |
This page shows a selection of the types of products which are produced
from the ECMWF Ensemble Prediction System to help customers understand
the uncertainty in forecasts looking several days ahead. The examples
shown are all historical examples, but all these are generated routinely
every day and can be provided to customers on request. For more details
please contact the Customer
Centre.
|
|
The Ensemble Meteogram shows the range of uncertainty in a weather
parameter such as windspeed at a specific location, plus some indication
of the most probable values. At each forecast time a symbol is plotted
showing the full range possible (usually with 95% confidence), the
central range with 50% probability and the mid-point (median) of the
ensemble distribution:
In the example below a solid line is drawn through the median values to
give an indication of how the mid-point of the forecast range varies
with time.
|
Probability charts can show the probability of a defined event
occurring at different places at a specific time or within a given time
period. The following examples show the probability of the wind speed
exceeding Beaufort Force 6, 8 and 9 around the British Isles on a
particular day.
|
|
| Stacked probability graphs |
|
Stacked probability graphs show the probabilities of exceeding several
different thresholds of a given parameter at a specific location over a
forecast period. They can be particularly useful for identifying
"weather windows" when the risk of weather disrupting a user's
operations is minimal. This example is for ensemble forecasts of
wave height at a location in the North Sea. The scale up the left-hand
side shows probability. It can be seen that if an operation is to be
performed which requires wave heights of below 4.0m (red) then a safe
weather window occurs between 36 and 60 hours ahead, and with only a
small risk up until 108 hours.
|
|
|
|
Tropical cyclones (including Atlantic Hurricanes and Pacific Typhoons)
are among the most dangerous weather systems on Earth. Predicting their
movements is therefore particularly important for operations at sea in
the tropics and for the areas where they are likely to make landfall,
but precise prediction is notoriously difficult. Use of ensembles
allows us to estimate the uncertainty and therefore the area at risk of
being struck. The charts below show two different presentations of
tracks for Hurricane Isabel as it approached the East Coast of the USA
in September 2003. The first shows the spread of tracks from individual
members of the ensemble, colour-coded by forecast time to show how many
hours ahead a place may be at risk. The observed track is shown in
black butterflies, with the larger symbol indicating the position at
the start of this particular forecast. The single forecast from the
ECMWF high-resolution model is shown with black circles. It can be seen
that the single forecast predicted landfall too far NE but that the
actual landfall is well-within the uncertainty indicated by the
ensemble.
The second chart shows the same forecast but plotted as the probability
that the storm will pass within 75 miles of any point during the next 3
days. This format eliminates some of the more extreme tracks predicted
by the ensemble and focuses on the areas at greatest risk in the
immediate future. The actual track is well-captured, slightly to the
right of what is predicted as the most probable.
|
|
|
|
|
Home 
