 |
| Titel |
Fire danger assessment using ECMWF weather prediction system |
| VerfasserIn |
Francesca Di Giuseppe, Florian Pappemberger, Fredrik Wetterhall |
| Konferenz |
EGU General Assembly 2015
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 17 (2015) |
| Datensatznummer |
250112387
|
| Publikation (Nr.) |
 EGU/EGU2015-12542.pdf |
|
|
|
|
|
| Zusammenfassung |
| Weather plays a major role in the birth, growth and death of a wildfire wherever there is
availability of combustible vegetation and suitable terrain topography. Prolonged dry periods
creates favourable conditions for ignitions, wind can then increase the fire spread, while
higher relative humidity, and precipitation (rain or snow) may decrease or extinguish it
altogether.
The European Forest Fire Information System (EFFIS), started in 2011 under the lead of
the European Joint Research Centre (JRC) to monitor and forecast fire danger and fire
behaviour in Europe. In 2012 a collaboration with the European Centre for Medium range
Weather Forecast (ECMWF) was established to explore the potential of using state of the art
weather forecast systems as driving forcing for the calculations of fire risk indices. From this
collaboration in 2013 the EC-fire system was born. It implements the three most
commonly used fire danger rating systems (NFDRS, FWI and MARK-5) and it is both
initialised and forced by gridded atmospheric fields provided either by ECMWF
re-analysis or ECMWF ensemble prediction systems. For consistency invariant fields (i.e
fuel maps, vegetation cover, topogarphy) and real-time weather information are
all provided on the same grid. Similarly global climatological vegetation stage
conditions for each day of the year are provided by remote satellite observations. These
climatological static maps substitute the traditional man judgement in an effort to create
an automated procedure that can work in places where local observations are not
available.
The system has been in operation for the last year providing an ensemble of daily
forecasts for fire indices with lead-times up to 10 days over Europe and Globally. An
important part of the system is provided by its (re)-analysis dataset obtained by using the
(re)-analysis forcings as drivers to calculate the fire risk indices. This is a crucial part of the
whole chain since these fields are used to establish the initial conditions from which the
forecast is subsequently run. The reanalysis dataset goes back to year 1980 (the starting
year of ERA-Interim integrations) and is updated in quasi real time. In addition of
providing the staring point for the operational forecasts it is a very useful dataset for the
scope of calibration and verification of the system. Assuming reanalysis fields are
good proxies for observations then, by comparison with fire events which really
occurred, this dataset can be used to assess the potential predictability of fire risk
indices.
In this work we will introduce the EC-fire system. Then the reanalysis dataset will be
used to identify regions of high fire risk predictability and where the system might be in need
of further refinement. |
|
|
|
|
|
|