MERCURE - Modelling European Regional Climate, Understanding
and Reducing Errors
MERCURE is a European framework IV project financed by
the European Commission under contract ENV4-CT97-0485
November 1997 - November 2000
Objectives and goals:
The project has the following five main objectives and associated
goals:
-
To improve regional climate models (RCMs) by understanding the source of
errors and reducing them by improving the model's representation of physical
processes (including increasing resolution). The focus will be on improving
the simulation of surface air temperature, reducing errors to 1 K over
most of Europe and 2 K locally, and precipitation, reducing errors to 1
mm/day over Europe. Higher resolution models will be expected to perform
better than equivalent standard resolution over their common area.
-
To improve the hydrological cycle in regional models via improved formulations
of land-surface processes, radiation and precipitation physics. New aspects
of the simulations will be validated by comparing with run-off, snow-pack
and soil-moisture data.
-
To assess the ability of regional models to reproduce observed precipitation
freqeuncy distribution, including the frequency of heavy events, not just
climatological means. The models should have realistic frequency distributions
and increases in resolution should improve the simulation on all scales.
-
To further characterise errors in regional climate simulations nested in
general circulation models (GCMs) derived from the GCM driving data. GCM
errors which induce large-scale errors in the RCMs will be identified and
GCM modellers will be encouraged to reduce them.
-
To provide a statistical-dynamical tool linking RCM and GCM simulations
to extend the length of RCM integrations. The method will be able to repruduce
the main features of long RCM simulations from large-scale driving data
given shorter sets of both for calibration.
Contribution of DLR:
Develop and use a statistical-dynamical postprocessor
-
Develop and verify statistical links between regional data from RCM simulations
and the large-scale driving data used by the RCM
-
Extend these statistical relashionships with the inclusion of a stochastic
post-processor to generate statistically consistent time-series of precipitation
and temperature
-
Validate the ability of the method to capture the features of multi-decade
RCM simulations using statistics from multi-annual simulations
-
Use the method to exend the validity of short RCM test simulations
Results are published in:
Busch U., Heimann D., 2001:
Statistical-dynamical extrapolation of nested regional climate simulations.
Clim.Res., 19, 1-13.
Top of page
| Last change 20 November 2001 |