Low (web) Quality - BALTEX
Low (web) Quality - BALTEX
Low (web) Quality - BALTEX
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
242<br />
Chances of the global-regional two-way nesting approach<br />
Philip Lorenz and Daniela Jacob<br />
Max Planck Institute for Meteorology, Hamburg, Germany; philip.lorenz@zmaw.de<br />
1. Introduction<br />
Large scale atmospheric processes influence smaller ones,<br />
which in turn affect the evolution of the regional climate:<br />
this paradigm is the basis for one-way nested simulations<br />
with regional climate models (RCMs) driven by data from<br />
general circulation models (GCMs).<br />
However, small scale processes are at least influencing large<br />
scale processes too and could have in key regions significant<br />
impacts on the evolution of the general circulation: this<br />
paradigm was the main motivation for the development of a<br />
two-way nested GCM-RCM climate model system, in which<br />
feedback from the RCM to the GCM takes place in a<br />
selected region (see figure 1). This feedback is accounting<br />
for processes which are not resolved by the relative coarse<br />
resolution of the GCM, but which are resolved by the finer<br />
resolution of the RCM.<br />
One-way nesting<br />
Two-way nesting<br />
GCM RCM GCM RCM<br />
Figure 1. Illustration of one-way nesting versus twoway<br />
nesting.<br />
Another motivation for performing two-way in contrast to<br />
one-way nested RCM simulations is a higher compatibility<br />
between the RCM's internal dynamics and the lateral<br />
boundary conditions. The GCM adapts to the large scale<br />
state of the RCM by the two-way nesting technique<br />
feedback, and is providing therefore more consistent<br />
boundary data.<br />
For investigating the former mentioned questions, a twoway<br />
nested global – regional climate model system has been<br />
developed and applied in several simulations. A description<br />
of the model system and introductions to different analysis<br />
will be given within the following sections.<br />
2. The two-way nested model system<br />
The developed two-way nested model system consists of<br />
Max Planck Institute (MPI-M) climate models, the GCM<br />
ECHAM4 and the RCM REMO. Both the spectral GCM<br />
ECHAM4 and the grid-point RCM REMO use the same set<br />
of physical parameterisations. The feedback from the RCM<br />
to the GCM on all prognostic variables takes place at every<br />
time of the GCM (24 min. in T42 resolution). More details<br />
about the two-way nested model system are given in Lorenz<br />
and Jacob (2005).<br />
space), and the RCM in a 0.5° (~55 km) horizontal<br />
resolution.<br />
3. Influence on the general circulation<br />
For the analysis of the effects of the two-way nesting<br />
approach on the general circulation, a stand-alone<br />
ECHAM4 run has been performed and compared against<br />
two-way nested GCM-RCM runs. At least for the two-way<br />
nested region covering the tropical Maritime Continent a<br />
positive influence on the general circulation was analyzed<br />
and published in Lorenz and Jacob (2005).<br />
4. One-way versus two-way nested RCM<br />
simulations<br />
Additional one-way nested RCM simulations have been<br />
carried out for some of the domains used for the two-way<br />
nested GCM-RCM simulations. The comparison of oneway<br />
versus two-way RCM results reveal a significant<br />
reduction of typical RCM boundary artefacts (like<br />
unrealistic precipitation close to the lateral boundaries) in<br />
the two-way nested simulations, and furthermore an<br />
influence on the interior of the regional model domains.<br />
5. Influence of the lateral boundary data update<br />
frequency in one-way nested simulations<br />
Most state of the art RCM’s use 6-hourly output from<br />
GCMs or (re-)analysis as lateral boundary data. Within<br />
the framework of the two-way nested model system it is<br />
possible to perform one-way nested RCM simulations<br />
using GCM output down to a time interval of 24 minutes,<br />
which is the internal time step of ECHAM4 in T42<br />
resolution. For the domain covering the Maritime<br />
Continent only very small differences were found between<br />
RCM runs with the usual 6 hourly update frequency and<br />
runs with an increased update frequency of 24 minutes.<br />
6. Conclusions<br />
An overview of the major results of the investigations<br />
within the two-way nested model system framework will<br />
be presented; and prospects and limits of the two-way<br />
nesting approach will be discussed.<br />
References<br />
Lorenz, P. and Jacob, D., Influence of regional scale<br />
information on the global circulation: a two-way<br />
nesting climate simulation, Geophysical Research<br />
Letters, Vol.32, L18706, doi:10.1029/2005GL023351,<br />
2005<br />
10-year integrations using observed sea surface temperature<br />
data (AMIP; 1980-1989) have been carried out with the twoway<br />
nested model system for different two-way nested<br />
regions. In all simulations the GCM was applied in a<br />
spectral T42 horizontal resolution (~250 km in grid-point