Geoscience Reference
In-Depth Information
ocean models, however, suggest only a 1-2°C
surface warming for century-long transient or
doubled CO
2
experiments (see Chapter 13).
recent years. One methodology is to embed a
regional climate model into a GCM or AOGCM
in a certain region of interest and to use the global
model information as a boundary condition for
the regional model. The typical regional climate
model has grid cells of approximately 50km on
a side providing a higher resolution climate
simulation over a limited area. In this way, small-
scale effects such as local topography, water bodies
or regionally important circulations can be
represented in a climate or weather simulation.
These local effects, however, are generally not
transmitted back to the larger scale model at
present. In addition, regional models often
have a more realistic treatment of smaller scale
processes (e.g., convective adjustment), which can
lead to more accurate simulations.
2 Simpler models
Because GCMs require massive computer
resources, other approaches to modeling climate
have developed. A variant of the GCM is the
statistical-dynamical model (SDM), in which
only zonally averaged features are analyzed and
north-south energy and momentum exchanges
are not treated explicitly but are represented
statistically through parameterization. Simpler still
are the energy balance model (EBM) and the
radiative convective model (RCM). The EBM
assumes a global radiation balance and describes
the integrated north-south transports of energy in
terms of the poleward temperature gradients;
EBMs can be one-dimensional (latitude variations
only), two-dimensional (latitude-longitude, with
simple land-ocean weightings or simplified geo-
graphy) and even zero-dimensional (averaged for
the globe). They are used particularly in climate
change studies. The RCMs can represent a single,
globally averaged vertical column. The vertical
temperature structure is analyzed in terms of
radiative and convective exchanges. These less
complete models complement the GCMs because,
for example, the RCM allows study of complex
cloud-radiation interactions or the effect of atmos-
pheric composition on lapse rates in the absence
of many complicating circulation effects. Simpler
models are also important for simulation paleo-
climate, as these models can represent thousands
or even millions of years of climate history.
C DATA SOURCES FOR
FORECASTING
The data required for forecasting and other
services are provided by worldwide standard
three-hourly synoptic reports (see Appendix 3),
similar observations made hourly, particularly in
support of national aviation requirements. Upper-
air soundings (at 00 and 12UTM), satellite data,
and other specialized networks such as radar
stations for severe weather provide additional
data. Under the World Weather Watch program,
synoptic reports are made at some 4000 land
stations and by 7000 ships (
Figure 8.5A
). There are
about 700 stations making upper-air soundings
(temperature, pressure, humidity and wind)
(
Figure 8.5B
). These data are transmitted in
code via teletype and radio links to regional or
national centers and into the high-speed Global
Telecommunications System (GTS) connecting
world weather centers in Melbourne, Moscow and
Washington and eleven regional meteorological
centers for redistribution. Some 184 member
nations cooperate in this activity under the aegis
of the World Meteorological Organization.
Meteorological information has been collected
operationally by satellites of the United States and
3
Regional models
Because of the necessity of transferring climate
and weather information representing averages
over grid cells which are hundreds of kilometers
on a side to point scales where information can
actually be applied, a variety of downscaling
techniques have been developed and applied in
