Modeling Climate
What is climate and why do we want to model it?
There are basically two main types of forecasting; daily weather predictions
and climate projections. The first deals with relatively short time scales
such as one to several days in the future, the second deals with much longer
time scales such as years or even centuries. Essentially, climate can be
thought of as being a long time series of daily weather averaged over extremely
long time periods. Among other things, if enough daily weather is accumulated
over time, one can establish trends which would indicate if the long-term
climate is systematically changing or not. An example of this is a record
of global mean surface temperature departure taken over the past one hundred
years shown below.
Here, it may be seen that there has been a gradual rise of global mean
surface air temperature during this time period. The question, then arises
as to whether this represents a trend toward future global warming or is
this simply an example of natural variability such as that illustrated
below which shows the paleoclimate record of both global surface air temperature
change and CO2 concentration over the past 400,000 years.
It is questions like this that have encouraged research scientists to study climate and it's potential future change if any. To attempt to answer these questions, research scientists need to understand the various physical processes that cause the earth's climate to behave the way it does. These processes include winds, temperature, rainfall, snowfall, cloud cover, water vapor, surface hydrology and radiational processes. In particular, long-term changes in the earth's climate could have serious implications for our society as a whole in many different ways.
How do we model climate?
To do this, atmospheric scientists have developed a "tool" called the
general circulation model (GCM) to study a variety of problems ranging
from daily forecasting to long-term climate change caused by increased
carbon dioxide and other greenhouse gases. In simple terms, a GCM is a
complex mathematical model composed of a system of differential equations
derived from the basic laws of physics and fluid motion and modified to
consider air as a fluid. These equations describe the dynamic, thermodynamic,
radiational and hydrologic processes of the atmosphere. The processes of
a typical GCM are illustrated below
where the various arrows and diagrams denote the main components of
the model. The equations represented by these arrows and pictures are too
complex to be solved analytically, so they must be converted into an arithmetic
form that is suitable for computation by a large digital computer. This
alternate mathematical form is generally referred to as "finite differences".
The next step is to divide the entire three-dimensional atmosphere into
a systematic series of "boxes" to which the basic equations must be applied
and evaluated. In each box, the motion of the air (winds), heat transfer
(thermodynamics), radiation (long and short wave), moisture content (relative
humidity and precipitation) and surface hydrology (evaporation, snow melt
and runoff) are calculated as well as the interactions of these processes
among each of the boxes. The model is, then, programmed to run or "integrated"
on this finite difference network in a series of time increments or "time
steps" until the particular time period is completed which is usually for
many decades or even centuries. A typical grid system employed for
this purpose is illustrated by the figure below.
The current state-or-the-art general circulation models now consist of both an atmospheric and oceanic model coupled together and run or integrated as a unit to produce projections of climate change. The same type of description as described above also applies to the ocean model only in this case, the processes involved include ocean currents, convection, salinity, the thermohaline circulation and sea ice.
It should be noted that despite the complexity of GCMs, they are limited by both resolution (number of boxes) and our ability to describe the complicated atmospheric and oceanic processes in mathematical terms; in particular precipitation and cloud formation. However, despite these imperfections, GCMs remain the most viable means available to study climate and investigate its future change. In fact, these models are continually being upgraded and improved with the idea that they will provide better and better simulations of the processes which govern our climate and its potential changes. The advent and construction of larger and faster computers is also an important factor in this GCM development in that it allows the inclusion of more detailed processes than was previously possible. The main point here is that the more we understand climate and how its various mechanisms and processes work, the better we will be able to project the future and plan for it.
What types of topics are studied using climate models?
General circulation models have been extensively used to study a variety of scientific problems such as the present climate, greenhouse warming , oceanic circulation, hurricane tracking, mesoscale systems, atmospheric chemistry and past paleoclimates. These topics can be outlined in more detail as follows.
Present Climate. Probably the most important aspect of climate
research is to verify how well a given general circulation model simulates
or reproduces the current climate. This generally consists of comparing
various physical variables generated by the GCM to corresponding observations
such as temperature, winds, precipitation, solar and long wave radiation
and cloud cover. This is especially important since the simulation of the
present climate is generally used as a "control" experiment to estimate
the effects of climate change due to a variety of forcings. An example
of this type of analysis is given below which consists of computed and
observed precipitation anomaly fields taken during EL-Nino years for the
1951-2000 December-January-Feburary period. Top panel is the computed distribution
obtained from the GFDL Climate Diagnostics Group Model and the bottom panel
is the corresponding Global Precipitation Climatology Project (CPCP) data
set of observations. Blue coloring denotes regions of decreased precipitation
and red denotes regions of increased precipitation.
Greenhouse Warming. GCMs have been used to estimate the effects
of increasing greenhouse gases up to and beyond the 21st century. Generally,
these studies are carried out with coupled atmosphere-oceanic GCMs in response
to a variety of greenhouse gas and aerosol growth scenarios ranging from
"no controls on emissions" to "severe controls on emissions". Analysis
here is focused mainly on changes of both temperature and hydrology. In
particular, theoretical changes of hydrology are extremely important for
estimating the potential effects of greenhouse warming on water resources
and agriculture which could impact seriously on our ability to grow crops.
An example of a surface air temperature response to greenhouse warming,
obtained from the GFDL Climate Dynamics Model, is given below.
Ocean Circulation. Oceanic modeling has been used for investigations
into the dynamics of the large-scale ocean general circulation in order
to gain a fundamental understanding of the ocean's role in the earth's
physical climate system and global biogeochemical cycles. In addition,
studies have also been conducted to gain insight into the mechanisms for
paleoclimate fluctuations and transitions, and ocean ecosystem dynamics.
In particular, ocean models have been used to simulate and understand the
well known reoccurring oceanic phenomenon of the El Nino-Southern Oscillation
(ENSO) and accompanying La Nina, both of which can have far reaching consequences
for weather patterns thousands of miles from the tropical Pacific Ocean.
Below is a simulation of the surface oceanic circulation around Antarctica
using an ocean model with a typical resolution used for climate studies
(left hand) and a much higher resolution ocean model (right hand) which
resolves individual eddies similar to those observed in the actual ocean.
Both versions were developed by the Oceans and Climate Group at GFDL and
are being used to investigate their relative responses to different forcings.
Hurricane Tracking. GCMs have been extensively used to track
tropical storms and hurricanes (typhoons). These can range from the global
GCMs to embedded limited high resolution models such as that developed
by the Geophysical Fluid Dynamics Laboratory/NOAA. The obvious goal here
is to be able to issue timely forecasts of storm tracks, intensity and
probable landfall of these violent storms. Also studies are being made
to understand the mechanisms of the formation and development of tropical
storms in general. An example of this type of study is given below which
shows a three-dimensional simulation of Hurricane Andrew (August 1992),
obtained from the GFDL Hurricane Model, as it approached the East Coast
of Florida.
Mesoscale Systems. Limited high resolution models have been used
to investigate and simulate small-scale but violent convective storms such
as squall lines and thunderstorms. Using models of this type can lead to
greater understanding to the mechanisms of formation and development of
these dangerous storm and may ultimately lead to better forecasts of the
so-called "super cells" or "rotating thunderstorms" which can produce tornados.
The figure below illustrates a three-dimensional simulation of a mesoscale
cyclone, obtained from the GFDL Mesoscale Model, depicting cloud cover
(colored blue) associated with both warm and cold fronts as well as the
general cloud cover associated with the northern and western parts of the
storm system where steady precipitation occurs.
Atmospheric Chemistry. GCMs are being used to investigate the
upper atmospheric general circulation and the chemical interactions that
occur there as well as air pollution reactions in the lower troposphere.
Particular emphasis has been placed upon understanding the very complex
chemical and photochemical interactions that have acted together to create
the observed depletion of ozone or "ozone holes" over both the north and
south polar regions at very high levels in the earth's atmosphere. An example
of a stratospheric ozone hole simulation in Dobson units produced by GFDL's
Atmospheric Physics and Chemistry Group is given below where the red colors
denotes a minimum of ozone concentration and the blue colors denote ozone
concentration increases. This illustration represents a latitude-time distribution
over a single year which explains the fluctuations noted in the figure.
Past Paleoclimates. One of the credibility tests that GCMs have
been subjected to is how well they can simulate climate changes brought
about by past climatic periods in our earth's history. Examples of
this are the Cretaceous and the Last Glacial Maximum which represent abnormally
warm and cold climates respectively. Studies like these could yield
insight as to what might have caused these abnormal climates as well as
what the resulting changes in these respective climates might have been
like. More recently, these models have been run to simulate the change
in temperature and other meteorological parameters associated with the
Mt. Pinatubo volcanic eruption, a physical event for which there are good
observations to compare with. Below is a simulation of surface temperature
change during the Last Glacial Maximum (approximately 18,000 years ago)
as compared with the present climate obtained from a coupled climate model
developed at the British Hadley Centre, United Kingdom. Dark blue colors
signify regions of maximum surface temperature decreases due to the colder
simulated climate.
Institutions around the world that model climate
British Hadley Centre, Exeter, United Kingdom
http://www.met-office.gov.uk/research/hadleycentre/
Bureau Meteorology Research Centre, Melbourne, Australia
http://www.bom.gov.au/bmrc/
Canadian Climate Centre, Victoria, Canada
http://www.cccma.bc.ec.gc.ca/
Centre for Scientific and Industrial Research Organization, Melbourne,
Australia
http://www.csiro.au/
European Centre for Medium-Range Weather Forecasts, Reading, United
Kingdom
http://www.ecmwf.int/
Geophysical Fluid Dynanics Laboratory, Princeton, New Jersey
http://www.gfdl.noaa.gov/
German Climate Computing Center, Hamburg, Germany
http://www.dkrz.de/
Goddard Institute for Space Studies, New York, New York
http://www.giss.nasa.gov/
Laboratoire de Météorologie Dynamique du C.N.R.S., Paris,
France
http://www.lmd.jussieu.fr/en/
Lawrence Livermore National Laboratory, Livermore, California
http://www.llnl.gov/
Los Alamos National Laboratory, Los Alamos, New Mexico
http://www.acl.lanl.gov/
Main Geophysical Observatory, St. Petersburg, Russia
http://www.mgo.rssi.ru/
Max-Planck-Institute for Meteorology, Hamburg, Germany
http://www.mpimet.mpg.de/en/web/
Meteorological Research Institute, Tokyo, Japan
http://www.mri-jma.go.jp/
National Center for Atmospheric Research, Boulder, Colorado
http://www.ncar.ucar.edu/