Radiation and Atmospheric
Chemistry
· Introduction
· Radiation in Forecast Models
· Model Descriptions
· Eta
· ECMWF
· Sources
Introduction
The presence of long wave and shortwave radiation in the atmosphere profoundly influences weather and weather prediction. On the most basic level, differential heating of the Earth's surface by the sun creates horizontal temperature, and thus pressure differences. These pressure differences drive the circulation of the atmosphere. The Earth's surface primarily absorbs solar radiation and emits longwave radiation. Clouds and greenhouse gases both reflect and absorb incoming shortwave radiation, emitting longwave radiation both to Earth and back to space. Thus, the presence of clouds or pollution could ensure cooler daytime temperatures by blocking incoming shortwave radiation or warmer nighttime temperatures by trapping outgoing terrestrial radiation. In short, forecast model developers must take the atmosphere's energy budget into account. The image below (Takle 1995) depicts the energy balance of the earth-atmosphere system:
The radiation balance between shortwave and longwave radiation depends on a combination of temperature, clouds, concentrations of water vapor, ozone, carbon dioxide, and aerosols. These variables must be taken into account before a model can be run in order to accurately account for their effect on short-term weather phenomena. The developers of each model decide which of these variables to keep and which to neglect. The processes involved are extremely complex and therefore difficult to treat in detail as a variable component in forecasting models. It would be impractical to set aside the computer time necessary for radiation to be dynamic in daily forecasting model iterations. It is, however, the basis for Energy Balance Models and is important in the General Circulation Models used in climate applications.
Each of the various weather forecast models take radiational
effects into account. The parameters of each model allow radiation effects to
be found and inserted into the model both before initiation and during a model
run.
All the models utilize two separate radiation schemes, one for shortwave and
the other for longwave radiation; however, each model's schemes differ in their
simulation of the radiation budget. Shortwave and longwave radiation schemes
must both simulate the absorption of radiation by water vapor, ozone, carbon
dioxide, clouds and the Earth's surface. In addition, models must demonstrate
other phenomena such as the reflection
of radiation by clouds and the Earth's surface, the scattering of radiation by clouds and aerosols, and the re-emission of longwave radiation by
ozone.
Models ultimately use different variables, such as absorption and emission, as parameters in the radiative transfer equation and can thus calculate the net fluxes of radiation in the atmosphere.
Atmospheric chemistry is not a focal point of weather forecasting models as it is in the air pollution models, tracing models, or climate change predictions. Forecast models primarily use atmospheric chemistry as it pertains to the radiative properties of the model's simulated environment. As mentioned in the previous discussion of shortwave and longwave schemes, different chemical compounds have different radiative properties. Water vapor is by far the most important compound in the atmosphere in terms of radiative properties because it absorbs multiple bandwidths of longwave radiation emitted by black bodies such as the earth. Ozone plays a key role in the modeling of the stratosphere, if that portion of the atmosphere is included in the radiation scheme. Other compounds included in particular models due to their radiative properties are methane, carbon monoxide, nitrous oxide, oxygen, and aerosols. While the bulk of the earth's atmosphere is made up of nitrogen and oxygen, these gases are not important to radiation schemes because they are almost entirely transparent to short and longwave radiation.
Model Descriptions
The global spectral Medium Range Forecast (MRF) model developed by the National Meteorological Center uses the shortwave radiation scheme developed by Lacis and Hansen (1974). It takes into account absorption by ozone, carbon dioxide, water vapor, and clouds. Also included is ground surface and cloud surface reflection due to their respective albedos. The ozone and carbon dioxide distribution profiles for the scheme are based on seasonal climatological data. Snow cover is derived from climatology and is accounted for as a modification to the surface albedo.
Fels and Schwarzkopf (1975) developed the longwave scheme for the global spectral model in 1975 at the Geophysical Fluid Dynamics Laboratory. It, like the Lacis and Hansen shortwave model, takes into account how water vapor, ozone, carbon dioxide, and clouds absorb and emit longwave radiation.
In producing clouds, the GSM uses an interactive cloud simulation that is diagnosed from the predicted relative humidity of the environment. Cloud coverage is important in considering reflection and absorption of short- and longwave radiation, and emission of longwave radiation. Cloud reflection and absorption are assigned according to cloud height and type. Longwave emission by clouds is also determined by height and type, but is also affected by latitude.
The Eta model's radiation package is nearly identical to the
global spectral MRF. Some minor adjustments have been made to it in the past
few years.
Radiation: The current version of the radiation package used in the model is
one developed at the Geophysical Fluid Dynamics Laboratory (GFDL). The latest
Eta model now views clouds on each layer of the model. This helps account for
insolation and insulation effects of cloud coverage. The initial surface albedo
is taken from climatology, but is allowed to evolve during the forecast. The
atmospheric temperature tendencies arising from the radiative effects are
applied after every adjustment time step. A newer version of the radiation
package that directly utilizes cloud in all model layers is being tested. The
Eta model also includes a correction for the eccentric orbit of the earth, an
improved ozone scheme, and aerosols. Chemistry: The model's carbon dioxide and
ozone distributions are taken from climatology and held constant.
The Nested-Grid Model (NGM) was developed to help improve short-range forecasting. The NGM's radiation package was developed at NASA's Goddard Space Flight Center (Hoke et al. 1984). In comparison to the shared scheme of the MRF and Eta models, the NGM's radiation scheme contains the essential pieces of a radiation package, but is otherwise relatively simplistic. Longwave radiative effects are determined to be a function of ground temperature, atmospheric temperature, specific humidity, and cloud amount (cloud amount is determined solely by relative humidity). Shortwave radiation is computed as a function of specific humidity, cloud cover, surface albedo, and solar zenith angle. The only improvements made to the NGM since its inception has been related to its horizontal and vertical resolution. Subsequent updates or improvements incorporated into the MRF and Eta models have not been and will not be added to the NGM's radiation scheme, as the model was "frozen" in 1990 (an effort to create a record of forecasts and verifications for comparison). The NGM does not contain an atmospheric chemistry package.
The original version of the Rapid Update Cycle (RUC) was on a 60-km grid. It contained a simple radiation scheme where a surface energy budget was constructed including short and longwave radiation. Also, interactions between clouds and radiation were also included. No substantial atmospheric chemistry was included in the 60-km grid version of the RUC model.
When the RUC model was updated from a 60-km grid to a 40-km grid, the radiation scheme used in the MM5 (Mesoscale Model 5) was included. This scheme was developed by NCAR, Penn State, and it was originally documented by Anthes and Warner (1978). The new RUC model (RUC-2) is considerably different from the original RUC model due to its parameterizations of physical processes such as cloud microphysics (stable precipitation), turbulent mixing, radiation, and convective precipitation. The RUC uses the energy balance equation to predict ground temperatures. Also included are additions for attenuation and scattering by hydrometers. In the RUC, there was no atmospheric radiation at all, only a surface radiation budget with clouds diagnosed from relative humidity. The solar flux at the top of the atmosphere is now variable, taking into account the elliptical orbit of the earth around the sun.
The European Center for Medium Range Weather Forecast model uses a radiation scheme simulated after Orcrette's work (1991). For clear sky conditions, shortwave radiation focuses on aerosol scattering and the effects of the absorption by water vapor, ozone, oxygen, carbon monoxide, methane, and nitrous oxide using line parameters of Rothman et al. (1983). Clear sky longwave radiation focuses on the absorptive properties of water vapor, carbon dioxide, and ozone. The temperature/pressure dependence of longwave gaseous absorption follows Morcrette et al. (1986). The carbon dioxide level is set at 345 ppm.
Cloudy skies are dealt with separately. Shortwave radiation in the cloudy scheme focuses on the absorption and scattering properties of cloud droplets. Radiative parameters include optical thickness, single-scattering albedo linked to cloud liquid path, and prescribed asymmetry factor. Longwave radiation also focuses on cloud droplet absorption but neglects scattering. The droplet absorption is modeled by an emissivity formulation from the cloud liquid water path.
The Weather Research and Forecast model (WRF) is currently under collaborative development by the U.S. Air Force, NCAR, NOAA’s Forecast Systems Laboratory (FSL), NCEP and the University of Oklahoma’s Center for Analysis and Prediction of storms. When completed in 2004, the WRF will give forecasters the ability to analyze mesoscale features, such as thunderstorms, with far greater precision. Several short and long-wave radiation schemes are being tested as part of the physics package, including the GFDL (currently in place in the ETA model). Also in place at the moment are the RRTM, GSFC and simple short wave. In addition, the WRF will include an extensive chemistry package, which will allow, among other things, simulation of aerosols, forecasting of chemical-weather, and testing of strategies to ease air pollution. Designers emphasize, however, that all packages being tested now are merely placeholders until a consensus is reached on which specific options to include.
Forecasting models serve varying purposes for which different parameterization requirements are necessary. The usage of a particular model and the goal set forth in its design stage determines the necessary complexity of the radiation scheme. For a short duration mesocale model, such as the RUC, an elaborate radiation and chemistry scheme is not required. On the other hand, a model of longer duration and greater physical coverage, such as the ECMWF, demands a detailed radiation scheme with the addition of multiple chemical compounds, which have been proven to be important in the radiation budget. The still under construction WRF model promises an advanced chemistry package, though final parameters are still a couple of years away.
The ability of the model to forecast cloud
coverage is of key importance to the radiation budget as it affects minimum and
maximum predicted temperatures. Much of this ability is determined by the
model's capability of generating appropriate amounts of convection. This
depends on the convective parameterization scheme used by a particular model
and is a key focus for further research and potential improvements to forecast
models. In fact, designers of the WRF
model are working on an improved physics package that will better handle
convection.
Radiation schemes, although different, all
perform similar functions. Improvements are made to individual
models by improving the particular scheme used by the model. As the
understanding of the radiative properties of chemical compounds is increased,
especially aerosols, further additions and improvements will be made to radiation
schemes thus increasing the accuracy of forecast models.
Recommending a forecast model to a specific user depends on
the detail and accuracy required by the user. The best results,
obviously, will be given by the models that most closely simulate
reality. Radiation schemes range from very basic (NGM) to extremely
complex (all others, to varying degrees). For a model such as the RUC,
which is designed to be run quickly for short-range forecasts, it is
impractical to include a large atmospheric chemistry package because of the
computer time required. However, chemistry becomes more important when
performing longer-range model forecasting, such as when using the Medium Range
Forecast (MRF) or the ECMWF models
In contrast, there is a significant difference between the Eta and the NGM. The Eta model contains well-developed radiation and atmospheric chemistry schemes that allow for those physical processes to be taken into account when the forecast is calculated. The NGM, on the other hand, does not include any chemistry package and has a primitive radiation scheme relative to its counterpart. Therefore, an appropriate recommendation can be made to use the Eta model as opposed to using the NGM. Down the line the WRF will likely offer a sophisticated chemistry regime. . The team working on that component promises their model will “have the capability to simulate coupling between dynamics, radiation and chemistry”.
The ECMWF and MRF models also forecast
for similar time periods. However, unlike the relationship between the
Eta and NGM, there are no significant differences between the complexities of
the respective schemes of the ECMWF and MRF. Although different, they
both adequately attempt to include radiation and chemistry processes when
creating forecasts.
Ongoing research will necessarily require
review of which model is best suited for which task. For example, the coming of the WRF promises significant progress
in finescale weather analysis. Indeed,
completion of the WRF will remove one model option from the table. FSL plans to replace the RUC with the WRF,
while NCEP will employ the latter as a high-resolution nest within the ETA
model.
Radiation schemes are crucial to the accuracy
of all model outputs. Appropriately dealing with the energy input into
and the energy budget of the atmosphere is essential. There are no
forecast models that can operate adequately without an appropriately complex
radiation package.
Anthes,R.
A., and T. T. Warner, 1978: Development
of hydrodynamic models suitable for air pollution and other mesometeorological
studies.Mon. Wea. Rev., 106,
1045-1078.
Benjamin,
S. G., K. J. Brundage, L. L. Morone, NOAA/NWS, 1994: Implementation of the Rapid Update Cycle. http://maps.fsl.noaa.gov/tpbruc.cgi
Fels,
S. B., and M. D. Schwarztkopf, 1975: The
simplified exchange approximation: A new method for radiative transfer
calculations. J. Atmos. Sci., 32,
1475-1488.
Hoke,
J. E., N. A. Phillips, G. J. Dimego, J. J. Tuccillo, and J. G. Sela, 1989: The regional analysis and forecast system of
the National Meteorological Center. Weather and Forecasting, 4, 323-334.
Lacis,
A., and J. E. Hansen, 1974: A
parameterization for the absorption of solar radiation in the earth's
atmosphere. J. Atmos. Sci., 22,
40-63.
Morcrette,
J.-J., 1991: Radiation and cloud
radiative properties in the ECMWF operational weather forecast model. J.
Geophys. Res., 96, 9121-9132.
Staudenmaier, Jr., M., NWSO Sacramento, 1996: Western Region Technical Attachment No. 96-06 a Description of the MESO ETA Model. http://nimbo.wrh.noaa.gov/wrhq/96TAs/TA9606/ta96-06.html
Stephens,
G. L., 1984: The parameterization of
radiation for numerical weather prediction and climate models. Monthly
Weather Review, 112, 826-865.
Takle, G, Iowa State University, 1995: 1-12 Global Energy
Balance.http://www.iitap.iastate.edu/gcp/forcing/forcing_lecture.html.
______ , Atmospheric Model Intercomparison Project, 1997:
Summary Documentation of the AMIP I Models: Main Document Directory. http://www-pcmdi.llnl.gov/phillips/modldoc/amip/01toc.html
______ , National Center for Atmospheric Research, 1998: MM5
Modeling System Overview. http://www.mmm.ucar.edu/mm5/overview.html
______ , NOAA/Environmental Modeling Center, 1997: MRF Documentation Files.http://sgi62.wwb.noaa.gov:8080/web2/web2/tocold1.html
back to top
Return
to the Model
Forecasting page
Page written and
encoded by:
Christopher Rogers
Michele Nordeen
April 27, 1998
Page updated by:
Paul Heggen
Jerome
Daniel Jamieson
February 1, 1999
Page updated by:
Robert Lee
Kathryn
February 22, 2001
Page updated by:
Julie
Lauren
William
Trey
February 22, 2002