EdGCM Forums  

Go Back   EdGCM Forums > FAQ > EdGCM FAQ

Search FAQ Search FAQ
Search Word(s):
Matching Options:
Search in:

How do I setup mixed-layer ocean model runs for running simulations that model sea surface temperatures (SSTs)?
There is a QFlux Tutorial that should be read to understand how to set up the ocean to model sea surface temperature (SSTs)

Can I change the climate model’s boundary conditions files?
Boundary conditions in the GISS GCM include topography, land coverage (continent distribution), ice sheets, vegetation, surface drag coefficient, and ocean energy characteristics. Oceans use either monthly specified sea surface temperatures (SSTs) and sea ice, or mixed-layer depths and ocean heat transports in the case of predicted “Qflux” runs.

Many types of experiments require the user to alter the boundary conditions. Some of the best examples of this are paleoclimate experiments in which many geographic features of the Earth are altered (e.g., larger ice sheets, shifted continents, higher sea levels, etc.). Unfortunately, altering the boundary conditions and getting the GCM to run successfully with them can be a significant technological challenge. EdGCM includes some specific sets of boundary conditions for some common experiments (e.g. Last Glacial Maximum files for Ice Age experiments), but we cannot allow users to change the boundary condition files themselves at present. We can try to accommodate certain requests; however, because of the complexity of interactions between boundary conditions, initial conditions and the GCM code, many problems that sound simple are not, and we cannot guarantee implementation.

How do I change the orbital parameters?
The GISS GCM allows changes to the orbital parameters, but EdGCM 2.4.1 does not fully implement that flexibility. In the Power tools section of the Setup Simulations window, there are fields where you can provide different values for eccentricity, axial tilt, and omegaT. If you enter a value into any of the three spaces then all three will be used to specify the orbital parameters for the simulation. Sensitivity tests can be performed using alterations to the modern values of the orbit.

In order to change the orbital parameters for a sensitivity test please follow these steps:
  1. Create a new run.
  2. Go to the Power tools section of Setup Simulation.
  3. Set the three orbital parameters to the modern values instead of zero. The modern values for the orbital parameters are:
    Eccentricity: 0.0167
    OmegaT: 282.9
    Axial tilt: 23.44
  4. The eccentricity or axial tilt may be furthered modified, but do not change OmegaT.
Historical orbital parameters require complex calculations of the precession of the orbit, which are not provided by EdGCM 2.4.1. You cannot calculate OmegaT because the values come from a table created by the Fortran program Orbpar, which calculates the three orbital parameters as a function of year. References for these calculations are:
  • Berger, André L., 1978, Long-Term Variations of Daily Insolation and Quaternary Climatic Changes, Journal of Atmospheric Sciences, v. 35(12), p. 2362-2367.
  • Berger, André L., 1978, A Simple Algorithm to Compute Long Term Variations of Daily Insolation. Institut D'Astronomie et de Géophysique, Université Catholique de Louvain, Louvain-la Neuve, No. 18.
  • Russell, Gary L., Determination of Earth’s Orbital Parameters: Orbpar.sub and Orbpar.for
available at http://aom.giss.nasa.gov/srorbpar.html

How do I create a trend file?
There are two ways to create a trend for a simulation. The simplest way is to use the various Trend sections within the Setup Simulations window. Alternatively, you can create text files, provided that the file format can be read by a Fortran program. Fortran's inflexibility requires spaces to separate columns, specific number formats, and Unix line endings. The Fortran program reads the files as I4,T6,F15.8. For example:

1958    314.89999390
1959    315.70001221
1960    316.50000000
1961    317.20001221
1962	317.89999390
Look in the Input Forcings folder of EdGCM for examples of functional trend files. Mac users should edit the files with TextWrangler, making sure to select “Show Invisibles” and “Show Spaces” from the toolbar to fully see the trend file. Similarly, Windows users should make edits with EditPlus, selecting “Tabs and Spaces” from the View menu to fully see the trend file.

What can you tell me about the observed data for solar luminosity and greenhouse gases that are included with EdGCM?
The solar luminosity values and greenhouse gas values supplied with EdGCM are the actual data sets used by the NASA/GISS climate group in their continuing studies of climate change (see http://www.giss.nasa.gov/ for full information on GISS's research and data sets). The greenhouse gas values from 1958-2000 are based on direct measurements of atmospheric trace gas levels (samples of air are taken in casks at several "pristine" locations around the world). Values from 1850-1957 have been measured from bubbles trapped in ice cores. Values from 2001-2050 are projections.

The IPCC uses many potential scenarios for greenhouse gas increase. If you obtain any other scenarios, or generate some of your own, EdGCM will allo you to utilize data sets as input in the trends sections of the Setup Simulations window.

Detailed references and plots of greenhouse gas forcings may be found here: http://www.giss.nasa.gov/data/simodel/ghgases/

A paper on the calculation of past solar irradiance by Solanki and Fligge (2000) may be downloaded here [PDF file]: http://www.astro.phys.ethz.ch/papers/fligge/GL10884W01.pdf

To what extent can the vegetation boundary conditions be modified?
The GISS GCM includes eight vegetation "types": Desert, Rainforest, Woodland, Evergreen Forest, Deciduous Forest, Grassland, Shrubs and Tundra. Each type is defined by the physical characteristics of Visible Albedo, NIR Albedo, Water Field Capacity, Roughness Length, and Snow Masking Depth. Despite the simple sounding names of the vegetation types, each type is actually a compilation of many tens of other vegetation categories and is assigned at a 1 x 1 degree resolution. The GCM uses a "fractional" scheme in which each 8 x 10 degree grid cell in the model is assigned a fraction of each vegetation type, based on the corresponding 1 x 1 cells. Thus, every grid cell in the GCM is made up of multiple vegetation types.

We can change certain grid cells to be all one type, or we can change the physical characteristics of any vegetation type, but we cannot change just one type of vegetation over just one region. For example, the “Desert” characteristics are adjusted, then all geographic locations that have some fraction of "Desert" will acquire the new characteristics. There is no way around this problem, short of re-coding and recompiling the GCM.

Can I run a deforestation simulation?
Open the Setup Simulations window, click on the run called "Modern_Predicted SST" (included with EdGCM 2.3.5 only) from the Toolbar's run list, then click "Duplicate" in the toolbar to create a copy of that run. In the Setup Simulations window, give the new run a sensible run ID. Then in the "Input Files" section, choose the "V8X10_NoForest" file from the Vegetation popup menu. This input file replaces all forests with desert.

Where can I find a description of the ocean model used by EdGCM's "Predicted SST (QFlux)" technique?
The simple ocean model used for the predicted sea surface temperature (SST) simulations in the GISS GCM is called a mixed-layer ocean model, because it simulates only the upper portion of the ocean that is well mixed with respect to temperature and salinity. Mixed-layer ocean models allow the GCM to account for the heat capacity of a large volume of water. For modern runs, the mixed layer thickness is dependent upon geographic location and season. Such data are not available for paleoclimate simulations, which use a constant 250-meter-deep mixed layer.

By using the mixed-layer ocean model, SSTs can adjust to changes in the climate forcings, thus allowing for a two-way feedback between atmosphere and oceans. The GISS mixed-layer ocean model employs a "Qflux technique" which explicitly alters heat in each grid cell to mimic horizontal ocean heat transports (ocean currents). The Qflux technique is described more fully in several publications:

What do the abbreviations mean in the tables of variables produced in Analyze Output?
More documentation will be forthcoming for some of the variables in the GISS GCM's diagnostic tables. Some of the more commonly examined variables include:
P0 = pressure level zero (top of the atmosphere)
P1 = pressure level one (first layer in the model's atmosphere)
Z0 = surface of the earth
Z1 = first ground layer
Z2 = second ground layer
G1 = ground layer 1 (ditto G2, G3)
TG1 = temperature of first ground layer (ditto for TG2, TG3)
T1 = temperature of atmosphere layer one
T AIR = vertically integrated air temperature
SW = shortwave radiation
LW = longwave radiation
MC = moist convective
SS = super saturated

Can I get regional averages of climate variables?
The resolution of the GCM used in EdGCM (8˚ latitude x 10˚ degrees longitude) is not designed for local studies. However, obtaining average values over large regions can be useful with some variables.

The GISS GCM includes 23 pre-set regions that have averaged variables for those regions only. The monthly summary tables created in the Analyze Output section include a series of files with the suffix "_reg.xls," which are Excel files containing the regional averages for numerous variables. The region names are abbreviated in the Excel file, but the following figure (from Hansen et al., 1983, included with the EdGCM download in the Documentation folder) shows the exact locations:



There is no legend for this figure, but you should be able to determine which abbreviation matches which shaded region.

Although the GISS GCM supports "special regions," the ability to modify those regions is not supported. However, a plug-in for EdGCM that will allow users to pick their own regions is planned for an upcoming release.

Can I use EdGCM to study interactions between air pollution and climate?
The GCM at the core of EdGCM is a physical model only, and does not include code for dealing with chemical tracers such as air pollution. Research scientists interested in changing chemistry typically employ separate simple “box models” for that purpose, using the box model output to adjust the GCM boundary conditions (such as levels of greenhouse gases) as needed. No chemistry box models are included in the EdGCM suite


All times are GMT -5. The time now is 04:38.


Powered by vBulletin® Version 3.8.5
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.