12.4. Tutorials

12.4.1. A single stack releasing a single gas

The simplest case will be addressed first. This scene contains a single factory stack that is releasing SO2 gas. The release rates used in these examples are meant for demonstration purposes and do not represent actual observed releases. For this case, in the directory TUTORIALS/ONE open the file plume1.cfg in cfg_edit and notice under the Options tab that the plume model and truth outputs have been turned on. You will have to alter the particular paths to match your current DIRSIG directory structure. This is done in the PATHS section of cfg_edit.

Under the "Plume" button, the following window will appear:

Next, open the file plume1.rel in a standard ASCII editor such as vi or nedit. Notice the entries for the COMPOUND NAME (SO2). There must be a file named SO2.gas in the directory specified in the ABSORPTION PATH section of the PATHS button in cfg_edit. At the bottom of the file, under the RELEASE section, the string of numbers correspond to the release parameters specified in the comment line directly above. Remember that in the current DIRSIG, the "time" variable is not used. These variables are:

  RELEASE {
    0.0 5.0 350. 301.15 3.6 0 40 10.0
  }
  

and represent the plume release parameters as described in the following table:

To run the simulation, in the directory where the config file resides create the ADB and IMG directories (if they do not already exist) with the command mkdir ADB IMG. This is where the config file is set up to place the atmospheric database and output image files. (Note that this is not required and can be changed in the config file manually or through cfg_edit.)

Next, create the appropriate atmospheric database with the command:

    prompt> make_adb plume1.cfg >& adb.log
  

Once the Atmospheric Database File (ADB) is successfully created, run DIRSIG by executing the command:

    prompt> dirsig plume1.cfg >& plume1.log &
  

This will run DIRSIG on the configuration file plume1.cfg and will send both the standard output and the standard error reports to the file plume1.log. The scene is small and contains only a few objects so it should not take too long to render (about 30 minutes on a Sun workstation).

The result of the simulation should be two image files (and associated header files) in the directory IMG. The truth output, plume1.T.img, will contain all the truth outputs selected in the configuration file, including the plume truth outputs. The synthetic image is the file plume1.img. An image taken at Band 22 (8.83 microns) is shown below along with a sample spectrum taken near the plume source. In the image, the plume is seen over the plates that have been placed out in the field of grass. Looking at the truth imagery will show how far downwind the effects of the plume are computed. In this image the plume is only seen for a short distance downwind (the black border was added to more clearly show the image). The SO2 feature between 8.5 and 9 microns is clearly evident.

Now, go back and change the name of the output files (under the "Platform"/"Band"/"Edit" and the "Truth" sections in cfg_edit), change some of the plume release parameters (such as release rate, wind speed/direction, or release temperature) and re-run the simulation. Notice how the plume radiometric and spatial characteristics are determined by these variables. This is the simplest plume release case. The next example is a slightly more realistic case with a single stack releasing multiple gases, in this case two. The extension to more gases is straightforward.

12.4.2. A single stack releasing two gases

This example considers the case of a single stack releasing two gas species, still within the context of the simple scene described above. The appropriate files are found in the "TWO" directory.

The steps to running this simulation are similar to the first example. Note however here, that in the plume release file, plume2.rel, there are two entries under COMPOUND:

  EFFLUENTS {
      COMPOUND {
	  NAME = SO2
	  WEIGHT = 64.0
      }
      COMPOUND {
	  NAME = CH3Cl
	  WEIGHT = 50.0
      }
  }
  

Here, two gases will be release from the same stack, SO2 and CH3Cl. Note that both files must separately exist in the ABSORPTION_PATH directory. Also note that at the bottom of the file, the RELEASE parameters are slightly different:

  RELEASE {
      0.0 5.0 350. 301.15 3.6 0 40 10.0 25.0
  }
  

In the first example, this line ended with the release rate of SO2 set to 10 g/s. Here, a second release rate of 25 g/s has been added for the CH3Cl. Note that all the other release parameters are the same for both gases. All species from the same stack are released with the same temperature and velocity.

Run this tutorial simulation exactly as before. Create the atmospheric database first with the command:

    prompt> make_adb plume2.cfg >& adb.log
  

And then run the simulation with the command

    prompt> dirsig plume2.cfg >& plume2.log &
  

Actually, because the only changes that have been made are to the plume parameters, the same atmospheric database can be used for both examples in this section. The output of the simulation, as in the first example, will be two image files and their associated headers. The file plume2.T.img contains the truth imagery for this simulation, and plume2.img contains the rendered image. Note that the truth output image now includes two "Plume Column Densities", one for each gas:

The figure below shows two spectra taken from the image near the stack. The black curve shows the spectrum taken from the plume while the red curve is a spectrum taken near but off the plume. Both features can be seen in the plume spectrum, the SO2 feature between 8.5 and 9 microns, and the CH3Cl feature between 9.5 and 10.0 microns. A second feature due to CH3Cl is visible around 13 microns as well.

Compare these spectra with those in the previous example. The emission feature due to the second gas is clearly evident. Again, adjusting the release parameters will change the spatial and spectral characteristics of the plume. For this case, changing the relative release rates will change the relative strengths of the two gas emission features.