Note on Terrain input to CTDMPLUS/CTSCREEN Donna Schwede August 9, 1993 The purpose of this note is to address the most commonly asked questions related to preparing the terrain file required by CTDMPLUS/CTSCREEN. Topics include selecting the coordinate system, contour digitizing, judging the fit, and obtaining a hard copy of your results. A description of the required inputs and technical basis of the terrain preprocessor is provided in The Complex Terrain Dispersion Model Terrain Preprocessor System User's Guide (Mills et al., 1987). Since the publication of this user's guide, the programs have been slightly modified in order to be compatible with the CTDMPLUS menu driver. Users should refer to the User's Guide to the Complex Terrain Dispersion Model (Perry et al., 1989) for information about these changes. We are currently reviewing all aspects of the terrain preprocessing procedures and expect to eliminate some of the below-mentioned concerns with future updates to the terrain preprocessing programs. 1. Coordinate system The coordinate system used in CTDMPLUS/CTSCREEN is user defined. Three common coordinate systems are stack centered, hill centered, and Universal Transverse Mercator (UTM). Users may use any coordinate system as long as the same system is used for locating the terrain features, receptors, stack(s), and meteorological tower. A conversion factor is input to CTDMPLUS/CTSCREEN to convert from user units to meters. Users have encountered problems with coordinates that have too many significant figures. For example, mathematical errors, due to lack of precision, can occur in the FITCON program when coordinates with many (6-7) significant digits are used. Also, the terrain preprocessor programs currently output results in an E10.4 or E15.4 format. If more than 4 significant digits were input to the FITCON, the contours drawn by PLOTCON may appear as straight lines which bear little resemblance to the input contours. Also, the parameters input to CTDMPLUS/CTSCREEN to describe the hill will not be accurate. Using the UTM coordinate system can cause problems such as these. With all coordinates, use no more than 4 significant figures (e.g. use 6453, 6462, 6537; not 706453, 706462, 706537). 2. Contour Digitization 2.1 Base elevation considerations CTDMPLUS requires terrain information from the base elevation (i.e. lowest ground elevation considering the ground elevations of all stacks and the meteorological tower) to hill top. This information can be obtained in two ways: (1) digitize contours starting at the base elevation; keep in mind that the lower contours can strongly influence the overall shape of the three- dimensional (3-D) terrain feature, or (2) digitize the upper part of the hill (where the plume is expected to interact) and allow the terrain preprocessor (HCRIT mode 2) to extrapolate the upper hill information down to the base elevation. The method chosen will depend on the terrain. If terrain information is not available down to base elevation or if inclusion of contours down to the base elevation would significantly bias the fitted shape of the hill at the important elevations of likely plume interaction, HCRIT mode 2 should be selected. Using this mode, the user must choose the lowest critical elevation (at or below base elevation) and the number of critical elevations. The number of critical elevations should be greater than the number of input contours, but no more than 20. Since the HCRIT program requires little time to run, selecting a large number of critical elevations is not unreasonable. Also, the number of critical elevations does not effect the run-time for CTDMPLUS. 2.2 Vertical spacing of contours The number of contours selected for digitization will depend on the vertical resolution needed to capture significant changes in the slope and shape of the terrain. The terrain preprocessor does not require a constant contour interval of input contours. Since CTDMPLUS linearly interpolates between levels where terrain is provided to it, it is important to provide enough information, particularly at the heights where the plume will interact with sections of the hill. 2.3 Complete vs. incomplete contours A contour is considered to be complete when the distance between the first and last points of the contour is less than 1.0 E-15 user units. Therefore, when digitizing a complete contour, users should make the last point on a contour a duplicate of the first point. In some cases, it may be inadvisable to digitize a complete contour. An example of this is when a contour closes, on the lee-side of the hill, at a considerable distance from the hill center; inclusion of the complete contour would bias the fit of the hill on the important windward side. However, users should digitize at least a 180ø arc around the user specified hill center. The terrain preprocessor will complete contours using a simple reflection process about the input hill center. The location of the hill center is an input to the FITCON program that is used in the contour completion process. The input hill center does not necessarily have to represent the true center of the hill or the top of the actual hill; it is the point about which the reflection will occur for contour completion. Since the location of the input hill center influences the contour completion, users should experiment with various locations for the hill center before accepting the results from FITCON. Also, note that, the true hill top elevation should be input to FITCON, not the elevation at the hill center. The hill top elevation is used in obtaining the parameters for the best-fit inverse polynomial. There may be contour shapes for which the reflection process (contour completion) does not provide a good completion of the contour. For these cases, the user may want to complete the contour manually. To complete a contour manually, one can follow the trend of existing closed contours and natural divisions of the topography such as stream beds. 2.4 Multiple contours at the same elevation Some terrain features include multiple peaks and therefore multiple contours at the same elevation. Each contour of the set of multiple contours must be input as a complete contour (i.e. distance between first and last point of the contour must be less than 1.0 E-15 user units). The terrain preprocessor will connect the contours with a zero area segment and fit the set to one ellipse. If PLOTCON shows a line segment connecting the contours in the set, then they have not been input as closed contours and the resulting fit will be incorrect. 3. Judging the fit The PLOTCON program is designed to give users a graphical representation of the input terrain and the results of the terrain fitting process. Users should use this program to judge the mathematical fit to the actual terrain. First, the user should look at the results from FITCON. The elliptical representations should be reasonable approximations of the input contours. An ellipse should have the same orientation and total area as the input contour that it is approximating. The spatial relationship between individual ellipses is not important. It does not matter if individual ellipses cross each other. The appearance of the completed contours (e.g. overlap) is also of little concern as long as the elliptical representations adequately fit the input part of the contour. After the elliptical representations have been judged to be acceptable, then the 3-D fit to the hill should be assessed. Users should compare the lengths of the major and minor axes and the orientation of the axes (represented by the hill base shape) at each critical elevation with those of the actual terrain above that height. The 2-D "shape" of the hill base (elliptical, diamond, square) that appears on the screen is otherwise of no importance in judging the fit. The only parameters for the 3-D fit that are used by CTDMPLUS are the hill centers, length scales and the profile exponents. 4. Hard copy output from PLOTCON The options available for obtaining a hard copy of the plots displayed by PLOTCON vary depending on the hardware of the computer system and the printer. On some systems, users may run GRAPHICS.COM before starting the menu driver, and then hit the Shift and PrtScr keys when it is time to print. If this does not work for your system, try using a screen capture utility. A screen capture utility saves the screen to a file which can be printed or incorporated into a document. The utility should be activated before starting the menu driver. Note that some screen capture programs save the screen as it appears, e.g. white on a black background. Since printing a black background would use a significant amount of ink, you may want to invert the image to black on white. 5. References Mills, M.T., R. Paine, E. Insley, B. Egan. 1987. The complex terrain dispersion model Terrain preprocessor system --user's guide and program description. EPA/600/8-88/003, U.S. Environmental Protection Agency, Research Triangle Park, NC. Perry, S.G., D. Schwede, L. Adams, R. Paine, M. Dennis, M. Mills, D. Strimaitis, R. Yamartino, E. Insley. 1989. User's Guide to the Complex Terrain Dispersion Plus Algorithms for Unstable Situations (CTDMPLUS): Volume 1. EPA/600/8-89/041, U.S. Environmental Protection Agency, Research Triangle Park, NC.