3.1 Tutorial Meteorological Data




Most of the CAPTEX example calculations in the following tutorial sections use one or more of the five meteorological data files provided in the Tutorial\captex directory on the CD. Additional meteorological files are provided for the Customized Concentration Simulations section at the end of this Tutorial in their own directories (Japan, Smoke, Dust, Sage). All meteorological files contain the data (u,v component wind speeds, temperature, and several other variables) on a regular grid at multiple levels and time periods. These files are already in a format that can be read by HYSPLIT directly. Converting data in other formats for use by HYSPLIT will be discussed in the next section. There are several programs available through the GUI that can be used to examine the contents of the meteorological data files.

If this tutorial is being run from a CD, then all the data files are provided. If these files need to be downloaded because this tutorial is being run through the web, try to insure that the data transfer is binary. Explicit external links are provided for some of the larger meteorological data files in section 2 below.

  1. To start, select the Meteorology / Display Data / Grid Domain menu tab to open the grid domain menu and set the file name to each of the meteorological files to see the spatial domain covered:

  2. Note that the menu settings default to show latitude-longitude lines at 5-degree intervals and only every other data grid point. The NARR file seems tilted because the native data grid is on a Lambert Conformal projection where the Northeast U.S. extract shown is toward the edge of the projection. The data on the two global files are on a regular latitude-longitude grid.

  3. To get more detailed information about the contents of any HYSPLIT meteorological data file, select the Meteorology / Display Data / Check File menu tab and select the data file in the check file menu which will then open a scroll box with more detailed information which is summarized below for each of the files. Do not try to open the link to the binary data files shown below with an editor.

    • captex2_narr.bin
    • Time period: 25th 15Z -> 28th 15Z
    • Number of grid points: 50 x 52
    • 20 levels: 1000 to 400 hPa
    • Resolution: 32 km, 3 hr (180 min)
    • External Link: captex2_narr.bin

    • captex2_wrf03.bin
    • Time period: 25th 15Z -> 26th 12Z
    • Number of grid points: 210 x 210
    • 28 levels: 0.998 to 0.0225 sigma
    • Resolution: 3 km, 15 min
    • External Link: captex2_wrf03.bin

    • captex2_wrf09.bin
    • Time period: 25th 12Z -> 28th 06Z
    • Number of grid points: 158 x 131
    • 25 levels: 0.998 to 0.310 sigma
    • Resolution: 9 km, 1 hr
    • External Link: captex2_wrf09.bin

    • captex2_wrf27.bin
    • Time period: 25th 12Z -> 28th 06Z
    • Number of grid points: 48 x 43
    • 25 levels: 0.998 to 0.310 sigma
    • Resolution: 27 km, 1 hr
    • External Link: captex2_wrf27.bin

    • captex2_era40.bin
    • Time period: 25th 12Z -> 28th 12Z
    • Number of grid points: 100 x 100
    • 14 levels: 1000 to 100 hPa
    • Resolution: 100 km (interpolated from 2.5 deg), 6 hr (360 min)
    • External Link: captex2_era40.bin

    • captex2_gblr.bin
    • Time period: 25th 12Z -> 29th 00Z
    • Number of grid points: 39 x 25
    • 15 levels: 1000 to 50 hPa
    • Resolution: 2.5 deg, 6 hr (360 min)
    • External Link: captex2_gblr.bin

    • RP198309.gbl
    • Time period: 1st 00Z -> 30th 18Z
    • Number of grid points: 39 x 25 (regional extract from the global analysis)
    • 15 levels: 1000 -> 50 hPa
    • Resolution: 2.5 deg, 6 hr (360 min)
    • External Link: RP198309.gbl

  4. A meteorological data file is composed of one or more time periods. Each time period begins with one or more ASCII index records that summarize the valid time, the grid definition, the variables, and level information. Each subsequent record contains one horizontal data field, consisting of 50 ASCII bytes of time, variable, and level information for that record, followed by X times Y bytes of data, where X and Y are the number of data points in the W-E and S-N directions, respectively. Floating point or integer data are packed as one byte per variable. Precision is maintained by packing the differences between adjacent grid points rather than packing the absolute values.

In summary, the WRF-ARW simulation (boundary conditions derived from NARR) provides the finest spatial and temporal resolution. The three global reanalysis files are just extracts for North America, two covering just the period of the experiment, while the third one covers the entire month. The NARR or WRF-ARL data will be used for most of the examples.