This page is not very easy to read. It is a basic "how to" for someone who already has good computer skills and also knows their way around WRF. Partially, it is just a place to store the procedure so I can do it again for another location. I am willing to share this hard earned information with others, but am not inclined to give detailed tutoring when I have so much more I have to get done. Good luck. I hope it helps you.
Restrictions and Decisions Concerning the Data The terrain height data files, like the other terrain data fed into geogrid, contains index values within the name of the file. It has starting and ending north-south indexes and starting and ending east-west indexes in its name. The index values must be 5 digits, so no set of data can have more than 99999 values in either direction. That is not enough to cover the entire USA, so I have divided the 10 meter terrain height data set into multiple parts, one part for each state. Each degree of longitude or latitude is about 10800 values. Only a few degrees are needed, in each direction for each state or some similar area. Each state's files are in a different directory which allows them to have files with the same names, but the starting longitude/latitude for index 1 is different for each state's directory. So each directory has files of the same name. The offset is controlled by a file in each directory named "index" which also has other important information about how to interpret the data files. Index values start with one, not zero. Data files can not overlap each other. Files can be missing within a set but each one must be complete. The size of the files is specified in the "index" file. Since unix allows links, you may want to consider creating a directory with each data file having an understandable name, like its longitude and latitude. Then use links with the required naming in a separate directory to build a set for you area of interest. That way you can cross borders or have multiple areas without duplicating the large files. A convenient size for this resolution is one arc degree in each direction. That size (232 MB) is almost the maximum size (250 MB) that is allowed to be downloaded in a contiguous file. I am using whole number degree values for the left and bottom of each file. It goes up to but does not include the next whole degree value. For Idaho I start with -118 as the longitude because someday I may want to create a detailed nest that goes farther west than the state border at -117. Data in each state's directory is independent of other directories, so directories can overlap. Details about the file and the "index" file can be found in the WRF users guide, Spring 2009 edition. It was at about page 3-34 with a subtitle heading of "Writing Static Data to the Geogrid Binary Format". Where to Find the Data and How to Get It The USGS site from which to get the data is: http://seamless.usgs.gov/website/seamless/viewer.htm On the left hand side, use the "Define Download Area by Coordinates" button, and on the resulting pop up page, use the "Switch to Decimal Degrees" option. Enter the coordinates you want to download. An example that worked for some Wyoming data is 41.00 to 41.99998 latitude and -107.00 to -106.0001 longitude. The amount less than the next whole degree seems to be different for longitude and latitude. I don't know why. It takes experimenting to get the exact 10800 values in both the rows and the columns. "Add the area" to get the next page: "Request Summary Page" which will default to the 1 arc second data. Choose "Modify Data Request" to get a page with very many data set options. About 90% down the list will be the 1 arc second option that will be checked. Also check the one below it, which is the 1/3 arc second. After the page updates, change the options for 1/3 arc second to BIL format, TGZ compression, and HTML. Then uncheck the 1 arc second option. After the page updates again, go to the bottom and "Save Changes and Return to Summary." That page will now show your chosen longitude latitude pairs. If you chose 1 X 1 degree block, expect 232 MB. Download it. Uncompress it. Check the file with .hdr extension; it can be read with gvim. Make sure it says 10800 X 10800 rows and columns. Then rename the file to fit with whatever was specified in the "index" file. Using topo_id starting at -118.00 and 42.00, the .bil file for -116.00 to -115.0001 and 42.00 to 42.99998 gets renamed to 21601-32400.00001-10800. Below is a "index" file for the Idaho example. I name the directories for each state similar to the other data that comes with WRF. For example, Idaho is topo_id. Below the index file is a copy of the changes I made to the GEOGRID.TBL so that it uses the topo_wy for the finest detail rather than the default topo_30s. It also seemed appropriate to change the way it interpolates since the 10 meter data will be smaller than the WRF grid cells. I removed the smoothing option because I specifically want all the detail rather than have it smoothed away. "index" File for 10 Meter Terrain Height Data. type = continuoussigned = yes projection = regular_ll dx = 0.00009259259259259259259 dy = 0.00009259259259259259259 known_x = 1.0 known_y = 1.0 known_lat = 42.000 known_lon = -118.000 wordsize = 2 endian = little tile_x = 10800 tile_y = 10800 row_order = top_bottom tile_z = 1 units="meters MSL" description="Topography height" GEOGRID.TBL File for Using 10 Meter Terrain Height Data This was done in version 3.0, so you may have to handmake these changes in the new 3.1 version which is a significantly different GEOGRID.TBL.# See options.txt for a (somewhat up to date) list of the # options that may be specified here. =============================== name = HGT_M priority = 1 dest_type = continuous df_dx=SLPX df_dy=SLPY # smooth_option = smth-desmth_special; smooth_passes=1 fill_missing=0. interp_option = wy:average_gcell(4.0)+four_pt+average_4pt # interp_option = wy:nearest_neighbor interp_option = 30s:average_gcell(4.0)+four_pt+average_4pt # interp_option = 30s:nearest_neighbor interp_option = 2m:four_pt interp_option = 5m:four_pt interp_option = 10m:four_pt interp_option = default:four_pt rel_path= wy:topo_wy/ rel_path= 30s:topo_30s/ rel_path= 2m:topo_2m/ rel_path= 5m:topo_5m/ rel_path= 10m:topo_10m/ rel_path= default:topo_30s/ =============================== name = HGT_U output_stagger = U priority = 1 dest_type = continuous interp_option = wy:average_gcell(4.0)+four_pt+average_4pt interp_option = 30s:average_gcell(4.0)+four_pt+average_4pt interp_option = 2m:four_pt interp_option = 5m:four_pt interp_option = 10m:four_pt interp_option = default:four_pt # smooth_option = smth-desmth_special; smooth_passes=1 fill_missing=0. rel_path= wy:topo_wy/ rel_path= 30s:topo_30s/ rel_path= 2m:topo_2m/ rel_path= 5m:topo_5m/ rel_path= 10m:topo_10m/ rel_path= default:topo_30s/ =============================== name = HGT_V output_stagger = V priority = 1 dest_type = continuous interp_option = wy:average_gcell(4.0)+four_pt+average_4pt interp_option = 30s:average_gcell(4.0)+four_pt+average_4pt interp_option = 2m:four_pt interp_option = 5m:four_pt interp_option = 10m:four_pt interp_option = default:four_pt # smooth_option = smth-desmth_special; smooth_passes=1 fill_missing=0. rel_path= wy:topo_wy/ rel_path= 30s:topo_30s/ rel_path= 2m:topo_2m/ rel_path= 5m:topo_5m/ rel_path= 10m:topo_10m/ rel_path= default:topo_30s/ =============================== name=LANDUSEF priority=1 dest_type=categorical z_dim_name=land_cat landmask_water=16 # Calculate a landmask from this field dominant=LU_INDEX interp_option = 30s:nearest_neighbor interp_option = 2m:four_pt interp_option = 5m:four_pt interp_option = 10m:four_pt interp_option = default:four_pt rel_path= 30s:landuse_30s/ rel_path= 2m:landuse_2m/ rel_path= 5m:landuse_5m/ rel_path= 10m:landuse_10m/ rel_path= default:landuse_30s/ =============================== name=SOILTEMP priority=1 dest_type=continuous interp_option=default:sixteen_pt+four_pt+average_4pt+average_16pt+search masked=water fill_missing=0. rel_path=default:soiltemp_1deg/ =============================== name=SOILCTOP priority=1 dest_type=categorical z_dim_name=soil_cat interp_option = 30s:nearest_neighbor interp_option = 2m:four_pt interp_option = 5m:four_pt interp_option = 10m:four_pt interp_option = default:four_pt rel_path= 30s:soiltype_top_30s/ rel_path= 2m:soiltype_top_2m/ rel_path= 5m:soiltype_top_5m/ rel_path= 10m:soiltype_top_10m/ rel_path= default:soiltype_top_30s/ =============================== name=SOILCBOT priority=1 dest_type=categorical z_dim_name=soil_cat interp_option = 30s:nearest_neighbor interp_option = 2m:four_pt interp_option = 5m:four_pt interp_option = 10m:four_pt interp_option = default:four_pt rel_path= 30s:soiltype_bot_30s/ rel_path= 2m:soiltype_bot_2m/ rel_path= 5m:soiltype_bot_5m/ rel_path= 10m:soiltype_bot_10m/ rel_path= default:soiltype_bot_30s/ =============================== name=ALBEDO12M priority=1 dest_type=continuous z_dim_name=month masked = water fill_missing = 8. interp_option=default:four_pt+average_4pt+average_16pt+search rel_path=default:albedo_ncep/ =============================== name=GREENFRAC priority=1 dest_type=continuous interp_option=default:four_pt+average_4pt+average_16pt+search z_dim_name=month masked = water fill_missing = 0. rel_path=default:greenfrac/ =============================== name=SNOALB priority=1 dest_type=continuous interp_option=default:four_pt+average_4pt+average_16pt+search masked = water fill_missing = 0. rel_path=default:maxsnowalb/ =============================== name=SLOPECAT priority=1 dominant_only=SLOPECAT dest_type=categorical z_dim_name=slope_cat masked = water fill_missing = 0. interp_option=default:nearest_neighbor+average_16pt+search rel_path=default:islope/ =============================== |