The standard hydrologic grid (SHG) is a grid system developed by USACE Hydrologic Engineering Center (HEC) to cover the conterminous United States for hydrologic modeling purpose in HEC-HMS. SHG is defined on the basis of Albers equal-area conic map projection on NDA83 datum (EPSG 5070 or ESRI 102039) – see Figure 1. It is worth emphasizing that SHG is NOT a new kind of map projection, and instead it is a grid system established on EPSG 5070 and thus it uses EPSG 5070’s projection parameters (Table 1). Other sources may refer SHG to EPSG 5071 or EPSG 5072, which is also correct since EPSG 5071 or EPSG 5072 uses a refined version of NAD83 datum such as NAD83 (HARN) or NAD83(NSRS2007), and the differences between them and NAD83 are negligible for H&H applications.
The SHG grid cells are squares and their sizes vary. In SHG, the default grid cell size is 2000-meter x 2000-meter (SHG two-kilometer) and other supported grid cell sizes include 10,000-meter x 10,000-meter, 1000-meter x 1000-meter, 500-meter x 500-meter, 200-meter x 200-meter, 100-meter x 100-meter, 50-meter x 50-meter, 20-meter x 20-meter, and 10-meter x 10-meter. For easy discussion purpose, from now on all the discussions in this post assume the SHG grid cell size is 2000-meter x 2000-meter.
SHG grid cells are arranged relative to the origin point (intersection of 960W and 230N where) and indexed by cell counts (i, j) of the southwest (lower left) corners as shown in Figure 2. For Grid Cell (-5,4) in Figure 2, its southwest corner easting/northing coordinates are -10000m, 8000m (the easting is negative because this cell is on the west side of 960W).
A point’s location is defined in EPSG 5070 by its easting/northing (or X, Y) coordinates with unit of meter. In SHG, a point’s location is described by which cell it is located in. To find the indices of the cell to which a point belongs, divide the point’s easting and northing by cell size (2000M for default SHG) and the indices are the quotients (See equations in Figure 3). An example: if a point X, Y coordinates are -9000m, 9000m in EPSG 5070, its index of i=floor(-9000/2000)=floor(-4.5)=-5 and its index of j=floor(9000/2000)=floor(4.5)=4. The results means this point is located within Grid Cell (-5,4) as shown in Figure 2.
Gridded data (either static gridded data or gridsets) are normally imported and stored in a HEC-DSS file in the format of SHG, which means the gridded data is projected to EPSG 5070 with grid cell sizes defined by SHG, either 2000-meter x 2000 meter by default, or other sizes. Refer to this post for more details about gridded data and HEC-DSS.
Two types of hydrology modeling methods are available in HEC-HMS: lumped methods and spatially variant distributed methods. For lumped hydrology methods, average values of a watershed are inputs and the entire watershed is to be modeled as one single element (Example: Clark Unit Hydrograph transform method, TOC and Storage Coefficient R, as well as watershed time-area relationship are provided). For distributed hydrology methods, a watershed is subdivided into a gridded system and each grid cell is a small sub-watershed where hydrology calculation happens (ModClark transform method, each cell has its own travel time).
ModClark requires subbasins in HEC-HMS to be discretized using “Structured Discretization” which usually employs SHG grid system as shown in Figure 4 and Figure 5.
After setting up Projection (SHG is the default option) & Cell size for each subbasin as shown in Figure 5, a terrain file needs to be linked with the basin (Figure 6) and preprocessed by tools in the menu of GIS including Step 1: Preprocess Sinks and Step 2: Preprocess Drainage (Figure 7) – if these two steps have not been performed before.
Finally, go to GIS—>Compute—>Grid Cells (Figure 8) to generate the grid systems.
In this example, the created grid system is saved in PistolCreek.sqlite with cell sizes of 50-meter x 50-meter. PistolCreek.sqlite includes data such as cell index i & J, cell areas, and cell travel distances which are to be used for travel time calculation for each cell in ModClark method (Figure 9).