Creating SCS Curve Number Grid using HEC-GeoHMS

Prepared by
Venkatesh Merwade
School of Civil Engineering, Purdue University
vmerwade@purdue.edu

October 2010

Introduction
SCS curve number grid is used by many hydrologic models to extract the curve number for
watersheds. The objective of this tutorial is to use soil and land use data to create a curve number
grid using HEC-GeoHMS (ArcGIS 9.2 version).

Computer Requirements
You must have a computer with latest windows operating system, and the following programs
installed:
1. ArcGIS 9.3 (with ArcInfo)
2. Arc Hydro tools (latest version)
3. Hec-GeoHMS for ArcGIS 9.3 (latest version)

Make sure the Arc Hydro tools (version that works with 9.2) are installed on the system. There
are lots of Arc Hydro tools available on the internet. The version that is used in this tutorial
including installation instructions is available at the following link.
ftp://ftp.ecn.purdue.edu/vmerwade/download/data/tools.zip

The HEC-GeoHMS set-up file for this tutorial is also included in this above tools.zip file. Just
unzip the file, and run the setup to install both ArcHydro and HEC-GeoHMS. You will need to
have administrative access to install Arc Hydro and HEC-GeoHMS.

Data Requirements and Description

This tutorial requires the following datasets:
(1) DEM for the study area
(2) SSURGO soil data
(3) 2001 land cover grid from USGS

The DEM and land cover grid from USGS and SSURGO data for Cedar Creek in northeast
Indiana are provided as zip file at the end of this paragraph. The SSURGO soil data that will be
used is provided in a geodatabase (cedar_ssurgo.mdb) that is created in the “Downloading
SSURGO Soil Data” tutorial available at:
(http://web.ics.purdue.edu/~vmerwade/education/download_ssurgo.pdf). You can look at the
SSURGO download tutorial if you want to learn how this dataset is processed. Also, if you are

1
interested in downloading the DEM and land cover data by yourself, instructions are available in
the following tutorial: http://web.ics.purdue.edu/~vmerwade/education/ned_nhd.pdf

It is highly recommended that you go through these exercises of downloading the data to make
yourself aware of the procedure involved.

(Notes:

1. The DEM, land use and soil data used in this tutorial are already clipped to the Cedar Creek
study watershed
2. This tutorial uses the SSURGO soil data, which is the highest resolution soil data available in
public domain from NRCS. This can be replaced by STATSGO data as long as you know
how to interpret STATSTO to follow the steps provided in this tutorial)

Download the data from ftp://ftp.ecn.purdue.edu/vmerwade/download/data/cngrid.zip

Unzip cngrid.zip in your working directory. The ArcCatalog-view of the data folder (or
whatever name you gave to your working folder) is shown below:

cedar_ssurgo is the geodatabase with SSURGO spatial and tabular data for cedar creek area.
cedar_dem is the raw 30 DEM for Cedar Creek obtained from USGS and clipped for the study
watershed, and cedar_lu is the 2001 land cover grid from USGS. All datasets have a common
spatial reference (NAD_1983_UTM_16).

Note: It is very critical to assign and use consistent coordinate system for all the datasets.

Getting Started
Open ArcMap. Create a new empty map, and save it as cngrid.mxd (or any other name). Right
click on the menu bar to pop up the context menu showing available tools as shown below.

2
Check the Arc Hydro Tools 9 menu. If the Arc Hydro Tools menu does not appear in the list,
click on Customize (Scroll down the list to see “Customize”). In the Customize dialog that
appears, check the Arc Hydro Tools box.

You should now see the Arc Hydro tools added to ArcMap as shown below. You can leave it
floating or you may dock it in ArcMap menubars.

Similarly add Spatial Analyst extension and activate it by clicking on ToolsExtensions…, and
checking the box next to Spatial Analyst.

Dataset Setup
All raster data created with the Arc Hydro are stored in a subdirectory with the same name as the
dataset or Data Frame in the ArcMap document (called Layers by default and under the directory
where the project is stored). The location of the vector, raster, and time series data can be
explicitly specified using the function ApUtilitiesSet Target Locations.

3
You can leave the default settings if they are pointing to the same directory where the ArcMap
document is saved. Now load data in the map document. Click on the Add icon to add the
raster data. In the dialog box, navigate to the location of the data; select cedar_dem raster file
containing the DEM for Cedar creek and click on the Add button. Save the map document.

Fill Sinks
(Note: If you have results from the terrain processing tutorial, you can use the Fil DEM from the
exercise directly skipping this step. Go to next section on preparing landuse data)

The first step in creating a curve number grid is to “hydrologically correct” the raw DEM by
filling sinks. The Fill Sinks function in Arc Hydro fills the sinks in a grid. If cells with higher
elevation surround a cell, the water is trapped in that cell and cannot flow. The Fill Sinks
function modifies the elevation value to eliminate these problems.

On the ArcHydro Toolbar, select Terrain PreprocessingData ManipulationFill Sinks.

Confirm that the input for DEM is cedar_DEM. The output is the Hydro DEM layer, named by
default Fil. This default name can be overwritten. Leave the other options unchanged.

4
Press OK. Upon successful completion of the process, the HydroDEM (in older ArcHydro, this
is called Fil) layer is added to the map. This process takes a few minutes. Save the map
document.

Preparing land use data for CN Grid

Add cedar_lu grid to the map document. You will see the grid is added with a unique symbology
assigned to cells having identical numbers as shown below:

5
These numbers represent a land use class defined according to the USGS land cover institute
(LCI). A description of some of the land classes and their associated numbers in the grid is
shown below by reproducing LCI webpage (http://landcover.usgs.gov/classes.php). You can visit
LCI website (publications link) to learn more about how the land use grid is created.

Eventually, we are going to use these land use classes and soil group type, in conjunction with
SCS curve numbers, to create the curve number grid. The SCS CN table gives CN for different
combinations of land use and soil group. The cedar_lu grid has 15 different categories which you
can leave unchanged, or reclassify the grid to reduce the number of land use classes to make the
task easier. If you open the attribute table of cedar_lu, you will see that majority of cells
represent grass/crops, followed by forest, developed land, and then water. We will reclassify
cedar_lu to represent these four major classes. The following table shows how we will
accomplish the reclassification of cedar_lu (you are free to have more or less classes).

6
 Original NLCD classification Revised classification (re-
classification)
Number Description Number Description
11 Open water 1 Water
90 Woody wetlands
95 Emergent herbaceous
wetlands
21 Developed, open space 2 Medium Residential
22 Developed, low intensity
23 Developed, medium intensity
24 Developed, high intensity
41 Deciduous forest 3 Forest
42 Evergreen forest
43 Mixed forest
31 Barren land 4 Agricultural
52 Shrub/scub
71 Grassland/herbaceous
81 Pasture/hay
82 Cultivated crops

To implement the above re-classification, use the Spatial Analyst Toolbar in ArcMap. If Spatial
Analyst is not visible in the menu bar, activate it in the same way you activated the Arc Hydro
toolbar. Click on Spatial AnalystReclassify….

In the reclassification window, confirm the Input raster is cedar_lu, Reclass field is Value, and
then manually assign the new numbers based on the above table as shown below (leave NoData
unchanged).

7
Save the output raster as lu_reclass in your working folder, and click OK. A new grid named
lu_reclass will be added to the map as shown below (you may not get the same colors in
symbology which is OK)

The final step in processing land use data is converting the reclassified land use grid into a
polygon feature class which will be merged with soil data later. Click on Spatial
AnalystConvertRaster to Features.. Confirm the Input raster is lu_reclass, the Field is
Value, output geometry type is Polygon, and save the Output features as landuse_poly.shp in
your working directory (this output is saved only as a shapefile without any other options). Click
OK.

8
You can symbolize the new landuse_poly.shp to match with lu_reclass grid or leave it
unchanged. Also, you can export landuse_poly.shp to cedar_ssurgo.mdb to keep all the data in a
single geodatabase. Save the map document.

The processing of land use data for preparing curve number grid is over. Now let us prepare soil
data.

Preparing Soil data for CN Grid

Add cedar_soil_clip feature class from spatial feature dataset within cedar_ssurgo.mdb. For
extracting CN numbers, we need soil group for each polygon in cedar_soil_clip feature class. If
you open the attribute table for cedar_soil_clip, you will notice that there is no field for storing
soil group (you can refer to SSURGO tutorial to understand the meaning of each field in
cedar_soil_clip). So the first step is creating an empty field for storing soil group data. Create a
field named “SoilCode” (type: Text) in cedar_soil_clip.

The soil group data are available in the component table (in hydgrp field) so add the component
table from cedar_ssurgo.mdb to the map document. The polygon features in cedar_soil_clip are
related to component table through mukey field. Right click on cedar_soil_clip in the ArcMap
table of contents, and click on Join and RelatesJoin …. Join component table to
cedar_soil_clip by using the common mukey field as shown below:

9
Click OK. You may get a message asking you to index fields. You can respond to this message
by selecting either yes or no - it does not matter in this exercise because we will join the table
only once. After you create the join, open the attribute table for cedar_soil_clip, and you will see
that the fields from component table are now available in cedar_soil_clip feature class. Now
populate the SoilCode (or cedar_soil_clip.SoilCode) field in cedar_soil_clip by equating it with
component.hydgrp field. With the attribute table open, right click on SoilCode field to open the
field calculator and then equate SoilCode to component.hydgrp as shown below:

Click OK. If there are rows in component with “Null” values (which is the case for this dataset),
you may get an error message saying the values are too large for the field. Just ignore this
message and continue. After the calculations are complete, you should see
cedar_soil_clip.SoilCode populated with letters A/B/C/D. Now remove the join and save the
map document.

Before we proceed, let us deal with <Null> rows for SoilCode field. Review the SoilCode field,
and you will see that most polygons have “C” soil group so let us assign a SoilCode of C for all
polygons that do not have soil group associated with them (this is just one way of dealing with
the issue for small number of Null rows. If the number of Null rows is significantly high, you
may want to consult other resources before assigning SoilCode to these rows). Select the rows in
cedar_soil_clip that have <Null> values for SoilCode and assign them a value of “C”. You can
use select by attributes option to do this. Now you should have a SoilCode (soil group) assigned
to each polygon in cedar_soil_clip.

10
Next create four more fields named PctA, PctB, PctC, and PctD all of type short integer in
cedar_soil_clip feature class. For each feature (polygon) in cedar_soil_clip PctA will define what
percentage of area within the polygon has soil group A, PctB will define what percentage of area
within the polygon will have soil group B and so on. This is critical when we have polygons with
more than one soil group (for eg. A-B-A/D would mean that group A, group B and group A/D
soils are found in one polygon; A/D would mean the soil behaves as A when drained and as D
when not drained, and so on). If we have classifications such as these, we need to define how
much area of a polygon is A/B/C/D. For Cedar Creek area we have only one soil group assigned
to each polygon so a polygon with soil group “A” will have PctA = 100, PctB = 0, PctC = 0, and
PctD = 0. Similarly for a polygon with soil group D, only PctD = 100, and other three Pcts are 0.
Now populate PctA, PctB, PctC and PctD based on SoilCode for each polygon. You can select
features based on SoilCode and then use field calculator to assign numbers to polygons. The
resulting attribute table should look like below:

The preparation of soil data is over at this point. The next step is to merge/union both soil data
and land use data to create polygons that have both soil and land use information. Save the map
document.

11
Merging of Soil and Landuse Data

To merge/union soil and landuse data, use the Union tool in ArcToolbox available under Analysis
ToolsOverlay. Browse/drag cedar_soil_clip and landuse_poly as input features, name the
output feature class as “cedar_soil_lu” in the same geodatabase (cedar_ssurgo.mdb), leave the
default options, and click OK (you can change the cluster tolerance to a small number, but this is
not necessary).

This process will take few minutes, and the resulting cedar_soil_lu feature class will be added to
the map document. Save the map document.

The result of union/merge features inherit attributes from both feature classes that are used as
input. However, if the outer boundaries of input feature classes do not match exactly, the
resulting merged feature class (cedar_soil_lu in this case) usually will have features that will
have attributes from only one feature class because the other feature do not exist in this area.
These features are usually referred to as “slivers”. If you open the attribute table for
cedar_soil_lu, you will find that there are several sliver polygons in this feature class that have

12
attributes only from landuse_poly and the soil attributes are empty, and vice versa as shown
below:

In the above table the columns that start with FID_.... give the object ids of features from
landuse_poly and cedar_soil_clip. A value of -1 for FID_... means one of the feature classes do
have features in that area to union with features from other feature classes. Basically a value of -
1 for FID_... means that feature is a sliver polygon. You can also verify this by looking at other
fields. For example features that have FID_cedar_soil_clip = -1 have attributes only from
landuse_poly and all attributes from cedar_soil_clip = 0.

One way to deal with sliver polygons is to assign missing values to all features. Another way
(easiest!) is to just delete them. For this exercise we will take the easy route, but you may want to
populate these features for other studies depending on your project needs.

Start the Editor. Select all the features that have “FID_...” = -1 and delete them. Save your
edits, stop the Editor, and save the map document.

This finishes the processing of spatial data for creating the curve number grid. The next step is to
prepare a look-up table that will have curve numbers for different combinations of land uses and soil
groups. In this case, we will use SCS curve numbers that are available from literature (SCS reports, or
SCS tables from text books). The spatial features in conjunction with the look-up table can then be used
to create curve number grid.

Creating CN Look-up table

Create a table named “CNLookUp” inside cedar_ssurgo.mdb. You can either do this in ArcMap
or in ArcCatalog. The quick and easy way is to use the command line in ArcMap. Once the table
is created create the following fields in it:

1. LUValue (type: short integer)

2. Description (type: text)
3. A (type: short integer)
4. B (type: short integer)
5. C (type: short integer)
6. D (type: short integer)

13
Now start the Editor to edit the newly created CNLookUp table, and populate it as shown
below.

Columns A/B/C/D store curve numbers for corresponding soil groups for each land use category
(LUValue). These numbers are obtained from SCS TR55 (1986). Save the edits and stop the
Editor. Save the map document.

Creating CN Grid

We will use HEC-GeoHMS to create the curve number grid. Activate the HEC-GeoHMS Project
View toolbar in the same way as ArcHydro toolbar. HEC-GeoHMS uses the merged feature
class (cedar_soil_lu) and the lookup table (CNLookUp) to create the curve number grid. The
format and the field names that we are used in creating the CNLookUp table are consistent with
HEC-GeoHMS. Before we proceed, one final step is to create a field in the merged feature class
(cedar_soil_lu) named “LandUse” that will have land use category information to link it to
CNLookUp table. We already have this information stored in GRIDCODE field, but HEC-
GeoHMS looks for this information in LandUse field. So create a field named LandUse (type:
short integer), and populate it by equating it to GRIDCODE.

On the HEC-GeoHMS Project View toolbar, click on UtilityCreate Parameter Grids.. Choose
the lookup parameter as Curve Number (which is default) in the next window, Click OK, and
then select the inputs for the next window as shown below:

14
HydroDEM for Hydro DEM, cedar_soil_lu (merged soil and land use) for Curve Number
Polygon, CNLookUp table for Curve Number Lookup, and leave the default CNGrid name for
the Curve Number Grid.

This process takes a while. Be patient and CNGrid will be added to your map document. You
can change the symbology of the grid to make it look pretty!

You now have a very useful dataset for use in several hydrologic models and studies. Save the
map document, and exit ArcMap.

OK. You are done!

15