Tutorial on using HEC-GeoHMS to develop Soil Moisture Accounting Method Inputs for

HEC-HMS

Prepared by

Jessica Holberg

Lyles School of Civil Engineering, Purdue University

[email protected]

June 2014

______________________________________________________________________________

Introduction

The objective of this tutorial is to create the grids necessary to complete HEC-HMS project setup
for the Soil Moisture Accounting (SMA) loss method using HEC-GeoHMS tools in ArcGIS. It is
expected that you are reasonably adept with ArcGIS and HEC-HMS. This tutorial was designed
to supplement the "HMS Model Development using HEC-GeoHMS (ArcGIS 10)" tutorial
provided on Dr. Venkatesh Merwade's research page and available at the following address:
https://web.ics.purdue.edu/~vmerwade/education/geohms.pdf.

The aforementioned tutorial uses SCS for the Loss Method, but this tutorial will take you
through the steps necessary to use SMA for the Loss Method. To continue with this tutorial, you
will need to have performed every step listed in Dr. Merwade's tutorial up to the "HMS
Inputs/Parameters" section that begins on page 15 of the tutorial.

Computer Requirements

You must have a computer with windows operating system, and the following programs
installed:

1. ArcGIS 10
2. HEC-GeoHMS version 10
3. Microsoft Excel

You can download HEC-GeoHMS for free from the US Army Corps of Engineers Hydrologic
Engineering Center website http://www.hec.usace.army.mil/software/.

Data Requirement

The datasets required beyond those listed in the aforementioned tutorial are:

(1) SSURGO soil data from the Web Soil Survey*

1
(2) 2006 land cover grid from USGS
(3) 2006 impervious surface percentage grid from USGS
(4) USGS Streamflow data
(5) Evapotranspiration data from NOAA (http://www.ncdc.noaa.gov/IPS/cd/cd.html)

*Before beginning this tutorial, you will need to generate a complete, linked SSURGO
geodatabase, as described in the tutorial titled “Downloading SSURGO Soil Data from Internet.”
This tutorial can be found at: http://web.ics.purdue.edu/~vmerwade/education/ssurgo.pdf.

**Important: Keep track of the units for the datasets you are using. You will need to make sure
that all variables are in the correct units once you create the HEC-HMS project file. For your
convenience, there is a table in the appendix listing the HEC-HMS units for the SMA
parameters. It is often easier to complete this tutorial with whatever units the raw data uses, and
then copy out the variables from the HEC-HMS file and convert them to the appropriate units,
before pasting them back into HEC-HMS.

Getting Started
Add your SSURGO geodatabase, land cover grid, and impervious surface percentage grid to
your existing map document.

Note that this tutorial is designed as a companion to the HMS Model Development tutorial and
you will need to reference it for certain elements.

Select HMS Processes

You can specify the methods that HMS should use for transform, routing, and loss using this
function. These designations are not final, but can be changed in HMS.

Select Parameters  Select HMS Processes. Confirm input feature classes for Subbasin and
River, and click OK. Choose SMA for Loss Method, Linear Reservoir for Baseflow Method, and
whichever methods you intend to use for transformation and routing. Click OK.

Save the map document.

Follow the instructions for “River Auto Name” and “Basin Auto Name” as outlined in the HMS
Model Development Tutorial. Then, continue with this tutorial.

Subbasin Parameters
Depending on the method (HMS Processes) you intend to use for your HMS model, each sub-
basin must have parameters such as tension zone depth for SMA method. These parameters are
assigned using the Subbasin Parameters option. This function overlays subbasins over grids and
computes an average value for each basin. We will explore only the grids required for the SMA
method.

Select Parameters  Subbasin Parameters from Raster. You will get a window in which you
will have to select the rasters that you wish to use for extracting parameters. The rasters listed for
the SMA method are as follows:

2
1. Total Storm Precipitation Grid
2. 2-Year Rainfall Grid
3. Percentage Impervious Grid
4. Max Canopy Storage Grid
5. Max Surface Storage Grid
6. Max Soil Infiltration Grid
7. Max Soil Percolation Grid
8. Soil Tension Storage Gird
9. Max Soil Storage Grid
10. GW1 Max Storage Grid
11. GW2 Max Storage Grid
12. GW1 Max Percolation Grid
13. GW2 Max Percolation Grid

Rasters 1 and 2 are optional for the SMA method and will not be developed over the course of
this tutorial. We already have Raster 3, since we downloaded the impervious surface percentage
grid from USGS. Rasters 4-9 will be developed during this tutorial. Rasters 10-11 are constant
value rasters, and thus do not need to need to be created. We will simply assign the constant
values to each subbasin in our Subbasin attribute table. Raster 12 can be taken as equivalent to
Raster 7. Raster 13 will not be created, because it is an extremely conceptual parameter. You will
simply assign a GW2 Max Percolation Rate during HMS model calibration.

Parameters estimated from Land Cover Grid

Raster 4: Max Canopy Storage Grid

First, clip the land cover raster to the project boundaries. Open the attribute table of the land
cover raster. Note that the Value field contains National Land Cover Database (NLCD) Classes.
Create a new field of type Double and name it Canopy_int. Use the NLCD values (see Table 1A
for NLCD definitions) and the canopy interception values shown in Table 1 to assign canopy
interception values to each NLCD. If you are unsure of any values, just use your best judgement.

Table 1. Canopy Interception Values

Canopy Interception
Type of Vegetation
in. mm
General Vegetation 0.05 1.270
Grasses and Deciduous Trees 0.08 2.032
Trees and Coniferous Trees 0.1 2.540

Your land use attribute table should look something like this:

3
Unfortunately, ArcGIS is unable to convert one raster directly into another raster based on a
different field. So, in order to create the canopy interception raster, we must first create a point
feature and then convert it into a raster. To do this, open the ArcToolbox. Go to Conversion
Tools  From Raster  Raster to Point. Select your clipped land cover raster for Input Raster.
Select Canopy_int for Field. Save the point feature class to your working geodatabase. Click
OK. Depending on the size of your watershed, this may take a long time.

Once the point feature class has been created, save the map document.

Next, we will convert the point feature class into a raster. Go to ArcToolbox  Conversion
Tools  To Raster  Point to Raster. Select your point feature class for Input Features. Select
GRID_CODE for Value Field. Save the raster to your working geodatabase. Select MEAN for
Cell assignment type. Type 30 for the cellsize. Click OK.

4
Save the map document. The Max Canopy Storage grid is now complete!

Parameters estimated from SSURGO Data

Preprocessing
Of the tables in your SSURGO Database, chorizon and component are of interest to us.

Export the chorizon table. Right click on the chorizon table in the ArcGIS Catalog window.
Select Export  To dBASE (single)… Select your working geodatabase as the Output Location.
Name the Output Table Chorizon_export.dbf. Leave everything else as default. Click OK.

5
Open the Chorizon_export table in Excel. Save it as a .xlsx file. In the Chorizon_export table,
the only fields you will need are: chkey, cokey, ksat_r, hzdepb_r, wsatiated_r, and wthirdbar_r.
If you want, you can delete the remaining fields to make the spreadsheet a little bit cleaner and
easier to work with.
Table 2. Chorizon Field Definitions
Field Definition
chkey Horizon ID
cokey Component ID
ksat_r Representative saturated hydraulic conductivity
hzdepb_r Representative depth from soil surface to bottom of layer
wsatiated_r Representative soil porosity
wthirdbar_r Representative field capacity

The Chorizon table contains information about each layer (horizon) of soil within a soil
component. Each component, identified with a single cokey, contains multiple layers, each
identified by a chkey. So, each cokey is associated with multiple chkeys, as seen below.

cokey chkey NoHorizons

1115678 1234567 1
1115678 1234568 2
1115678 1234569 3
1115679 1234570 1
1115679 1234571 2

Sort the entire spreadsheet so that the chkey field is arranged from lowest to highest. Then,
create a new field titled “NoHorizons.” Using the cokey field, fill the NoHorizons field with a
running count of the number of layers in each component. See the table above for an example.

For each component, find the average ksat_r, wsatiated_r, and wthirdbar_r values. Also,
determine the ksat_r value of the topmost horizon for each component and the hzdepb_r value of
the bottommost horizon. These values can be determined easily enough by implementing the
NoHorizons field and a little bit of creativity with your Excel formulas. Copy the results of this
exercise to another sheet. Once this is complete, you should have a table that looks something
like this:

cokey ksat_avg ksat_Layer1 hzdepb_r wsat_avg wthirdbar_avg

9391673 14.89 21.88 152 57.33 44.73
9391674 15.83 23.29 203 67.00 33.33
… … … … … …

Save your spreadsheet. Delete the first sheet with all of your calculations. Save the spreadsheet
as a .cvs file. Title it chorizon_re.

6
If you closed your map file, re-open it and import chorizon_re. To do this, Right click on your
working geodatabase. Select Import  Table (single)… Browse to the location of your .cvs table
and select it for the Input Rows. Name the Output Table chorizon_re.

Open chorizon_re in ArcGIS. Create three new fields: (1) titled “comppct” of type Short Integer,
(2) titled “slope” of type Float, (3) titled “mukey” of type Text. Join the existing component
table to the chorizon_re table using the “cokey” field. Comppct stands for component percent
and displays the percent of that specific map unit that is occupied by that particular component.
See the figure below for a graphical description of the relationship between map units,
components, and horizons.

Map Unit (mukey)

Component
(cokey)

Horizons (4)
(chkey)

Top View Side View

Figure 1. SSURGO Organization

Using the field calculator, equate chorizon_re.comppct with component.comppct_r,

chorizon_re.slope with component.slope_r, and chorizon_re.mukey with component.mukey.
Once this is complete, remove the join.

Save the map document.

Re-export the chorizon_re table. This time, name it SSURGO_Export. As before, open the table
in Excel and save it as a .xlsx file. Convert wthirdbar_avg and comppct to decimal form (divide
by 100).

You will notice that there are multiple cokeys for each mukey, similar to the multiple chkeys for
each cokey that we saw previously. Similar to the horizons, we need to create a running count of
the number of components associated with each mukey. Create a new field titled “muIndex” for
this purpose. Now we are ready to perform the calculations necessary to create the various grids.

Surface Depression Storage

Calculate the weighted slope of each map unit. This can be done by multiplying the comppct
(in decimal form) by the slope of each component and then summing the values for each map
unit. Create a new column titled “SurfDepStor” for your surface depression storage values.
Select the appropriate surface storage value from Table 3 according to the weighted slope you
calculated previously. So, you should have only one surface storage value for each map unit.

7
 Table 3. Surface Depression Storage Values
Surface Storage
Description Slope (%)
in. mm
Paved Impervious Areas NA 0.125-0.25 3.18-6.35
Flat, Furrowed Land 0-5 2.00 50.8
Moderate to Gentle Slopes 5-30 0.25-0.50 6.35-12.70
Steep, Smooth Slopes >30 0.04 1.02
*taken from Fleming, 2002

Maximum Infiltration Rate

Calculate the weighted saturated conductivity of layer one (topmost horizon) for each map unit.
Similar to the surface depression storage, this can be done by multiplying the comppct (in
decimal form) by the ksat_Layer1 of each component and then summing the values for each
map unit. We will take this weighted form of layer one’s saturated conductivity as the maximum
infiltration rate. Title the column containing these values “MaxInfilRate.”

Maximum Soil Profile Storage

For each component, multiply the corresponding comppct (decimal form), wsatiated_avg
(decimal form), and hzdepb_r together. Sum these values for each map unit in a column titled
“MaxSoilStor.”

Maximum Tension Zone Storage

For each component, multiply the corresponding comppct (decimal form), wthirdbar_avg
(decimal form), and hzdepb_r together. Sum these values for each map unit in a column titled
“MaxTensZoneStor.”

Percolation Rate
For each component, multiply the corresponding comppct (decimal form) by the ksat_avg. Sum
these values for each map unit in a column titled “PercRate.”

Save your spreadsheet. Copy the mukey, SurfDepStor, MaxInfilRate, MaxSoilStor,

MaxTensZoneStor, and PercRate to a new spreadsheet. Save the spreadsheet as a .cvs file. Title
it SSURGOImport.

Import the SSURGOImport table into your working geodatabase. Open your SSURGO polygon
feature class you created while building your SSURGO geodatabase. Create five new fields of
type Float for SurfDepStor, MaxInfilRate, MaxSoilStor, MaxTensZoneStor, and PercRate. Join
the SSURGOImport table to your SSURGO polygon feature class using the mukey as the
common field.

You are now ready to begin creating Rasters 5, 6, 7, 8, 9, & 12!

Raster 5: Max Surface Storage Grid

We will convert the SSURGO polygon feature class into a raster. Go to ArcToolbox 
Conversion Tools  To Raster  Feature to Raster. Select your SSURGO polygon feature

8
class for Input Features. Select SurfDepStor for Field. Save the raster to your working
geodatabase. Type 30 for the cellsize. Click OK.

The Max Surface Storage Grid is now complete!

Raster 6: Max Soil Infiltration Grid

Repeat the steps for Raster 5 but use MaxInfilRate for the Field. Name the raster
MaxSoilInfiltration, or something similar.

Raster 7 & 12: Max Soil Percolation and GW1 Max Percolation Grids
Repeat the steps for Raster 5 but use PercRate for the Field. Name the raster .

Raster 8: Soil Tension Storage Grid

Repeat the steps for Raster 5 but use MaxTensZoneStor for the Field. Name the raster.

Raster 9: Max Soil Storage Grid

Repeat the steps for Raster 5 but use MaxSoilStor for the Field. Name the raster.

You are now finished using ArcGIS tools to create the necessary rasters for the SMA loss
method!

Assign Subbasin Parameters

Select Parameters  Subbasin Parameters from Raster. You will get a window in which you
will have to select the rasters you wish to use for extracting parameters. Select the appropriate
subbasin feature class for Input Subbasin. Select ImpSurface for Input Percentage Impervious
Grid. Select CanopyInt for Input Max Canopy Storage Grid. Select SurfDepStorage for Input
Max Surface Storage Grid. Select MaxSoilInfiltration for Input Max Soil Infiltration Grid. Select
PercolationRate for Input Max Soil Percolation Grid and Input GW1 Max Percolation Grid.
Select SoilTensStorage for Input Soil Tension Storage Grid. Select MaxSoilStorage for Input
Max Soil Storage Grid. Click OK.
9
This process will calculate average parameter values for each subbasin and copy the values into
the subbasin attribute table. If you wish, you can open the subbasin attribute table to see how the
values are stored. If any values were not properly transferred, simply run the process again.

Save your map document.

Parameters estimated from USGS Streamflow Data

For this part of the tutorial, you will need daily streamflow data for 3 to 4 storms occurring
during different months of the year. Look for storms that are fairly isolated; storms where the
streamflow hydrograph is allowed to return to normal for a couple days before runoff from the
next storm is visible.

Streams convey stored water from three different sources: stream channels, surface soil
(interflow), and groundwater. In this portion of the exercise, we will learn how to break up a
streamflow hydrograph into its various components and calculate the variables necessary for soil
moisture accounting in HEC-HMS.

Download the streamflow data and open it in Excel. Save the file. Create a hydrograph of the
streamflow data on a semi-logarithmic plot.

10
 100000
Streamflow
Flow (cfs)

10000

Receding Limb
1000
17-Feb 22-Feb 27-Feb 4-Mar 9-Mar

The tail-end of the receding limb represents the time when groundwater is the only source
contributing to streamflow, as both surface runoff and interflow have stopped. There should be
an inflection point visible in this area of the graph to help you identify the correct portion of the
hydrograph.

To begin, project a line backwards from the tail-end of the receding limb to the time of peak
flow, maintaining the slope of that tail-end portion. Connect the line to the point at which the
hydrograph begins to rise as a result of runoff. This line represents the groundwater contribution
to streamflow, or GW2. See the figure below.

100000
Streamflow
Groundwater
Flow (cfs)

10000

1000
17-Feb 22-Feb 27-Feb 4-Mar 9-Mar

Next, subtract the groundwater from the streamflow. Plot the result on the same graph. This line
represents the contribution to streamflow from surface runoff and interflow. See the figure
below.

11
 100000
Streamflow
Runoff + Interflow
Groundwater
Flow (cfs)

10000

1000
17-Feb 22-Feb 27-Feb 4-Mar 9-Mar

We are only interested in the portion of the runoff + interflow receding limb with the shallowest
slope. So, you can either just ignore the tail-end of it, or delete the last few points, whichever is
the easiest for you. Using the same method as we used to create the groundwater line, create an
interflow line, as seen in the figure below. The interflow line represents GW1.

100000
Streamflow
Runoff + Interflow
Groundwater
Interflow
Flow (cfs)

10000

1000
17-Feb 22-Feb 27-Feb 4-Mar 9-Mar

Now that we have the graphs determined, we are ready to begin the calculations. In SMA,
Groundwater 1 variables represent interflow, and Groundwater 2 represents groundwater, or
baseflow.

The SMA inputs we will be calculating are:

 Groundwater 1 Recession Coefficient
 Groundwater 1 Storage Depth
 Groundwater 2 Recession Coefficient
 Groundwater 2 Storage Depth

12
The recession curve, or receding limb of a hydrograph, can be described by Equation 1, below.

𝑞1 = 𝑞0 𝐾𝑟 = 𝑞0 ∗ exp⁡(−𝛼𝑡) (1)

Where 𝑞0 is the initial streamflow, 𝑞1 is the streamflow at a later time, t, 𝐾𝑟 is a recession

constant less than 1, and 𝛼 = −ln⁡𝐾𝑟 ⁡. The recommended time stop for streamflow regression
analysis is 1 day, but you can use a shorter time step for a smaller basin. Using the area of
shallowest slope of the streamflow hydrograph and Equation 1, calculate the Groundwater 2 𝛼-
value for each step. If the calculation results in any negative 𝛼-values, simply delete them.
Average the remaining 𝛼-values and calculate your Groundwater 2 Recession Coefficient using
Equation 2, below.

𝑅𝑒𝑐𝑒𝑠𝑠𝑖𝑜𝑛⁡𝐶𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡 = 1/𝛼 (2)

Using the same section of the streamflow hydrograph and Equation 3, calculate the Groundwater
2 Storage depth for each step. Average the values for your final Groundwater 2 Storage Depth.
𝑞𝑡
𝑆𝑡 = (3)
𝛼

Where 𝑆𝑡 is the storage in the basin at time, t. Repeat the same calculations using the Runoff +
Interflow graph to determine the Groundwater 1 Recession Coefficient and Groundwater 1
Storage Depth.

You have just finished calculating the GW 1&2 Recession Coefficients and Storage Depths for
one storm! Repeat the same process for the remaining storms.

Once complete, summarize your values in one spreadsheet. This will allow for an easy
comparison. Examine how the values change over different months or seasons. Look for any
patterns or drastic differences in values; this will be an indication that you may need to consider
creating two models, instead of just one. Depending on the climate of the area you are modeling,
you may want to create a bi-annual model. You can split the model into wet and dry seasons,
spring/summer and fall/winter, etc. Once you have determined how to split your model, if
necessary, average each set of values for the relevant months. For example, you will have one
GW1 Recession Coefficient for your dry model, one GW1 Storage Depth for your dry model,
etc.

The best way to create two models is to fully develop one model, and then simply copy it and
change the relevant parameters before calibrating the model. For most models, the only variables
that will be different are the GW 1&2 Recession Coefficients and Storage Depths. Additional
variables will most likely change after you perform an independent sensitivity analysis and
calibration of each model.

You are now finished calculating the GW 1&2 Recession Coefficients and Storage Depths!

13
Assign Subbasin Parameters

Open the subbasin attribute table. Start an edit session. Use the field calculator to assign your
GW 1&2 parameters to the appropriate fields. Note that there is no field for the GW 1&2
Recession Coefficients, but these will need to be added to your loss parameters once the HEC-
HMS project file is complete.

Save your map document.

You are now finished calculating all of the required input parameters for the SMA method in
HEC-HMS!

Note
If you are using SCS for the transform method, you will need to resume the “HMS Model
Development using HEC-GeoHMS (ArcGIS 10)" tutorial exactly where you left off. You will
need to repeat the “Select HMS Processes” and “Subbasin Parameters” exactly as the tutorial
suggests—including selecting SCS method for loss. This is because you will need to use a curve
number grid to calculate the basin lag for the SCS transform method, but for whatever reason,
ArcGIS does not include this option while assigning subbasin parameters from raster if SMA is
selected for the loss method. So, you need to perform the steps from the tutorial to ensure that the
basin lag data is generated. Do not worry; this will not cause problems with any of the
parameters we already assigned during this tutorial.

Once this is complete, resume the “HMS Model Development using HEC-GeoHMS (ArcGIS
10)" tutorial from page 18 under the heading “HMS.”

Remember to convert your units once the HEC-HMS project file has been created!

My HEC-HMS project file has been created! I have converted my units! I am ready to start
calibrating the model!

Evapotranspiration
Not quite so fast. SMA requires the use of evapotranspiration data. In HEC-HMS, click on your
Meteorologic Model. Next to Evapotranspiration, select Monthly Average.

14
You will notice that each subbasin is now listed beneath the Meteorologic Model. Using the pan
evapotranspiration data for your region, provide the appropriate evapotranspiration values for
each month for each subbasin. Use 0.70 as the evaporation coefficient. If you do not have data
during the winter months, an estimate of 0.5 to 1.0 inches is appropriate.

Baseflow
Earlier in this tutorial, we selected linear reservoir to model the baseflow but did not calculate
any parameters for this. For the GW 1 & 2 Coefficients, select values that permit the
groundwater flow to travel through the reservoirs with little to no attenuation. The GW 1 & 2
initial flows and number of reservoirs are best determined during calibration.

Final Note
When all is said and done, you will notice that you still need values for initial storage for various
parameters. If you do not have any specific data pertaining to the actual field values, just use
your best judgment to determine these values during calibration.

15
Primary Resources

Bennett, T. (1998). Development and Application of a Continuous Soil Moisture Accounting

Algorithm for the Hydrologic Engineering Center Hydrologic Modeling System (HEC-
HMS). Davis, CA: University of California, Davis.

Fleming, M. (2002). Continuous Hudrologic Modeling with HMS: Parameter Estimation and
Model Calibration and Validation. Cookeville, TN: Texas Technological University.

Linsley, R., Kohler, M., & Paulhus, J. (1982). Hydrology for Engineers. New York: McGraw-
Hill.

16
Appendix

Table 1A. NLCD Land Cover Class Descriptions

Class Description
11 Open Water
12 Perennial Ice/Snow
21 Developed, Open Space
22 Developed, Low Intensity
23 Developed, Medium Intensity
24 Developed, High Intensity
31 Barren Land (Rock/Sand/Clay)
41 Deciduous Forest
42 Evergreen Forest
43 Mixed Forest
51 Dwarf Scrub
52 Shrub/Scrub
71 Grassland/Herbaceous
72 Sedge/Herbaceous
73 Lichens
74 Moss
81 Pasture/Hay
82 Cultivated Crops
90 Woody Wetlands
95 Emergent Herbaceous Wetlands
For more information, see http://www.mrlc.gov/nlcd06_leg.php

Table 2A. SMA Inputs in HEC-HMS (Required Parameters)

U.S. Customary SI Units
Loss
Initial Soil Storage* % %
Initial Groundwater 1 Storage* % %
Initial Groundwater 2 Storage* % %
Max Infiltration in/hr mm/hr
Impervious % %
Soil Storage in mm
Tension Storage in mm
Soil Percolation in/hr mm/hr
GW1 Storage** in mm
GW1 Percolation in/hr mm/hr
GW1 Coefficient** hr hr
GW2 Storage** in mm
GW2 Percolation in/hr mm/hr

17
GW2 Coefficient** hr hr
Surface
Initial Storage* % %
Maximum Storage in mm
Canopy
Initial Storage* % %
Maximum Storage in mm
* Calibrated Parameters- don’t need to build a raster
**Constant- don't need to build raster

18