Download PDF
Download page Applying Channel Loss Methods in HEC-HMS.
Applying Channel Loss Methods in HEC-HMS
Last Modified: 2026-09-30 08:42:59.303
HEC-HMS version 4.14-beta.2 was used to created this tutorial. You will need to use HEC-HMS version 4.14-beta.2 or newer, to open the project files.
Download the initial model files here: Mojave_River_Channel_Loss_Initial.zip
Overview
In this workshop, you will become familiar with several channel loss methods that are available within HEC-HMS. Specifically, you will:
- Apply the Green-Ampt Channel Loss Method
- Apply the Deficit and Constant Channel Loss Method
- Analyze and compare how the inclusion of channel loss impacts model calibration
Background
Channel losses are an important hydrologic process in many arid and semi-arid watersheds, where streamflow infiltrates through permeable channel beds and recharges underlying aquifers. These transmission losses can represent a substantial portion of runoff, particularly in ephemeral stream systems where channels are hydraulically connected to highly permeable alluvial deposits. Along the Mojave River, for example, infiltration from the river is the primary source of groundwater recharge, with losses varying spatially according to channel characteristics, sediment properties, groundwater levels, and flow conditions. These losses strongly influence both downstream flood routing and groundwater recharge. To represent these processes in HEC-HMS, channel loss methods can be assigned to river reaches to simulate infiltration from the channel during flow routing. By accounting for these losses, the model more accurately represents runoff attenuation, downstream hydrograph timing and magnitude, and the partitioning of surface water into groundwater recharge, making channel loss methods particularly important for watersheds with permeable alluvial channels and losing stream reaches.
The Mojave River has been described as the lifeblood of the Mojave Desert and serves as a major contributor to the floodplain-aquifer through groundwater recharge. Conversely, as seasonal flows within the river began to fluctuate with withdrawals from the aquifer, it became apparent that groundwater discharge supplied some streamflow along certain gaining reaches of this river (Lines, 1996). Channel Loss methods can help better simulate with water balance in your model. In this tutorial, you will compare two different channel loss methods applied to reaches in a portion of the Mojave River Watershed. Four channel loss methods are available in HEC-HMS: Constant Loss/Gain, Deficit Constant Loss/Gain, Green Ampt Loss/Gain and Percolation Loss/Gain. Additional details about each method can be found here: Channel Loss/Gain - Technical Reference Manual.

Channel Loss Methods Reference
The underlying computations for channel loss methods are similar to those of the subbasin loss methods. For background information pertaining to the estimation of loss parameters, refer to Applying Loss Methods in HEC-HMS.
Parameterize the Green Ampt Channel Loss Method
- Open the MRW (Mojave River Watershed) project file and expand the MRD_to_VV_NoLoss Basin Model.
- Copy the Basin Model and rename it MRD_to_VV_GreenAmpt.
- Expand the MR-Hesperia>OG W Reach element and change the Loss/Gain method to Green Ampt.
- Navigate to the Loss/Gain tab and fill in the following values:
- Initial Deficit: 0.3
- Wetting Front Suction (IN): 10
- Hydraulic Conductivity (IN/HR): 0.65
- Fill in the same values for MR-Oro Grande>VV reach.
- The Green Ampt Channel Loss Component Editor for both reaches should resemble the figure below.

Parameterize the Deficit Constant Channel Loss Method
- Open the MRW (Mojave River Watershed) project file and expand the MRD_to_VV_NoLoss Basin Model.
- Copy the MRD_to_VV_NoLoss basin model and rename it MRD_to_VV_DeficitConstant.
- Expand the MR-Hesperia>OG W Reach element and change the Loss/Gain method to Deficit Constant.
- Navigate to the Loss/Gain tab and fill in the following values:
- Initial Deficit: 1
- Max Deficit (IN): 8
- Percolation Rate (IN/HR): 0.65
- Fill in the same values for MR-Oro Grande>VV reach.
- The Deficit Constant Channel Loss Component Editor for both reaches should resemble the figure below.

Setup and Run the Simulations
- Navigate to the Compute Tab
- Copy the MRD_to_VV_NoLoss Simulation Run and rename it MRD_to_VV_GreenAmpt.
- On the editor panel for the MRD_to_VV_GreenAmpt simulation, keep Simulation parameters the same, but set the basin model to MRD_to_VV_GreenAmpt.
- Repeat step 2 and rename the Simulation Run MRD_to_VV_DeficitConstant and select the corresponding basin model.
- Run all three Simulation Runs. You can compute each individually or select the three simulations within Compute | Multiple Compute dialog from the toolbar.
Compare Results with and without Channel Losses
Navigate to the Results tab and select the MR at LN NR VV USGS computation point for each Simulation Run. Compare the results of each of the methods. Notice how much the simulation without channel losses overestimates the overall volume of the hydrograph. Conversely, both the Green Ampt and Deficit Constant simulation runs perform better across all performance metrics.

| Calibration Results for MR at LN NR VV USGS computation point | |||||
|---|---|---|---|---|---|
| Channel Loss Method | RMSE | NSE | Percent Bias | R2 | MKGE |
| No Loss | 0.81 | 0.343 | 69.03 | 0.96 | 0.27 |
Green Ampt | 0.219 | 0.952 | -5.42 | 0.96 | 0.94 |
Deficit Constant | 0.215 | 0.954 | -4.26 | 0.96 | 0.95 |
Download the final model files here: Mojave_River_Channel_Loss_Final.zip
References
Lines, G. C. (1996). Ground-water and surface-water relations along the Mojave River, southern California. U.S. Geological Survey Water-Resources Investigations Report 95‑4189, v, 43 p. https://doi.org/10.3133/wri954189.