Introduction

This section documents the validation and verification (V&V) effort for the Energy Budget Snow method which was added within HEC-HMS v4.11.  This new method is based on the Utah Energy Budget (UEB) snow model developed at Utah State University (Tarboton et al, 1995).  The UEB snow model was originally developed using FORTRAN, which required conversion to Java for use within HEC-HMS.  The HEC-HMS Java implementation was evaluated using data from the Earth System Model - Snow Model Intercomparison Project (ESM-SnowMIP).  ESM-SnowMIP provides boundary conditions and observations at numerous locations throughout the world (Menard et al, 2019).  These data sets were used to validate and verify the Energy Budget snow method computations within HEC-HMS.

Goals

The goals of the Energy Budget Snow V&V effort were:

  • Make model calculations consistent between the FORTRAN (UEB Model) and Java (HEC-HMS) implementations,
  • Refactor HEC-HMS code for improved readability and consistency with HEC-HMS coding standards,
  • Ensure energy and mass conservation, and
  • Develop automated tests that will flag altered results when future modifications are made to the HEC-HMS Energy Budget snow method.

Locations

In situ meteorological forcing and evaluation data, and bias-corrected reanalysis forcing data was available for ten locations: one maritime (Sapporo, Japan), one arctic (Sodankylä, Finland), three boreal (Old Aspen, Old Jack Pine and Old Black Spruce, Saskatchewan, Canada) and five mid-latitude alpine (Col de Porte, France; Reynolds Mountain East, Idaho, USA; Senator Beck and Swamp Angel, Colorado, USA; Weissfluhjoch, Switzerland) (Menard et al, 2019).  A summary of these locations is shown below.

Site

Abbreviation

Latitude

Longitude

Elevation (m)

Vegetation

Soil

Snow-Free Albedo

Temperature Measurement Height (m)

Wind Measurement Height (m)

Data Availability

BERM Old Aspenoas53.63-106.260021 m high forest with winter SAI 1sandy clay loam0.14373801Oct1997 - 30Sep2010
BERM Old Black Spruceobs53.99-105.1262911 m high forest with LAI 3.5 - 3.8sandy loam0.08252601Oct1997 - 30Sep2010
BERM Old Jack Pineojp53.92-104.6957913 m high forest with LAI 2.5 - 2.6sand0.11282901Oct1997 - 30Sep2010
Col De Portecdp45.35.771325short grasssandy clay loam0.21.51001Oct1994 - 30Sep2014
Reynolds Mountain Eastrme43.19-116.782060short grasssilty clay0.23301Oct1988 - 30Sep2008
Sapporosap43.08141.3415short grassclay0.21.51.501Oct2005 - 30Sep2015
Senator Becksnb37.91-107.733714alpine tundrathin soil and exposed bedrock0.23.8401Oct2005 - 30Sep2015
Sodankyläsod67.3726.63179short grasssandy loam0.22201Oct2007 - 30Sep2014
Swamp Angelswa37.91-107.713371short grasscolluvium0.23.43.801Oct2005 - 30Sep2015
Weissfluhjochwfj46.839.812540barrenmoraine0.14.55.501Sep1996 - 30Sep2016

Validation Locations

Approach

An HEC-HMS project was constructed for each of the aforementioned ESM-SnowMIP locations.  Basin and meteorologic models were created to represent the land surface and meteorologic processes, respectively, at each location.  The following sections describe the model construction, parameterization, and calibration techniques that were used.

Basin Model

A basin model was created for each location consisting of a single subbasin with an insignificant area.  This was done in order to treat each location as a point.  Only snow accumulation and melt processes were simulated.  As such, no canopy, surface, infiltration, transform, or baseflow methods were selected for the subbasin elements.  The following values were initially used within the Energy Budget snow method:

ParameterValue
Rain Threshold Temperature (deg C)3
Snow Threshold Temperature (deg C)-1
New Snow Albedo0.85
Min Snow Albedo0.5
Albedo Refresh Depth (mm)3
Albedo Decay Coefficient (hr)96
Snow Thermal Conductivity (J/m/deg C/hr)300
Liquid Water Retention Fraction0.05
Snow Hydraulic Conductivity (m/s)0.005
Soil Depth for Energy Balance (m)0.1

Meteorologic Model

A single meteorologic model was created for each location.  Specified time series were used as boundary conditions for each location using the following methods:

ProcessMethod
Shortwave RadiationSpecified Pyranograph
Longwave RadiationSpecified Pyrgeograph
PrecipitationSpecified Hyetograph
TemperatureSpecified Thermograph
WindspeedSpecified Anemograph
Atmospheric PressureSpecified Barograph
Dew PointSpecified Humidograph

Input Data

Boundary condition time series consisting of in situ measurements and atmospheric model outputs along with in situ observations are available in NetCDF format at this location: https://doi.pangaea.de/10.1594/PANGAEA.897575?format=html#download.  The NetCDF data was downloaded and converted to DSS for use within HEC-HMS.  These DSS files can be downloaded here: snowMIP DSS 20220726.zip.

Computational Settings

Two control specifications were created for each location: 1) using the entire period of record (POR) of available data and 2) using a single WY.  A 1 hour computational time step was used with each control specification.

Model Calibration

Following construction and parameterization of each HEC-HMS project, the initial Energy Budget snow parameter estimates were manually calibrated until a reasonable recreation of the observed data was achieved.

Validation and Verification

Following model calibration, the energy and mass balance was computed for each time step for a single water year (WY).  A WY spans 01 October through 30 September.  The energy balance was computed using the following:

\Delta \mathrm{Q}_{\mathrm{I}}=\mathrm{Q}_{\mathrm{S} \text {,net }}+\mathrm{Q}_{\mathrm{L} \text {,net }}+\mathrm{Q}_{\mathrm{H}}+\mathrm{Q}_{\mathrm{LE}}+\mathrm{Q}_{\mathrm{R}}+\mathrm{Q}_{\mathrm{G}}

where \Delta \mathrm{Q}_{\mathrm{I}} = net change rate of internal snowpack energy, \mathrm{Q}_{\mathrm{S} \text {,net }} = net shortwave radiation, \mathrm{Q}_{\mathrm{L} \text {,net }} = net longwave radiation, \mathrm{Q}_{\mathrm{H}} = turbulent exchange of sensible heat with atmosphere, \mathrm{Q}_{\mathrm{LE}} = turbulent exchange of latent heat with atmosphere, \mathrm{Q}_{\mathrm{R}} = heat input from rain, and \mathrm{Q}_{\mathrm{G}} = conductive exchange of sensible heat with ground.  The mass balance was computed using the following:

\Delta \mathrm{SWE}=\mathrm{LWASS}-\mathrm{Precipitation}+\mathrm{Condensation}

Incremental mass balance and energy budget errors less than 0.1 millimeters and 1E-8 Watt / square meter, respectively, over 1 hour were desired.  Cumulative mass balance and energy budget errors less than 10 millimeters and 1E-8 Watt / square meter, respectively, for the entire WY were desired.

Results

The following sections describe the V&V results for each of the aforementioned locations.

BERM Old Aspen

The calibrated Energy Budget snow parameters for this location are shown here:

ParameterValue
Rain Threshold Temperature (deg C)0
Snow Threshold Temperature (deg C)-2.75
New Snow Albedo0.85
Min Snow Albedo0.2
Albedo Refresh Depth (mm)3
Albedo Decay Coefficient (hr)575
Snow Thermal Conductivity (J/m/deg C/hr)425
Liquid Water Retention Fraction0.05
Snow Hydraulic Conductivity (m/s)0.005
Soil Depth for Energy Balance (m)0.1

Two simulations were made for this location: 1) using the entire POR and 2) using only WY2006.  The results of each simulation are shown below:

BERM Old Aspen WY1997 - WY2010 ResultsBERM Old Aspen WY2006 Results

The total and maximum mass and energy balance errors for the WY2006 simulation are shown below:

ParameterValue
Sum Mass Balance Error (mm)-1.1622
Max Mass Balance Error (mm)0.0098
Sum Energy Balance Error (W/m^2)3.6687E-9
Max Energy Balance Error (W/m^2)2.6126E-11

The complete energy and mass balance verification for this location can be found here: OAS_Energy_and_Mass_Balance.xlsx.  The completed HEC-HMS project for this location can be found here: BERMOldAspen.zip.

BERMS Old Black Spruce

The calibrated Energy Budget snow parameters for this location are shown here:

ParameterValue
Rain Threshold Temperature (deg C)1
Snow Threshold Temperature (deg C)-1
New Snow Albedo0.85
Min Snow Albedo0.1
Albedo Refresh Depth (mm)3
Albedo Decay Coefficient (hr)300
Snow Thermal Conductivity (J/m/deg C/hr)400
Liquid Water Retention Fraction0.05
Snow Hydraulic Conductivity (m/s)0.005
Soil Depth for Energy Balance (m)0.1

Two simulations were made for this location: 1) using the entire POR and 2) using only WY2002.  The results of each simulation are shown below:

BERM Old Black Spruce POR ResultsBERM Old Black Spruce WY2002 Results

The total and maximum mass and energy balance errors for the WY2002 simulation are shown below:

ParameterValue
Sum Mass Balance Error (mm)-1.6386
Max Mass Balance Error (mm)0.0169
Sum Energy Balance Error (W/m^2)8.8114E-9
Max Energy Balance Error (W/m^2)2.6343E-11

The complete energy and mass balance verification for this location can be found here: OBS_Energy_and_Mass_Balance.xlsx.  The completed HEC-HMS project for this location can be found here: BERMOldBlackSpruce.zip.

BERMS Old Jack Pine

The calibrated Energy Budget snow parameters for this location are shown here:

ParameterValue
Rain Threshold Temperature (deg C)0
Snow Threshold Temperature (deg C)0
New Snow Albedo0.85
Min Snow Albedo0.1
Albedo Refresh Depth (mm)3
Albedo Decay Coefficient (hr)300
Snow Thermal Conductivity (J/m/deg C/hr)400
Liquid Water Retention Fraction0.05
Snow Hydraulic Conductivity (m/s)0.005
Soil Depth for Energy Balance (m)0.1

Two simulations were made for this location: 1) using the entire POR and 2) using only WY2007.  The results of each simulation are shown below:

BERM Old Jack Pine POR ResultsBERM Old Jack Pine WY2007 Results

The total and maximum mass and energy balance errors for the WY2007 simulation are shown below:

ParameterValue
Sum Mass Balance Error (mm)-1.0307
Max Mass Balance Error (mm)0.0075
Sum Energy Balance Error (W/m^2)1.5736E-8
Max Energy Balance Error (W/m^2)2.7921E-11

The complete energy and mass balance verification for this location can be found here: OJP_Energy_and_Mass_Balance.xlsx.  The completed HEC-HMS project for this location can be found here: BERMOldJackPine.zip.

Col De Porte

The calibrated Energy Budget snow parameters for this location are shown here:

ParameterValue
Rain Threshold Temperature (deg C)0
Snow Threshold Temperature (deg C)-2.75
New Snow Albedo0.85
Min Snow Albedo0.2
Albedo Refresh Depth (mm)3
Albedo Decay Coefficient (hr)575
Snow Thermal Conductivity (J/m/deg C/hr)425
Liquid Water Retention Fraction0.05
Snow Hydraulic Conductivity (m/s)0.005
Soil Depth for Energy Balance (m)0.1

Two simulations were made for this location: 1) using the entire POR and 2) using only WY2014.  The results of each simulation are shown below:

Col De Porte WY1995 - WY2014 ResultsCol De Porte WY2014 Results

The total and maximum mass and energy balance errors for the WY2014 simulation are shown below:

ParameterValue
Sum Mass Balance Error (mm)-1.6232
Max Mass Balance Error (mm)0.021
Sum Energy Balance Error (W/m^2)7.3251E-9
Max Energy Balance Error (W/m^2)2.7640E-11

The complete energy and mass balance verification for this location can be found here: CDP_Energy_and_Mass_Balance.xlsx.  The completed HEC-HMS project for this location can be found here: ColDePorte.zip.

Reynolds Mountain East

The calibrated Energy Budget snow parameters for this location are shown here:

ParameterValue
Rain Threshold Temperature (deg C)0
Snow Threshold Temperature (deg C)-1.5
New Snow Albedo0.85
Min Snow Albedo0.2
Albedo Refresh Depth (mm)3
Albedo Decay Coefficient (hr)350
Snow Thermal Conductivity (J/m/deg C/hr)250
Liquid Water Retention Fraction0.05
Snow Hydraulic Conductivity (m/s)0.005
Soil Depth for Energy Balance (m)0.1

Two simulations were made for this location: 1) using the entire POR and 2) using only WY2008.  The results of each simulation are shown below:

Reynolds Mountain East WY1988 - WY2008 ResultsReynolds Mountain East WY2008 Results

The total and maximum mass and energy balance errors for the WY2008 simulation are shown below:

ParameterValue
Sum Mass Balance Error (mm)-2.4373
Max Mass Balance Error (mm)0.0225
Sum Energy Balance Error (W/m^2)1.6858E-9
Max Energy Balance Error (W/m^2)2.7945E-11

The complete energy and mass balance verification for this location can be found here: RME_Energy_and_Mass_Balance.xlsx.  The completed HEC-HMS project for this location can be found here: ReynoldsMountainEast.zip.

Sapporo

The calibrated Energy Budget snow parameters for this location are shown here:

ParameterValue
Rain Threshold Temperature (deg C)0
Snow Threshold Temperature (deg C)0
New Snow Albedo0.85
Min Snow Albedo0.2
Albedo Refresh Depth (mm)3
Albedo Decay Coefficient (hr)

150

Snow Thermal Conductivity (J/m/deg C/hr)300
Liquid Water Retention Fraction0.05
Snow Hydraulic Conductivity (m/s)0.005
Soil Depth for Energy Balance (m)0.1

Two simulations were made for this location: 1) using the entire POR and 2) using only WY2014.  The results of each simulation are shown below:

Sapporo WY2006 - WY2015 ResultsSapporo WY2014 Results

The total and maximum mass and energy balance errors for the WY2014 simulation are shown below:

ParameterValue
Sum Mass Balance Error (mm)-1.1523
Max Mass Balance Error (mm)0.0157
Sum Energy Balance Error (W/m^2)2.6768E-9
Max Energy Balance Error (W/m^2)2.4222E-11

The complete energy and mass balance verification for this location can be found here: SAP_Energy_and_Mass_Balance.xlsx.  The completed HEC-HMS project for this location can be found here: Sapporo.zip.

Senator Beck

The calibrated Energy Budget snow parameters for this location are shown here:

ParameterValue
Rain Threshold Temperature (deg C)0
Snow Threshold Temperature (deg C)0
New Snow Albedo0.85
Min Snow Albedo0.2
Albedo Refresh Depth (mm)3
Albedo Decay Coefficient (hr)

150

Snow Thermal Conductivity (J/m/deg C/hr)300
Liquid Water Retention Fraction0.05
Snow Hydraulic Conductivity (m/s)0.005
Soil Depth for Energy Balance (m)0.1

Two simulations were made for this location: 1) using the entire POR and 2) using only WY2011.  The results of each simulation are shown below:

Senator Beck WY2006 - WY2015 ResultsSenator Beck WY2011 Results

The total and maximum mass and energy balance errors for the WY2011 simulation are shown below:

ParameterValue
Sum Mass Balance Error (mm)-8.7383
Max Mass Balance Error (mm)0.0315
Sum Energy Balance Error (W/m^2)3.6468E-8
Max Energy Balance Error (W/m^2)2.7610E-11

The complete energy and mass balance verification for this location can be found here: SNB_Energy_and_Mass_Balance.xlsx.  The completed HEC-HMS project for this location can be found here: SenatorBeck.zip.

Sodankylä

The calibrated Energy Budget snow parameters for this location are shown here:

ParameterValue
Rain Threshold Temperature (deg C)0
Snow Threshold Temperature (deg C)0
New Snow Albedo0.85
Min Snow Albedo0.2
Albedo Refresh Depth (mm)3
Albedo Decay Coefficient (hr)

400

Snow Thermal Conductivity (J/m/deg C/hr)300
Liquid Water Retention Fraction0.05
Snow Hydraulic Conductivity (m/s)0.005
Soil Depth for Energy Balance (m)0.1

Two simulations were made for this location: 1) using the entire POR and 2) using only WY2010.  The results of each simulation are shown below:

Sodankyla POR ResultsSodankyla WY2010 Results

The total and maximum mass and energy balance errors for the WY2010 simulation are shown below:

ParameterValue
Sum Mass Balance Error (mm)-1.3751
Max Mass Balance Error (mm)0.0079
Sum Energy Balance Error (W/m^2)1.3830E-8
Max Energy Balance Error (W/m^2)2.7704E-11

The complete energy and mass balance verification for this location can be found here: SOD_Energy_and_Mass_Balance.xlsx.  The completed HEC-HMS project for this location can be found here: Sodankyla.zip.

Swamp Angel

The calibrated Energy Budget snow parameters for this location are shown here:

ParameterValue
Rain Threshold Temperature (deg C)0
Snow Threshold Temperature (deg C)0
New Snow Albedo0.9
Min Snow Albedo0.2
Albedo Refresh Depth (mm)3
Albedo Decay Coefficient (hr)

300

Snow Thermal Conductivity (J/m/deg C/hr)600
Liquid Water Retention Fraction0.05
Snow Hydraulic Conductivity (m/s)0.005
Soil Depth for Energy Balance (m)0.1

Two simulations were made for this location: 1) using the entire POR and 2) using only WY2008.  The results of each simulation are shown below:

Swamp Angel POR ResultsSwamp Angel WY2008 Results

The total and maximum mass and energy balance errors for the WY2008 simulation are shown below:

ParameterValue
Sum Mass Balance Error (mm)-0.8204
Max Mass Balance Error (mm)0.0047
Sum Energy Balance Error (W/m^2)3.7297E-8
Max Energy Balance Error (W/m^2)2.7482E-11

The complete energy and mass balance verification for this location can be found here: SWA_Energy_and_Mass_Balance.xlsx.  The completed HEC-HMS project for this location can be found here: SwampAngel.zip.

Weissfluhjoch

The calibrated Energy Budget snow parameters for this location are shown here:

ParameterValue
Rain Threshold Temperature (deg C)3
Snow Threshold Temperature (deg C)-1
New Snow Albedo0.9
Min Snow Albedo0.1
Albedo Refresh Depth (mm)3
Albedo Decay Coefficient (hr)

350

Snow Thermal Conductivity (J/m/deg C/hr)400
Liquid Water Retention Fraction0.05
Snow Hydraulic Conductivity (m/s)0.005
Soil Depth for Energy Balance (m)0.1

Two simulations were made for this location: 1) using the entire POR and 2) using only WY2012.  The results of each simulation are shown below:

Weissfluhjoch POR ResultsWeissfluhjoch WY2012 Results

The total and maximum mass and energy balance errors for the WY2012 simulation are shown below:

ParameterValue
Sum Mass Balance Error (mm)-6.3895
Max Mass Balance Error (mm)0.0612
Sum Energy Balance Error (W/m^2)3.0574E-8
Max Energy Balance Error (W/m^2)2.7256E-11

The complete energy and mass balance verification for this location can be found here: WJF_Energy_and_Mass_Balance.xlsx.  The completed HEC-HMS project for this location can be found here: Weissfluhjoch.zip.

Summary

The Energy Budget snow method within HEC-HMS was successfully validated and verified using the ten ESM-SnowMIP locations (Menard et al, 2019).  Individual HEC-HMS projects were constructed for each of the ten locations and model parameters were calibrated using observed data to ensure that reasonable results were produced.  Following model calibration, the energy and mass balances were calculated to verify that the computations were producing appropriate results.