SUMMA Test Cases

This tar archive contains a series of test cases for SUMMA for SUMMA 3.0.

There are two sets of test cases: 1) Test cases based on synthetic/lab data; and 2) Test cases based on field data.

To run the test cases you first will need to install SUMMA or alternatively you can use a dockerized version of the code. Further details are provided below.

Contents

• readme.md: this file
• check_bit_4_bit.py: python script to compare the contents of two directories with NetCDF files and highlight and summarize any differences.
• installTestCases_local.sh: bash script to install the test cases if you use a locally compiled version of SUMMA.
• installTestCases_docker.sh: bash script to install the test cases if you use a dockerized version of SUMMA.
• runTestCases_local.sh: bash script to run through all the test cases if you use a locally compiled version of SUMMA.
• runTestCases_docker.sh: bash script to run through all the test cases if you use a dockerized version of SUMMA.
• settings_org/: directory with settings files for test cases.
• testCases_data_org/: directory with input data for test cases.
• output_benchmark/: directory with benchmark outputs for all test cases.

Installation

• If you plan to use a locally compiled version of SUMMA:
• Install SUMMA. This is described in the readme files that are part of the summa distribution. Note the location of the SUMMA executable (SUMMA_EXE).
• In the top directory of this tar archive, run ./installTestCases_local.sh or /bin/bash installTestCases_local.sh. This will create three new directories (output, settings, testCases_data) with the correct path names. The original input files in settings_org and testCases_data_org remain untouched so you can recover if something goes wrong.

• Edit runTestCases_local.sh to set the SUMMA_EXE variable in that file.

• If you plan to use the dockerized version of SUMMA:

• Install docker. See https://www.docker.com/.
• In the top directory of this tar archive, run ./installTestCases_docker.sh or /bin/bash installTestCases_docker.sh. This will create three new directories (output, settings, testCases_data) with the correct path names. The original input files in settings_org and testCases_data_org remain untouched so you can recover if something goes wrong.
• Edit runTestCases_docker.sh to set the DOCKER_TEST_CASES_PATH and the SUMMA_EXE variable in that file. The first variable should be the full path to the parent directory of summaTestCases_2.x. The second variable is the docker image you want to run, e.g. bartnijssen/summa:latest, which will run the latest version of the SUMMA master branch (generally this is what you'd want to run) or bartnijssen/summa:develop although that version may not be fully compatible with this test data set.

Running and comparing the test cases

• After installing the test cases, you can run them from the top directory of the tar archive as ./runTestCases_local.sh or /bin/bash runTestCases_local.sh if you use the locally installed version of SUMMA or ./runTestCases_docker.sh or /bin/bash runTestCases_docker.sh if you use the dockerized version of SUMMA. Note that this may run for quite a while (hour) depending on the speed of your machine.
• Model output is stored in the output directory.

Test cases

Test cases based on synthetic or lab data, as described in Clark et al. (2021)

• Synthetic test case 1: Simulations from Celia (WRR 1990).
• Synthetic test case 2: Simulations of drainage through snow pack from Clark et al. (WRR 2016) based on Colbeck (1976).
• Synthetic test case 3: Simulations from Miller (WRR 1998).
• Synthetic test case 4: Simulations of the lab experiment of Mizoguchi (1990) as described by Hansson et al. (VZJ 2005).
• Synthetic test case 5: Simulations of rain on a sloping hillslope from Wigmosta (WRR 1999).

Test cases based on field data, as described in Clark et al. (2015b)

• Figure 1: Radiation transmission through an Aspen stand, Reynolds Mountain East.
• Figure 2: Wind attenuation through an Aspen stand, Reynolds Mountain East.
• Figure 3: Impacts of canopy wind profile on surface fluxes, surface temperature, and snow melt (Aspen stand,Reynolds Mountain East).
• Figure 4: Form of different interception capacity parameterizations (no model simulations conducted/needed).
• Figure 5: Snow interception at Umpqua.
• Figure 6: Sensitivity to snow albedo representations at Reynolds Mountain East and Senator Beck.
• Figure 7: Sensitivity of ET to the stomatal resistance parameterization (Aspen stand at Reynolds Mountain East).
• Figure 8: Sensitivity of ET to the root distribution and the baseflow parameterization (Aspen stand at Reynolds Mountain East) (NOTE: baseflow simulations conducted as part of Figure 9).
• Figure 9: Simulations of runoff using different baseflow parameterizations (Reynolds Mountain East).

File provenance

This README file was part of the SUMMA 2.0 test cases that were available via the NCAR website, and also of the SUMMA 3.0 test cases that were distributed via Google Drive. It is now (Oct-2022) updated to go along with the distribution of SUMMA 3.0 test cases through Hydroshare.

Author contributions

The contents of the SUMMA test cases, as well as the team working on the model code, have evolved over time. Please get in touch if you believe you have reasons to be listed as a contributor on this resource. Authors on this repository are listed in no particular order.