Checklist for v2 Developments

 

Checklist To-Do Item

Description

Evaluation

Evidence of evaluation

Group lead/deputy group lead sign off

Checklist To-Do Item

Description

Evaluation

Evidence of evaluation

Group lead/deputy group lead sign off

Run in BGC mode (offline mode)

Ensure development works with active BGC in a transient run (for land features this means running at least one offline simulation with CTC)

Verify model runs without crashes for 50 years for the land model, or 25 years for other components. “Reasonable” outputs of GPP, ET, QRUNOFF, TOTVEGC, TLAI, FSA or FSDS, TSA, Ocean DIC and DOC, but does not have to be journal-level scientifically valid.

Code review document

June 11, 2019

https://acme-climate.atlassian.net/wiki/spaces/DOC/pages/992641025

 

Run in coupled mode with BGC

Ensure development works in a B-case transient compset with BGC on

Verify model runs without crashes for 10 years. Provide outputs of e3sm_diags package for the simulation.

MOSART-heat currently is not implemented in a energy-conservation mode, so can’t not be run in a coupled mode.

 

Test with other BGC features turned on

Ensure the development works with other (already merged) BGC features turned on (if there are options defaulted off in the previous two rows)

Verify model runs without crashes for 50 years for the land model, or 25 years for other components. “Reasonable” outputs of GPP, ET, QRUNOFF, TOTVEGC, TLAI, FSA or FSDS (absorbed or incident solar radiation), TSA (2m air temperature), Ocean DIC and DOC, but does not have to be journal-level scientifically valid.

MOSART-heat does not interact with any other BGC features.

 

Input check

Ensure all inputs are available on the e3sm input data server AND document how any new inputs are generated for the feature

Inputs on server, documentation on Confluence, script to generate archived in [prepostprocessing script repository]

MOSART-heat does not require its own inputs, but rather leveraging the inputs for MOSART-wm (hydrograph parameters) and ELM (forcing).

 

Performance testing

Ensure development does not significantly increase run time

Provide timing information with and without development for BGC/performance leads to assess. Tests can be offline cases, but should be for 10 days using an ne30 grid. Please be sure that you use the same number of processors for simulations with and without the development.

Code review document

June 30, 2019

https://acme-climate.atlassian.net/wiki/spaces/DOC/pages/992870401

 

Validation testing

Add a test (or verify a test exists) using the feature so we can ensure future E3SM developments do not break your new feature/development.

Test in the E3SM testing suite with the feature enabled. At a minimum, a smoke test is needed.

June 30, 2019

https://acme-climate.atlassian.net/wiki/spaces/DOC/pages/993001473