Timeline & File Naming Convention#
Attention
Ready to make your model outputs accessible to other MIP participants? Please refer to this page to learn how to upload your model outputs to the repository.
Processed model outputs are invited for commit to the GitHub repository under /comble-mip/output_xxx/YOUR_MODEL_NAME/ where xxx can be les or scm. Within your model’s subdirectory, you should place any intermediate or sensitivity simulations in a directory called devel. Once you are happy with your simulations for the specifications outlined in the tables below, you may host them in a directory called sandbox. While these can be committed and removed at any time, note that we expect sandbox simulations to be finalized by the dates listed below.
We request that you name your simulations as outlined below so that they can be readily compared with other runs using the same test specification. Please add a prefix to each of the file names to distinguish between other models. For example, for the SCM, liquid-only specification in Part I, outputs from the ModelE3 SCM model should be named ModelE3_FixN_noice. If your submission is for an LES model, then we ask that you also include a prefix indicating the domain size (units of km) and grid cell spacing (units of m). For example, for the Small-domain LES, liquid-only specification in Part I, outputs from the DHARMA LES model should be named DHARMA_Lx25_dx100_FixN_noice.
Our requested prefixes for the domain size and grid cell spacing are as follows for the LES runs:
Small-domain (Lx=25.6 km, dx=100 m): Lx25_dx100
Large-domain (Lx=128 km, dx=100 m): Lx125_dx100
If you have additional LES submissions that have different domain sizes and/or grid cell spacings, then we ask that you adjust your prefix information accordingly.
Deviations from requested specification: If you are unable to configure your model according to the requested specification for a particular simulation, then please append to the end of your file by including _alt. We ask that you also include an alt descriptor as a global attribute to the file to explain how exactly your simulation deviates. Be as descriptive as possible! We will do our best to include deviating simulations with the official MIP results.
Sensitivity simulations: We encourage participants to conduct their own sensitivity simulations. If you would like to contribute a sensitivity run, then please append to the end of your file by including _sen_x where x is a number (begin numbering at 1). We ask that you also include a sen descriptor as a global attribute to the file to explain what sensitivity you’ve conducted. Again, please be as descriptive as possible!
Part I#
Product |
Simulation Name |
Request Date |
|---|---|---|
SCM/Small-domain LES, liquid-only |
{prefix}_FixN_noice |
Apr. 1, 2024 |
Part II#
Product |
Simulation Name |
Request Date |
|---|---|---|
SCM/Small-domain LES, prognostic aerosol, diagnostic ice |
{prefix}_ProgNa |
May 1, 2024 |
Part II (optional; all SCM/small-domain LES)#
Product |
Simulation Name |
Request Date |
|---|---|---|
Prognostic droplet, no ice |
{prefix}_ProgNc_noice |
TBD^ |
Notes#
*This simulation will use the default roughness length formulations in each host model. All other simulations will use the fixed roughness lengths as defined on the Main Model Configuration page.
**Given its high computational cost, we suggest that participants refrain from running this setup in case there are any last minute tweaks to the model forcing; for instance, large-scale forcing required to generate realistic roll structures.
^We plan to first evaluate model outputs before recommending a request date for these additional sensitivity simulations.