Structure
contributor_folders(optional) Each contributor can make a folder here and push their work here during the week. This will allow everyone to see each others work but prevent any merge conflicts.final_notebooksWhen the team develops shared final notebooks, they can be shared here. Make sure to communicate so that you limit merge conflicts.scriptsShared scripts or functions can be added here.dataShared dataset can be shared here. Note, do not put large datasets on GitHub. Speak to the organizers if you need to share large datasets.
Ideas
- Does the 3D CHLA estimation work with Sentinal-3 bands?
- Rerun the BRT with fewer spectral bands
- What improves estimation? Does including other Argo variables improve?
- For global prediction would need to use variables that we have globally. Temp & O2 from GOBAI?
- Does closer spatially help? PACE L2 (1k) versus L3 (4km)
- Try other ML approaches with the training dataset
- Test classic surf CHLA to integrated CHLA and DSL (deep scattering layer)
- Look for deeper peaks in Argo profiles? Are there any?
- Look for relationships with MLD (mixed layer depth)
- Look for interesting areas with where CHLA 3D patterns might be of interest. See https://github.com/fish-pace/chla-z/blob/main/notebooks/examples.ipynb
| Name | Role | Affiliation |
|---|---|---|
| Elizabeth Eli Holmes | Project Facilitator | NOAA Fisheries, University of Washington |
| Andrea | Participant | |
| Jodi | Participant | |
| Peter | Participant | |
| Charlie Turner | Participant | University of California, Davis |
| Nadir | Participant | |
| Haipeng | Participant |
- Initial idea: "short description"
- Ideation jam board: Add link
- Ideation Presentation: Add link
- Slack channel: fp25_proj_name
- Project google drive: Add link
- Final presentation: Add link