Dev: Determine if model_run_predictions can be dropped

**Describe the task**
`model_run_predictions` looks to be used only to fill `weather_station_model_predictions`, determine if that's true and we can drop `model_run_predictions`.

**Acceptance Criteria**
- [ ] Be able to answer yes or no to dropping `model_run_predictions`
- [ ] Verify we can use `weather_station_model_predictions` in place of `model_run_predictions` for machine learning
- [ ] Any follow up tasks written based on the above (if necessary)

**Additional context**
- Add any other context about the task here.
- Or here

Search the repository and database-related configuration for references to model_run_predictions and weather_station_model_predictions, then trace how each is populated and consumed. Check the machine-learning workflow's inputs and document whether weather_station_model_predictions can replace model_run_predictions. Done means a supported yes-or-no decision and any required follow-up tasks are recorded.

Written by the indexing model from the issue text.