Partition Operational Database for Historical Data Archiving

[conbrad]

Describe the task
Based on #4031 , we can use pg_slice to generate native partitioning sql.

Acceptance Criteria

• Large tables are partitioned appropriately (e.g. date range columns that frequently exist in where clauses)
• weather_station_model_predictions should being the largest would make sense to partition first
• Partitioning of tables should be captured in migrations
• An automated job runs to create new future partitions, as well as archive (dump to s3) and delete old partitions

Additional context

• Depends on Update Postgres DB to v. 16: after OCT 31 #2365
• Or here