Scheduled Subdaily-to-Daily Aggregation ETL Task

[daniel-slaugh]

Implement a dedicated scheduled ETL task that aggregates sub-daily HydroServer datastreams to daily datastreams (mean or end_of_interval), with UI configuration, run status, logs, and failure notifications.

DWRi's current setup:

• Most Legacy TCP stations aggregate sub-daily data into a time-weighted daily mean dataset (trapezoidal)
• One of their stations (beaver_mt) contains a daily summary dataset which uses the last value of each day for a reservoir storage dataset

Frontend Requirements

Task form:

• Add a select field that defaults to 'ETL' but "Aggregation" is also an option. This should map to task.type.
• if the task.type == "Aggregation":
  1. Data connection isn't needed for aggregation tasks, so remove the data connection select field and set data connection to null.
  2. Add a timezone selector similar to the other timezone selectors on the frontend except we know hydroserver will always give us a standard ISO timestamp. So the user only needs to select a two button group for either fixed offset or daylight savings aware. Then, they'll select the specific offset or IANA timezone from the list.
• The schedule section will remain the same
• The swimlane section will require a datastream selection as the source and target instead of just the target.
• Add a dropdown to select the aggregation statistic with three choices: 1. simple mean. 2. Time-weighted daily mean 3. Last value of day

hydroserverpy Requirements

• Add aggregation execution flow:
  • Check phenomenon_end_time for source and destination datastream. Short circuit if there's not at least a day between them.
  • Fetch 3 days behind destination datastream's phenomenon_end_time to current date.
  • for each closed day time range, apply aggregation statistic.
  • Save new data to destination datastream
  • Instead of using hydroserverpy's extract, transform, load pipeline, add a custom path that fetches, aggregates, and loads more efficiently. Make sure this works for both Celery and the Streaming Data Loader.

Backend Requirements

• Add user friendly logs to allow a workspace owner to easily see the status of their aggregation tasks

Acceptance Criteria

• Owner/Editor can configure an aggregation task with one or more mappings.
• Task runs on schedule and via run-now.
• Output daily values match selected statistic.
• Aggregation window reliably converts sub daily into daily values
• UI displays task state and logs.
• Failures are logged and notification emails are sent.
• Viewer cannot modify task configuration

Testing

• Backend: aggregation correctness, schedule/run-now, error+notification paths, permissions.
• Frontend: form validation, role-based actions, task status/log rendering.