Quality Control¶

🚀 Introduced in gwml2 v4.12.4 — August 13, 2025

Well quality control is implemented via the WellQualityControl model in gwml2/models/well_quality_control.py. Each well has a one-to-one WellQualityControl record that is created automatically when a new Well is saved (via a post_save signal).

Calling WellQualityControl.run() executes all three checks sequentially. Each check is skipped if it has already been run (its _generated_time field is already set), so repeated calls are safe and cheap.

Model Fields¶

Field	Type	Description
`groundwater_level_time_gap`	JSON	Measurement pairs where the time gap exceeds the configured threshold
`groundwater_level_time_gap_generated_time`	DateTime	Timestamp of last time-gap check
`groundwater_level_value_gap`	JSON	Measurement pairs where the level jump exceeds the configured threshold
`groundwater_level_value_gap_generated_time`	DateTime	Timestamp of last level-gap check
`groundwater_level_strange_value`	JSON	Measurements matching the strange-value SQL filter
`groundwater_level_strange_value_generated_time`	DateTime	Timestamp of last strange-value check

Time Gap Check¶

Method: gap_time_quality() — well_quality_control.py:60

Finds the longest gap in days between consecutive level measurements using a LAG() window function over well_level_measurement, partitioned by well_id and parameter_id. If the largest gap meets or exceeds the threshold, the result is saved to groundwater_level_time_gap.

Result shape (JSON):

{
  "parameter_id": 1,
  "current": "2024-01-15 00:00:00",
  "previous": "2021-01-01 00:00:00",
  "gap": 1110.0
}

Threshold: SitePreference.groundwater_level_quality_control_days_gap — configurable via Site Preferences, default 1095 days.

Level Gap Check¶

Method: gap_level_quality() — well_quality_control.py:92

Finds the largest jump in value_in_m between consecutive level measurements using a LAG() window function, ordered by time. If the largest jump meets or exceeds the threshold, the result is saved to groundwater_level_value_gap.

Result shape (JSON):

{
  "parameter_id": 1,
  "time": "2023-06-01 00:00:00",
  "current": 45.2,
  "previous": -5.1,
  "gap": 50.3
}

Threshold: SitePreference.groundwater_level_quality_control_level_gap — configurable via Site Preferences, default 50 metres.

Strange Value Check¶

Method: strange_value_quality() — well_quality_control.py:124

Queries well_level_measurement using a raw SQL WHERE clause from SitePreference.groundwater_level_strange_value_filter. All matching rows are saved to groundwater_level_strange_value.

Result shape (JSON — list):

[
  { "parameter_id": 1, "time": "2022-03-10 00:00:00", "value": 0 },
  { "parameter_id": 1, "time": "2020-11-05 00:00:00", "value": -9999 }
]

Filter: SitePreference.groundwater_level_strange_value_filter — configurable via Site Preferences, default value_in_m IN (0, -9999).

Site Preferences¶

All thresholds and filters are configured in the Django admin under Site Preferences (gwml2/models/site_preference.py):

Preference	Default	Description
`groundwater_level_quality_control_days_gap`	`1095`	Max allowed gap in days between measurements
`groundwater_level_quality_control_level_gap`	`50`	Max allowed jump in metres between measurements
`groundwater_level_strange_value_filter`	`value_in_m IN (0, -9999)`	PostgreSQL `WHERE` fragment to match erroneous values

Running Quality Control¶

Automatic¶

Quality control runs automatically in two situations:

On well creation — a post_save signal (well_quality_control.py:146) creates a WellQualityControl record for every new Well.
Via Celery task — the run_well_quality_control task (tasks/well_quality_control.py) runs on the update queue and triggers the management command internally.

Management Command¶

python manage.py generate_well_quality_control
python manage.py generate_well_quality_control --ids 123,456 --force

Use --force to re-run checks even if they have already been generated.

Admin Action¶

In the Django admin, go to Gwml2 → Well Quality Controls, select one or more wells, then choose Force generate from the actions dropdown.

Quality Control Action

Programmatically¶

from gwml2.models.well_quality_control import WellQualityControl  # well_quality_control.py

qc = WellQualityControl.objects.get(well_id=<well_id>)
qc.run()