Data from Aims weather stations are subjected to two quality control processes.
The first quality control process involves applying automatic rules to the raw data to flag data points that
are unlikely to be correct. These rules flag:
- Values frequently associated with sensors which are faulty, in need of a service or are not working
properly.
- Values outside believable ranges.
- Values that are out of range compared to other nearby stations.
The second quality control process involves manual visualisation of all data. Data from all sensors are
individually graphed and compared to sensors on the same station (e.g. water temperature 1 and water
temperature 2), calibrated temperature loggers, predicted values (e.g. PAR) or compared to sensors from
nearby stations (incl. BOM stations, esp. barometric pressure, wind speed and direction).
After these processes have been applied the data can be categorised in the following three levels.
Level 0 Data
Raw unprocessed data as received from the AWS. These data points have had no quality control process applied
to
them.
Level 1 Data
Level 1 data has had all suspect data points removed but no suspect data points are corrected.
Level 2 Data
Level 2 data has had all suspect data points that were removed in Level 1 corrected where possible.
