Time Series Editing and Correction
Records from remote environmental monitoring stations usually contain at least some anomalous or erroneous data. To ensure the highest quality of published data these anomalous records need to be quality controlled. The anomalies could range from sensor drift that needs to be corrected to data that needs to be deleted.
Performing these tasks requires two key components:
|
|
|
Spike removal, drift adjustment, pro-rated percent correction, copy and paste, and recession curves are just a few of the many data correction tools available. |
|
Audit Log for Defensible Retraceable DataThe second key component of the AQUARIUS Data Correction toolbox is that all changes (edits, corrections, approvals, grades, notes) are non-destructive and fully logged in a session-independent undo/redo audit track. That is, AQUARIUS never lets a user edit the raw data; rather all changes are recorded into a reversible change stack. |
![]() |
| The completeness of this audit log allows for the re-construction of a time series “as of” any time in the past. This feature makes it easy for real-time operational decision makers to defend their decisions since data can be rolled back in time to show exactly what level of correction had been applied at any point in history. | |
Data Grading (Quality Coding)Many organizations are not only required to apply corrections to their time series data, but also to apply a data grade (sometimes called a quality code) as a subjective indicator of the quality of their datasets. The Data Correction toolbox includes a facility for the mark up of datasets with data grades. AQUARIUS is pre-populated with a standard list of data grades however the grades can be customized and extended to meet the needs of any specific agency. Data grades are also propagated from source to derived datasets (e.g. grades on a stage time series are inherited by a derived discharge time series). |
|