In Reply Refer To: September 28, 2006
Mail Stop 415
OFFICE OF SURFACE WATER TECHNICAL MEMORANDUM 2006.05
SUBJECT: Approval and Revision of Time-Series Data in ADAPS 4.6
Unit Values of discharge and stage have become important products of the U.S. Geological Survey (USGS). There has also been a continuing move toward processing surface water records continuously and finalizing them more frequently than annually, when and where appropriate. Unfortunately, the revision criteria published in the WRD Data Reports Preparation Guide (Open File Report 85-480; Novak, 1985; http://pubs.er.usgs.gov/usgspubs/ofr/ofr85480) have not kept up with these changes and have become dated. This memo supplements those revision criteria and codifies them as Office of Surface Water policy, and defines the National Water Information System (NWIS) data aging status classifications.
When Annual Data Reports were published in paper formats, revisions to discharge records were formalized and documented by publishing the revision in the Annual Data Report. According to Novak (1985, p. 103), these revisions were limited to “only those published records of discharge that are substantially in error – and only when the revisions are reliable. Revisions may result from additional data, re-examination and reinterpretation of data, or from the discovery of errors in computation. If revisions are published, an analysis should be prepared explaining the basis for making the revisions and the reasons why other periods perhaps do not need revision. This analysis should be completed and filed for reference.” The basic principles guiding revisions described by Novak (1985) are still sound and should be applied when using electronic databases and publication formats. This memorandum goes beyond Novak (1985) by providing guidance regarding revisions of Unit Values and revisions of stage and elevation data.
Approval of Data - Definitions of Data Aging Status
In the NWIS, data can be assigned three different aging statuses, “Approved,” “In-Review,” or “Working.” With the release of NWIS version 4.6 in September 2006, the data aging status can be assigned to days rather than for only a whole water year. This policy memorandum establishes the definitions, described below, for use in documenting the data aging status of time-series data following the release of NWIS 4.6:
Approved – data analysis is complete and this value is deemed to be final, supported by the USGS, and useable by the public without caveat. Data are displayed in NWISWeb as approved. This is equivalent to published data as defined in Novak and any changes to data are considered revisions. These revisions should be carefully tracked and documented.
In-Review – data analysis is complete but the data may still need review and can not be fully supported without caveat. Changes can be made and are not considered revisions. Data are displayed in NWISWeb as provisional.
Working – data analysis is still in progress up through the checking phase and is fully provisional with potential changes possibly occurring at several levels in that analysis. Data are displayed in NWISWeb as provisional.
Data Revision Criteria
Revisions to time-series data should be made if changes to “Approved” data are greater than the following criteria. Revisions can be made at the discretion of the Water Science Center for changes less than these criteria, if desired.
The criteria for the revision of extremes listed below are the same criteria stated in Novak (1985). If the difference between the old data and the new (revised) data is greater than about 10%, then the extreme discharge should be revised.
However, when Unit Value extremes associated with either annual or secondary peaks are changed based on the criteria for Unit Values, the discharges for those same events in the peak flow file should be revised so that the peaks in the peak flow file match the revised instantaneous peak Unit Value Discharges in the database. If the instantaneous value comes from a source other than the data recording device (e.g. a crest stage gage or surveyed high water mark), it is recommended that this value be inserted into the edited Unit Values in ADAPS, again to maintain database consistency. Each Water Science Center should exercise judgment regarding the value of making minor revisions of minimum values. Revisions of minimums should only be made if the change is significant.
Monthly Mean Discharges >10%
Monthly Mean Discharge is not stored in the database but is calculated from daily mean values for inclusion in the data display. If an accumulation of errors in daily mean values creates a change in the Monthly Mean Discharge that is greater than the criteria, both the daily and monthly mean values should be revised. If the Monthly Mean Discharge is below 10 cubic feet per second, the revision criteria may be relaxed unless the change is truly significant.
Daily Mean Discharges >50%
Daily Mean Discharges should be revised if the difference between the old data and the new (revised) data is greater than the criteria of about 50%, or the Unit Value Discharges used to calculate the Daily Mean Discharge are revised.
Unit Value Discharges >50%
Unit Value Discharges should be revised if the difference between any one old Unit Value and a new (revised) Unit Value is greater than the criteria of about 50%. Unit Values should also be revised if the Daily Mean Discharge values or peaks are revised. This will maintain database consistency. When a peak flow value is revised, all the data for that flood event should be revised as well.
Stage and Elevation Values
Extremes >0.2 feet
Daily Means >0.2 feet
Unit Values >0.5 feet
As with discharge values, the stage and elevation values need to be consistent in the data base. If revisions are made to Unit Values of stage or elevation, then the Daily Mean Discharges and possibly the extremes need to be revised as well. If an extreme stage or elevation is revised, the Unit Values used to determine that extreme should be revised as well. If stage values are revised and these stage values are used to compute discharge Unit Values, the discharge values should be revised as well to keep the databases consistent.
All revised data should be documented in the Manuscript portion of the station data page. Changes to the NWIS database to track data revisions are being considered for the future.
Stephen F. Blanchard (signed)
Chief, Office of Surface Water