Hydrometeorological Data Quality Control
No hydrology program can be successful if the underlying data is inaccurate. The quality of the data directly affects the quality of the hydrometeorological guidance, forecasts and warnings.
WFOs are responsible for quality controlling data in their HSA and CWA. At a minimum, they are responsible for the following types of data:
Data sources at a WFO are numerous, and much of the data used by the NWS comes from other agencies. Some of the more common sources of hydrologic data include the USGS and ALERT networks. Even when the data source comes from outside the NWS, data must be monitored and the appropriate individual or agency notified when problems are noticed. Many USGS owned gages have NWS owned telemetry equipment attached to the gage. In this case, it is a joint NWS/USGS responsibility to diagnose the problem.
What needs to be done?
The following tasks need to be done in order to monitor and quality control data. Procedures need to be developed, automatic or manual, to ensure that these duties are performed:
Tools that can be used to carry out quality control duties:
v WHFS:WHFS has the ability to check for data outside a user specified range (under data ingest, range check). If incoming data is outside that range, it does not get posted to the database.
WHFS has the ability to create a list of data that has not been received within a certain time frame set by the user. WHFS can be used to scan data graphically or in table form to look for sudden jumps in values etc.. Bad data can be edited or removed from the database.
WHFS can also display forecasts along with observed data, to see how well forecasts are tracking.
Currently WHFS does not have the ability to send corrected data to other offices (RFC for example) or other databases. This feature should become available with later versions of AWIPS.
Each office should set up procedures to ensure that data is quality controlled before it goes out to the public, or to other offices. Although quality control may be assigned as someone's responsibility (an individual or a position in the office), it should be up to everyone to keep an eye out for erroneous values. Everyone working operations who comes into contact with data should be attentive to bad values, this includes the HMTs, SH, forecasters, etc. Every operational staff member should be made aware of quality controlling procedures, including the procedures that need to be followed to correct data, or notify appropriate individuals and agencies of data problems. The HSM manual should contain names and phone numbers of responsible parties for each gage in the HSA.