Quality Assurance

Data from the Grazinglands Research Laboratory micronets are quality assured by the Oklahoma Mesonet. The Oklahoma Mesonet's quality-assurance (QA) system consists of four principal components:
1) laboratory calibration,
2) on-site intercomparison,
3) automated QA, and
4) manual QA.

In the instrument laboratory, all sensors (except the soil moisture sensors) are calibrated to validate or improve upon factory calibrations.

At least three times per year, the accuracy of several types of sensors in the field is verified through comparison to calibrated reference sensors housed in a portable system.

To make a first pass through the observations that are received from the micronets each day, automated QA software evaluates the data in real time.

Finally, the Mesonet’s QA meteorologists employ numerous manual techniques to complement automated QA including analysis of monthly statistics to detect sensor drift or bias (click on a month and year above to download a PDF version of a Monthly Report). For another example of manual QA techniques, download the Summary of Obstructions to Pyranometers. In addition to detecting problematic sensors, the QA meteorologists trace the true start time of each problem so that appropriate data can be manually flagged as erroneous. The QA meteorologists are responsible for communicating problems to, and coordinating with, appropriate field technicians to ensure proper resolution (see a list of the current outstanding tickets below).

Station	Variable	Date Reported	Problem Description
A133	TS05	02-Sep-2008 13:58	TS05 has developed a 1deg C low bias.
A144	TAIR	12-Aug-2008 09:02	Sensor began reporting RELH values that are 40% lower than neighbors and TAIR values that range from 20C to -75deg C.
A149	RELH	11-Sep-2008 08:52	RELH has a low bias during high humidity.
A152	TS05	04-Nov-2008 13:08	TS05 has developed a low bias.
F108	VW45	25-Nov-2008 10:47	Data spike to near zero values. Previous ticket indicated that if problem persisted, which it has, to please replace sensor.
F110	TS05	02-Oct-2008 15:16	Monthly QA indicates that TS05 has developed a low bias compared to TS10/TS15/TS30.
F115	XXXX	15-Dec-2008 12:35	TS10 and TS30 stepped down to report large negative values, and have not recovered. Only TS10 and TS30 report errant data, TS05 and TS15 look fine.