Lessons Learned from 17 Years of Quality Assurance System Assessment using Accuracy Measures

The Oklahoma Mesonet employs a diverse approach to ensuring a high quality dataset for their measurement network. Quality assurance meteorologists closely coordinate with field and lab technicians to ensure proper identification and resolution of data problems. Starting in 2007, a procedure was added that reviewed all sensor post-field calibration results and compared the findings to whether staff pre-identified the replaced sensor as having a problem. Accuracy measures were calculated to identify sensors that are over- or under-ticketed. While ongoing analysis helped identify and correct short-term procedural problems, 17 years of routine analysis have provided multiple opportunities for lessons learned for not only identification of errors, but also improved field troubleshooting procedures to resolve problems and prevent premature sensor replacement.