A statistical validation procedure for wind pressure data

Producing high quality field data is expensive and time consuming. A major portion of this expense is the investment in personnel to insure that the collected data is valid and not corrupted with signals from instruments that are malfunctioning or signals that exhibit anomalies such as electronic spikes. For a single 15-minute duration run, the processing, printing, plotting and validation process often take over two hours, with the majority of that time spent in the validation of the data. With over 1000 15-minute duration runs collected per year the required time investment is intractable for the number of researchers and the amount of resources available to perform the task. This required high investment of time has several consequences: (1) experienced wind researchers cannot perform the validation of all the data. Since much of the validation depends upon judgement and experience (which inexperienced researchers do not have), inconsistencies in the validated data naturally arise; and, (2) instrument malfunctions can go undetected for several weeks, which results in a loss of data and extends the experiment duration significantly. The existing validation procedure that is used by Texas Tech University is rather timed consuming. There is a need for a faster and better way to validate the processed data.