No News is Good News? A Look at the Reliability of Implied Fair Conditions in Aviation Weather

Among the challenges in verifying aviation weather forecasts is that the primary observation platform for many phenomena is in situ reports. As a result, information is available only where planes are flying, but pilots will generally avoid areas expected to be hazardous. Additionally, there is a known bias in pilot reports (PIREPs) such that the fair weather conditions are underreported. In order to augment the official reports, an attempt was made to determine if areas without reports where planes were known to be present could be treated as indicative of fair conditions. The presence of air traffic was determined from Traffic Flow Management System (TFMS) data aggregated to the 13-km Rapid Refresh (RAP) model grid. The weather phenomenon of interest for this work is turbulence, which is observed in two different ways: human-generated PIREPs and an automated calculation of eddy dissipation rate (EDR) from aircraft-based wind measurements. The EDR data is used as an underlying truth field against which to judge whether the absence of a PIREP is associated with smooth turbulence conditions. While the rate of non-smooth turbulence is lower where no PIREP is issued even relative to where a smooth PIREP is issued, there are confounding factors that make application of the method suspect. The presentation will explain not only the results of this investigation but also a discussion of these confounding factors and possible approaches to dealing with them.