A Climatology of Turbulence Reports

 Tressa L. Kane, Barbara G. Brown, and Robert Sharman
 Research Applications Program
National Center for Atmospheric Research
Boulder, Colorado

Abstract

In the interests of safety, much effort is being expended to predict aircraft hazards such as turbulence. Observations of turbulence are essential to this effort. Researchers can learn much by studying prior cases, and forecasts can be improved through use of a climatology. Unfortunately, turbulence, particularly the clear air variety, can be difficult to observe. The greatest number of turbulence observations come from the pilots of aircraft experiencing it. These pilots' reports (PIREPs) are used in this paper to develop a climatology of turbulence.

Reports of turbulence collected for 6 years (between April 1992 and March 1998) form the basis for the analyses. In an attempt to differentiate reports of clear air turbulence, the subset of PIREPs from above 20,000 feet in altitude are studied separately.

This study considers the distributions of time, location, and altitude of turbulence reports, both overall and by season. Additionally, PIREP characteristics are investigated by location. The frequencies of PIREPs are summarized and displayed in 1 degree latitude-longitude "squares" across the United States. Maps of the frequencies are used to show seasonal frequencies, including indications of whether the variability of turbulence PIREPs includes seasonal migration. Also, a map of severe turbulence reports, which are infrequent, is used to evaluate whether these important events are concentrated in certain seasons or geographic areas.

Finally, the persistence of turbulence observations is considered. Persistence forecasts are not currently used in turbulence algorithms, but could prove to be very useful. However, very little is currently known about the persistence of turbulence events. For this analysis, the conditional probabilities of reports of turbulence events in each 1 degree "square", given that a turbulence event was recorded there earlier, are computed. To determine how long turbulence events tend to persist, if at all, persistence probabilities are computed for a set of increasing times since the event.

Basic results of the analysis indicate that the subset of turbulence PIREPs from over 20,000 feet are similar to the overall set of PIREPs with respect to many of the characteristics examined. Both sets exhibit hourly, daily, and seasonal variability. Reports of severe turbulence are infrequent at all levels and above 20,000 feet. Finally, more PIREPs are received from certain geographic areas than from others. These results are expected to aid researchers in their efforts to improve turbulence forecasting.