Deriving Operationally Useful Turbulence Measurements from ADS-B Reports

Atmospheric turbulence remains a serious concern for the aviation community for several reasons, including the cost of passenger and flight attendant injuries, its effect on the efficient use of airspace, and the shortened lifespan of aircraft due to fatigue effects. Better observations of turbulence would help enhance the physical understanding of the turbulence phenomenon and its operational predictability. However, dense, routine, and quantitative observations of turbulence in the National Airspace System (NAS) are still lacking and without them it is not possible to provide complete and accurate information about the turbulent state of the atmosphere for use in tactical and strategic turbulence avoidance. Additionally, observations are required for verification and tuning of automated turbulence forecast systems such as the Graphical Turbulence Guidance System (GTG). Routine observations of turbulence are available through verbal reports from pilots (PIREPs) and through automated in situ measurements on some aircraft. Inference of turbulence from radar and satellite imagery is also available. But all of these combined sources are still not sufficient to provide timely and complete representations of turbulence in the NAS.

One viable option to augment existing turbulence observations is to use the Automatic Dependent Surveillance – Broadcast (ADS-B) one-second position and velocity data to indicate the presence of turbulence. This information can be collected by ground station receivers and can also be accessed by other aircraft to provide situational awareness. ADS–B is an element of the U.S. NextGen and Europe’s SESAR enhancements of the airspace system. ADS–B will be required on most aircraft in controlled U.S. airspace by 2020, and as of 2017 it became mandatory for some aircraft in Europe. Canada is already using ADS-B for air traffic control. Since ADS-B will soon be implemented on most aircraft and the data are available using low-cost commercial receivers, this strategy would provide much more coverage than other methods and at minimal cost. The amount of data potentially available from ADS-B is unprecedented. For example, as of July 1, 2019, there are approximately 91,000 aircraft equipped with ADS-B datalink units, of which approximately 65,000 are General Aviation aircraft.

The purpose of this research and development activity, sponsored by the Federal Aviation Administration’s (FAA)Weather Technology in the Cockpit (WTIC) program, is to determine whether an operationally useful turbulence detection algorithm using routine ADS-B reports is feasible. The key phrase here is “operationally useful.” Experience has shown that it is often a straightforward matter to show an algorithm’s capability over a limited set of case studies, but it is much more difficult to show that it has acceptable detection and false alarm qualities in day-to-day applications. Furthermore, it is desired that the ADS-B turbulence information be aircraft-independent, i.e., a “state-of-the-atmosphere” measure of turbulence intensity. Due to the significant numbers of reports that could be made available from ADS-B-equipped aircraft, along with more precise location/time information, augmenting existing turbulence PIREPs with aircraft-dependent reports would provide tangible operational benefits. Nevertheless, providing an aircraft-independent turbulence measure from the ADS-B reports would garner even more benefit. The ADS-B reports, specifically, altitude and vertical rate, are aircraft-dependent quantities, as they provide information on the aircraft response to the turbulence (and/or other quantities, such as maneuvers or wavelike phenomena). Thus, the idea is to work backwards to estimate the turbulence intensity that produced the measured response. Other challenges with the ADS-B data relate to its relatively low update rate (on the order of once per second) and the quantization of the vertical rate parameter (64 ft/min). Another aspect is that this parameter is calculated – not measured – on board. Therefore, there are signal processing artifacts (e.g., filtering) that contaminate the data.

This presentation will discuss these challenges in deriving an operational turbulence metric from ADS-B downlinks, as well as some encouraging results from the feasibility study.

This research is in part in response to requirements and funding by the FAA. The views expressed are those of the authors and do not necessarily represent the official policy or position of the FAA.