Terminal Avian Hazard Detection

Birds in vicinity of airport create potentially hazardous conditions for aircraft taking off and landing. (Photo: Tomas Mellias)

Bird strikes are an increasing concern for military and civil aviation as airport operations and bird populations continue to increase. The number of reported bird strikes by civil aviation has more than quadrupled in the period 1990-2007, with the large majority of strikes occurring in the immediate vicinity of the airport.

During the same period, 92 percent of strikes occurred below 3,000 feet AGL, and 97 percent of strikes occurred in during take-off and landing. The increase is primarily attributed to three factors: increased bird populations near airports, increased air traffic, and quieter engines on commercial air carriers that are less likely to alarm birds in time to avoid collision. Read the latest joint FAA / USDA report, " Wildlife Strikes to Civil Aircraft in the United States 1990-2007” for more information.

Under FAA sponsorship, MIT Lincoln Laboratory has developed an initial automated detection capability for a proposed Terminal Avian Hazard Advisory System (TAHAS) utilizing data obtained from the Airport Surveillance Radar-9 (ASR-9) with Weather Systems Processor (WSP) enhancement. The primary objective of TAHAS is to provide continuously updated information on locations and density of avian targets in the immediate vicinity of an airfield and to generate real-time warnings of bird activity for dissemination to pilots by air traffic controllers or direct data link. A secondary goal of TAHAS is to provide continuous updates of bird locations to an external database that can be referenced by airport wildlife management personnel.

System Architecture

The proposed TAHAS architecture is shown in Figure 1. With its on-airport siting, rapid scan rate, and wide-area surveillance pattern, the ASR-9 is well-suited as the primary sensor for monitoring birds in the airport vicinity.

Roost Departure movie stillMovie 1: Animation sequence of ASR-9 reflectivity (left) and Doppler velocity (right) scans, showing several small bird roost departures and one very large roost exodus observed near sunrise at Austin-Bergstrom International Airport, TX on October 22, 1999.
View movie (6.3MB)

The ASR-9 can readily detect large flock movements such as those associated with roost departures (see Movie 1 above), and has sufficient sensitivity to detect individual or small groups of birds within 10 km (see Movie 2 below left).

Movie 2: Animation sequence of reflectivity (dBZ) images from ASR-9 scans taken ten seconds apart during a period of bird activity at Austin-Bergstrom International Airport, TX.
View movie (6.2MB)

Where available, a Terminal Doppler Weather Radar (TDWR), or a nearby Next Generation Weather Radar (NEXRAD) could provide improved altitude information and additional small target detection capability. One or more interfaces to internal or external databases are envisioned as components of the system. The databases could assist the detection algorithm by providing cueing based on historically favored bird locations (e.g., roosts) as well as regional-scale migratory bird movements obtained from the Avian Hazard Advisory System (AHAS) currently used by the US Air Force.

ASR-9 Data Processing

TAHAS detection sub-algorithms utilize knowledge-based pattern matching techniques and fuzzy logic to recognize multiple ASR-9 radar features indicative of bird flocks and individual birds. Classifying logic and heuristics are also incorporated to discriminate bird targets from other signals such as precipitation echoes. Figure 2 shows example input ASR-9 WSP reflectivity and resulting detections from the TAHAS flock detection module for a bat roost departure event observed at Austin, TX. The detected flock regions are indicated with six contour levels of intensity out to a range of 20 km and are updated every 55 seconds. Figure 3 shows the TAHAS bird tracker module diagnostic display indicating the progression of processing stages from left to right, top to bottom. Tracking movements of individual birds are shown in the lower right panel with white trails out to 10 km range and are updated every 10 seconds.

Display Concept

A user display similar to those utilized by current FAA weather detection systems such as TDWR and WSP is envisioned (Figure 4). A graphical area would depict current locations and intensity of bird flocks as well as individual bird tracks. User-selectable overlays would depict airport runways, navigational aids, and other geographic reference marks. An automated alerting algorithm could check for intersections of bird detections with runway corridors and generate formatted alert messages that could be displayed and uplinked to pilots.

top of page