BAM comes to Alaska

Flying Safety, Sept, 2004 by Russell P. Defusco, Wilhelm Ruhe

The state of Alaska could have its first Bird Avoidance Model (BAM) before the end of the year. The new BAM, a computer-based risk predictor for bird strikes, builds upon the familiar US BAM for the contiguous 48 states, which has been online for the past several years (http://www.usahas.com/bam/).

[FIGURE 1 OMITTED]

A coordinated effort from several agencies has produced an operational Alaska Bird Avoidance Model (AK BAM). The Natural Resources division of the Air National Guard provided major funding for the effort, with the USAF BASH Team contributing monies as well. Germany provided Mr. Wilhelm Ruhe for one year as an invaluable visiting scientist from their Geophysical Institute and is a member of the International Bird Strike Committee. Data were provided by a myriad of sources from federal, state, local, and private agencies. A coordinated team effort from government personnel and contractors was necessary to tackle the complex nature of producing the initial version of the AK BAM.

The AK BAM research team is hosted by the Institute for Information Technology Applications (IITA). Located at the USAF Academy in Colorado Springs, Colorado, the IITA is an independent research center supported by the Air Force Office of Scientific Research. The institute conducts research for the Department of Defense, the Air Force, and the USAF Academy. IITA supports acquisition, educational and operational IT needs, develops an information-rich environment to prepare graduates for the high-tech Air Force, and applies multidisciplinary expertise to IT research. They help develop research topics, select researchers, administer sponsored research, publicize results, and host conferences and workshops that facilitate the dissemination of information to a wide range of private and government organizations. With their multidisciplinary approach, the IITA was the ideal sponsor of the research leading to development of the new AK BAM.

The AK BAM operates just as the US BAM, by allowing users to analyze potentially hazardous concentrations of birds in their operational airspace. The crux of the model is the color-coded "relative risk surface" depicting distribution and abundance of birds in time and space over the entire state of Alaska (see Figure 1). Risk is defined as the likelihood of encountering a hazard and the severity of that hazard. Individual layers in the BAM define the hazard level of birds in units of airspace; thus, relative risk can be assessed by comparing one physical location with another, by comparing one time of day with another, or by comparing a period of the year with another.

Relative risk layers of the model are defined by the cumulative biomass, in ounces, of all hazardous bird species within a square kilometer of airspace from the surface to 3000 feet above ground level. Bird risk surfaces are depicted for every two-week period of the year and four daily time periods. These surfaces may be overlaid with a variety of environmental, infrastructure, and airspace depictions in a dynamic, web-based mapping application (see Figure 2). The surfaces were derived from over thirty years of ornithological data on seventy species of birds deemed most likely to cause catastrophic loss of aircraft, damage to components, or injury and/or loss of life. Species were determined by examining data from historic bird strike records provided to the USAF BASH Team by safety officers around the globe and from bird population levels as determined from numerous sources. Behavioral characteristics and activity patterns were also key in determining potential hazards posed by these species.

[FIGURE 2 OMITTED]

These same data were important in developing the US BAM, but there are several improvements made to the AK BAM. Methodological changes in the modeling techniques were made based on two decades of experience in developing the US BAM, improvement in computer processing technology and programs, and the nature of the data available in Alaska. These new techniques are now being reexamined to make future improvements and updates in the US BAM as well.

From the start of the AK BAM project, it was obvious the scarcity of ornithological data, especially in space, would require a more sophisticated approach. When observation sites are in reasonable proximity to each other, a standard interpolation technique is valid, as was used in the US BAM. However, ornithological data collection sites are unevenly and widely spaced over the state of Alaska as a consequence of inaccessible areas and low human population densities; attractive properties to many, but difficult to deal with in this instance. As a result, there are several major changes in the methodology that led to creation of the new bird risk surfaces. These changes involved additional data processing and judgment from experts in the field. The resultant calculations for the risk surface creation increased by about an order of magnitude over comparable US BAM elements. The major improvements involve three main areas, as briefly explained below.