Bird migration detection with KIEVIT (~1970 – 1989)
The start of radar ornithology at the Royal Netherlands Air Force
Especially with the low flying ‘en route’ there is a risk of a bird strike with an aeroplane. That risk is even larger during the bird migration in spring and autumn. Since military aircraft often fly low during exercises, there is a risk of aircraft loss, for example, due to a motor failure. But a bird through the canopy also presents a risk to the pilot. As a pilot is not always able to use its ejector seat, the risk to lose his life is high as well.
The Royal Netherlands Air Force wanted to make that the risk manageable. The question was how?
The period 1965 – 1966
In 1965, the Royal Netherlands Air Force Division of Scientific Research and the Section for Flight and Occupational Safety conducted research into a reliable method for the photographic registration of bird movements. The aim was to gain insight into meteorological situations and the relative intensity of bird movements to produce a bird forecast in order to prevent bird strikes.
The TNO Physics Laboratory was tasked to make a simple prototype system for filming radar images. This system was placed over the Teletrack screen of the 23 cm radar at the Navigation Station Noord (NS-N) in Den Helder. Use was made of:
- A Canadian Air Force provided 16mm Vinten film camera which captured images at a low frame rate: one image per full radar rotation; four images per minute.
- A Robot recorder 36 camera with a large cassette of 35 mm film that took one image per hour with a shutter speed of 15 minutes.
A black painted hardboard casing was placed over the Teletrack screen. A small timepiece with day number indicator was mounted in the casing. This timepiece was illuminated by a shielded lamp. In addition, the system provided a 12 V voltage supply, a control unit for the Vinten camera, and a timer that controlled the magnetic release of the Robot recorder camera.
Between the end of August 1965 and mid-January 1966, the prototype system was satisfactorily used to capture screen images. The shots were sharp, and had good contrast and coverage. However, the results were strongly dependent on the screen settings of the NS-N radar.
In February 1966, the project recommended to purchase recording equipment for creating an Airforce capability to record the 1966 spring migration of birds, and to collaborate internationally with other air forces in the field of radar bird behaviour analysis. These recommendations were shelved for several years.
Ing. Im Tengeler from the radar group of the Electronic Developments for the Armed Forces Laboratory (LEOK) and some Canadians jointly developed ideas for an electronic bird migration device. He thereby assumed that birds reflect radar radiation (see, for example, the Dutch discovery of radar). Therefore, birds can be made visible using radar, even at great distances.
In 1971, these ideas result in an assignment by the Materiel Directorate of the Royal Netherlands Air Force (DMKLu) to make contact with the Bird Strike Committee Europe (BSCE). Tengeler becomes a member of the Radar Working Group that would study the radar technical aspects of bird registration in an international context. In the first meeting of that group, he presents a LEOK design of an electronic warning system for determining bird density.
In 1973, Tengeler defines the concepts of bird intensity and bird density. He makes proposals to define international scales for bird intensity and bird density, regardless of the type of radar used or the location of the radar. He also researched the usability of a Medium Power Radar (MPR) radar, the intended successor of the Thomson-CSF ER 438 radar at the navigation stations, for bird registration. The then Chief of Air Force Staff, however, appears to disregard this work …
From 1973, the Royal Netherlands Air Force no longer requests research by TNO. In a joint decision of the Air Force and the LEOK, the research project is cancelled. At the same time, it is known from contacts in foreign countries, including Canada and Denmark, that they are developing bird detection systems.
In the winter of 1973/74, the Air Force takes the decision not to include a bird watching capability in the new Medium Power Radar (MPR) of Radarpost Noord (RP-N) in Wier in North Friesland. It is feared that such a capability will hinder the Operations.
In 1974, Drs. L. Buurma graduates from the VU in Amsterdam. For his military duties, he is put to work at the Department of Air Force Management (ALBV) within the Air Force Staff. There he studies the problem of increasing numbers of bird strikes. He discovers that there are boxes full of unused recordings of photos and videos of the Thomson CSF-ER 438 radar search screens of the Navigatiestation Noord-Holland (NS-N).
Those recordings were made in the ’60s as an attempt to visualise the bird migration process. That project dies prematurely because the radar cannot be optimally tuned for visualising birds. The NS-N search radar namely had an important NATO air defence task: the detection of enemy aircraft at a large distance and remote combat control.
Buurma spends his military service mainly on a thorough analysis of the many aspects of flight safety related to bird strikes. He discovers that there is a connection between bird migration, the weather, and bird strikes. Convinced of these results, the Air Force decides to set up a bird migration warning system.
From 1975 onwards, Buurma continues his work on the prevention of bird strikes as a civilian investigator at the Air Force. With the radar movies made by his predecessors, the current understanding of bird migration at that time is adjusted and expanded.
The prevention program does not only concern radar measurements. He appoints so-called birdmen on the flight bases. Their main task is to chase birds away from the runways and the immediate vicinity of an Air Force base.
The KIEVIT system
Buurma had met Tengeler during his military service at one of the conferences of the BSCE. In December 1975, they jointly view the FAUST system of the Danish Air Force. FAUST used, among other things, the raw video signal from the logarithmic receiver of – selectable- one of the two antenna bundles. The necessary taps did not in any way affect the operational and technical characteristics of the radar.
For that reason, LEOK is commissioned to develop a FAUST-like system with some improvements and adjustments in 1977. At the beginning of 1978, the plan is ready. A phased implementation of the KIEVIT project (Kast met Integrale Elektronische Vogeltrek Intensiteits Tellers – cabinet with integral electronic bird migration intensity counters) starts. The purpose of that KIEVIT system is to determine bird migration in five ‘counting windows’: areas around the radar and in two antenna beams located one above the other. A density figure according to international standards is electronically derived. If the density exceeds a limit value, a flight ban will be announced for that area. Not everyone was happy: an intervention on Operations by a technical system!
In October 1978, the experimental KIEVIT system is connected to Radar Post Noord (RPN) in Wier. From the end of that month, both the bird video signal and the images from the KIEVIT displays are recorded. In March 1979, an important bird migration month, some days of very intense bird migration are recorded. These results allow making some statements about the functioning of the system: “The system is not yet suitable for operational use, as some essential functions do not meet the requirements. Among other things, the system’s discriminatory capacity between bird echoes and clouds, rain, and ground echoes is insufficient”.
That is why a few modifications are made to improve the bird alarm function. While the measurements continue with this system, LEOK starts a study project to improve the reliability of bird migration observations with the KIEVIT system in 1981. Reducing false alarms due to atmospheric disturbances is a priority. The focus of KIEVIT shifts from radar technology to improvements in image processing.
Birds do not take aeroplanes into account, and since they are not under Air Force command, the question remains as to whether we as Air Force should take birds into account.
Air Force biologist L. Buurma in Veilig Vliegen (March 1983)
The method of bird intensity counting as used in the KIEVIT system is further analysed by the LEOK. The identified shortcomings can be explained. Even with a low bird density and clutter suppression, the KIEVIT system fails to deliver a clear result due to the use of one-dimensional filters. Therefore, the LEOK advises: “A strong improvement can be achieved through the introduction of two-dimensional filters. With greater clutter suppression, a more accurate bird count will be possible”. That moment must be seen as the start of the follow-on development: the Radar Observation of Bird Intensity and Notification (ROBIN) system – see the next web page.
The book [in Dutch] “KIEVIT wordt ROBIN en vliegt uit“.