Runway determination using two-point time difference method Taichi Higashioka 1 Yoshio Tadahira Manabu Sugiura Etsushi Fujita Osamu Kohashi Nihon Onkyo Engineering Co., Ltd. 1-21-10 Midori Sumida-ku Tokyo, Japan

ABSTRACT Around airports having multiple runways, aircraft noise affected areas vary depending on the runway operation because the flight path changes by the operation. Therefore, it is important to figure out exactly which runway was used for analyzing the noise impact on the surrounding area. For this purpose, runway information included in flight data provided by the airport operator can be used in most commercial airports. However, for some airports, the data is unavailable or not always accu- rate. We developed “Takeoff/ landing runway determination system; DL-TLS” for determining run- way operations and have been utilizing it to aircraft noise analysis. However, that system requires to place measurement stations at the end of each runway to ensure the high level of accuracy. To solve this issue, we focused on difference in acquisition time of the same transponder signal at two meas- urement stations. This new approach achieved satisfactory level of runway determination accuracy, by matching the time difference changes with template for each runway. We proved that this method can be used in combination with the current DL-TLS with even higher accuracy. In addition, we showed the possibility of reducing number of measurement stations and eliminating the measurement location constraint.

1. INTRODUCTION

Aircraft noise measurement in Japan is required not only the calculation of annual assessment values at the monitoring point, but also noise analysis from various perspectives, such as the time period of noise occurrence, used runway, aircraft model, etc. Particularly, around airports having multiple runways, aircraft noise affected areas vary depending on the runway operation because the flight path changes by the operation. Therefore, it is important to figure out exactly which runway was used, for analyzing the noise impact on the surrounding area. In most commercial airports, flight log data provided by the airport operator can be used for understanding flight operations. The basic data required for noise analysis, such as takeoff and landing times, destinations, flight numbers and aircraft models, are included in the flight log data at most airports, however, in some airports, used runway information is unavailable or not always accurate.

The authors developed the aircraft noise monitoring system [1] and automatic aircraft noise iden- tification by sound source discrimination [2] to monitor aircraft noise accurately at each measurement

1 taichi_higashioka@hibino.co.jp

worm 2022

station. It has been proved that aircraft noise can be automatically identified with high accuracy. In addition, the “Takeoff / landing runway determination system; DL-TLS” [3] was developed to iden- tify the aircraft as noise source by identifying runway operations and linking measured noise with flight log data. Using these technologies, detailed noise analysis has been carried out around many airports in Japan. However, it is also become clear that conventional DL-TLS methods are not always sufficiently accurate, and this paper proposes a new method to solve this issue.

2. THE ISSUE OF CURRENT DL-TLS

In current DL-TLS, measurement stations are located at the end of each runway. The system ana- lyzes the transponder signal measured at each station to identify the takeoff and landing times. Then, by determining which station the aircraft passed over, the system identifies which runway was used.

Figure 2.1: DL-TLS conceptual diagram Therefore, stations for DL-TLS should be located at the end of each runway to identify the runway operation accurately. Alternatively, even if it is not possible to place them at ideal location, runway determination can be carried out if the station is located directly under the flight path of the target runway. However, depending on the relationship between the station placement and the aircraft flight path, sufficient accuracy is not achieved in some cases, e.g., where aircraft does not pass over the station or pass over the station targeting a different runway.

Monitoring points on the south side of Osaka International Airport with parallel runways are given as a specific case. As it was not possible to install equipment at the end of the runways, the two noise measurement stations directly under the northerly landing paths are configured to be also used for DL-TLS. The two southern stations target Runway 32 landing (L32R, L32L) for northbound opera- tions and Runway 14 takeoff (T14R, T14L) for southbound operations. Runways are labelled by the first two digits of the compass bearing, and when parallel runway points in the same direction, the runways are labelled Left (L) or Right (R). For Runway 32 landing, runway determination can be carried out with high accuracy because the aircraft passes over each station as shown in Figure 2.2.

worm 2022

Figure 2.2: Runway 32 landing track map However, in the case of Runway 14 takeoff, the aircraft turning eastwards after takeoff often do not pass over the station of the target runway or pass over the station of the adjacent runway, making it impossible to determine the used runway accurately. Figure 2.3 shows the examples of relationship between the station placement and the aircraft flight path.

Figure 2.3: Examples of Runway 14 takeoff track map

worm 2022

3. TWO-POINT TIME DIFFERENCE METHOD

3.1. Prerequisite

The proposed method uses the same transponder signal measured at two stations, to judge runway operations that are impossible or difficult to determine by using current DL-TLS. Since the acquisi- tion time of the same transponder signal is used as a parameter, the time of two stations must coincide. Also, before using this method, it is necessary to identify the takeoff and landing times of aircrafts to be judged, based on the flight altitude data and the vertical status code obtained by analyzing the transponder, as in the conventional method.

3.2. Making the template data

From the difference in acquisition times of the same transponder signal at the two measurement stations, the distance difference from the aircraft to both stations was calculated according to the Equation 1:

𝑑= ሺ𝑇 1 −𝑇 2 ሻ∙𝑐 , (1)

where 𝑇 1 is the acquisition time of Station 1, 𝑇 2 is the acquisition time of Station 2, and 𝑐 is prop- agation velocity: 3 ∙10 8 ሺ𝑚𝑠 Τ ሻ .

As an example, the results of calculating the distance difference from the aircraft position using two stations nearby the Osaka International Airport are shown in Figure 3.1.

Figure 3.1: The calculated distance difference at Osaka International Airport.

worm 2022

For each runway operation, a template of the distance difference waveform was set up by deter- mining the amount of change in the distance difference as the aircraft moves. The templates were obtained by averaging the measured distance difference waveform data for each operation. The data for each operation were moved on the time axis so that they overlapped as much as possible, and the results obtained by calculating the average of the data at the same time were set as the templates. Figure 3.2 shows the templates set for each runway operation.

L14R L14L L32R L32L

T14R T14L T32R T32L

Figure 3.2: The Templates of distance difference waveform set for each runway operation.

3.3. Runway determination

The analyzed subject data was for a total of 3 minutes, one minute before and after the takeoff / landing time in minutes of the aircraft to be judged. From the difference in acquisition times of the same transponder signal of the target aircraft received at the two stations, the amount of distance difference change from the aircraft to both stations was calculated according to Equation 1. The cal- culated distance difference change was matched to the template according to Equation 2:

σ ሺ𝑑 𝑡𝑖 −𝑑 𝑚𝑖 ሻ 2 𝑗 𝑖=1

𝑚𝑎𝑡𝑐ℎ_𝑣𝑎𝑙ሾ𝑘ሿ=

𝑗 2 , (2)

where 𝑑 𝑡 is template data, 𝑑 𝑚 is distance difference waveform data and 𝑗 is matching data length.

The runway operation with the lowest match_val , i.e., the operation most consistent with the tem- plate, was adopted as the runway operation of target aircraft.

worm 2022

4. RUNWAY DETERMINATION ACCURACY

4.1. Accuracy using the time difference method

The accuracy of runway determination using the time difference method was verified using actual data at Osaka International Airport. The verification was targeted on 14 to 16 July 2021, when Run- way 14 operation took place, and the runway determination accuracy could not be ensured by current DL-TLS. The number of takeoffs and landings at the airport during these three days was 757, accord- ing to the scrutinized flight log data.

First, the results of runway determination using the conventional method are shown in Table 4.1.

Table 4.1: Runway determination accuracy of current DL-TLS

Runway Determination

Accuracy

L14R L14L L32R L32L T14R T14L T32R T32L

L14R 29 63 32%

L14L 21 1 95%

Correct Runway

L32R 66 3 96%

L32L 6 189 97%

T14R 58 11 84%

T14L 26 39 1 59%

T32R 1 113 99%

T32L 2 1 127 98%

As for Runway 32 operation, all of them could be determined with a high accuracy of more than 95%. In contrast, for Runway 14 operation, only L14L exceeds 95%, however the accuracy was low across the board for the other operations.

Next, Table 4.2 shows the results of runway determination using the time difference method.

Table 4.2: Runway dete r mi nat ion accuracy using the time difference method

Accuracy Runway Determination

L14R L14L L32R L32L T14R T14L T32R T32L N / A

L14R 85 1 1 5 92%

L14L 22 100%

Correct Runway

L32R 69 100%

L32L 190 5 97%

T14R 69 100%

T14L 3 1 61 1 92%

T32R 7 1 106 93%

T32L 1 129 99%

Good matches were shown for each operation. Runway 14 operation with low accuracy using the conventional method were also determined with an accuracy of more than 92%, and all data for L14L

worm 2022

and T14R were correct. On the other hand, there were a total of 10 data for which distance difference waveform data could not be calculated, and it was not possible to carry out judgments for all aircraft operations using this method.

Errors in the pattern of the wrong runway for the same operation was not observed on the three days analyzed in this study, however, errors in takeoff and landing, or in the direction of the operation, occurred for each operation. Although the overall accuracy was good, from the perspective of noise analysis, errors in takeoff/landing and operational direction are not desirable, so it is necessary to consider methods to eliminate these errors if only this method is used to determine the runway.

4.1. Accuracy in combination with conventional methods

Finally, we confirm the accuracy combining this method with the conventional method using at Osaka International Airport. The way of the combination was to override the results of the runway determination by this method for results that were judged to have low accuracy using conventional method. The results are shown in Table 4.3.

Table 4.3: Runway dete r mi n ation accuracy in combination with DL-TLS

Runway Determination

Accuracy

L14R L14L L32R L32L T14R T14L T32R T32L

L14R 88 4 96%

L14L 22 100%

Correct Runway

L32R 69 100%

L32L 195 100%

T14R 69 100%

T14L 3 62 1 94%

T32R 114 100%

T32L 130 100%

Runway determinations could be made with a very high degree of accuracy for all operations, especially for Runway 14 operation that were not well performed by the conventional method. For L14L and T14R were all correct, and for L14R and T14L, there were a few cases of incorrect judge- ments on the used runway, although takeoffs and landings were judged correctly. In addition, there was one error in operational direction.

It was shown that the combination of the conventional method and the present method enables accurate runway determination even when it is not possible to place a measuring station at the end of the runway. This shows the possibility of flexible installation of measuring stations for runway de- termination for each target airport.

worm 2022

5. CONCLUSION

The two-point time difference method was proposed as one of the runway determination methods. For those which distance difference data could be measured, it was shown that runway determination was possible with a high degree of accuracy.

Combined with current DL-TLS, runway determination carried out with very much high accuracy, even for operations that could not be accurately determined by the conventional method.

The combination of this method and the conventional method enables accurate runway determi- nation even when it is not possible to place measurement stations at ideal position, indicating that the installation of measuring stations for runway determination can be flexibly adapted to each target airport. 6. ACKNOWLEDGEMENTS

The authors would like to thank Kansai Airports Co., Ltd. for providing measurement data and flight log data. 7. REFERENCES

1. T. Okuda, H. Tsuru, N. Hayashi, T. Okabe and S. Ohashi, Noise monitoring by local self-govern-

ing bodies around new Tokyo International Airport. Proceedings of Inter-Noise 96, pp. 1931- 1936. Liverpool, U.K., July-August 1996. 2. K. Yamashita, Y. Tadahira and S. Ohashi, Unattended aircraft noise monitoring using the omni-

directional noise source analyzing method. Proceedings of Inter-Noise 07 , pp. 380-387. Istanbul TURKEY 2007. 3. Y. Tadahira, K. Yamashita and S. Ohashi, Unattended monitoring technique for identifying the

aircraft noise: the method for correlating the observed aircraft noise with an airport mode and an aircraft operation. Proceedings of Inter-Noise 07 , pp. 398-406. Istanbul TURKEY 2007.

worm 2022