.
Update of the arrival time
The arrival time mentioned in the headers are updated with the time of the last event time from the flight legs.
.
Move to the inconsistency table of flights having:
o
A missing element like a leg.
o
A first event time later than 950 minutes.
o
A single leg.
.
Update of legs with the same event time
The rounding to the nearest minute of event time given by the radar leads to the existence of points with a different position but an identical time. To avoid this problem, the average position of the legs with the same event time is calculated and a single leg is kept.
.
Removal of trajectory kinks
The purpose of this step is mainly to eliminate kinks in flight trajectories that are common when successive positions originate from different radar centers especially when these positions are close to each other. Kinks in flight trajectories are due to two factors: limited accuracy of radar data and times rounded to the closest minute.
The method carried out handles short flight legs less than 10 Nautical Miles. Two successive points are averaged if at least one of the following conditions is met:
If a flight leg is:
.
Shorter or equal to 1 NM always calculate the average.
.
Longer than 1 NM but shorter than 5 NM and the Reported Ground Speed is over 350 knots.
.
Longer or equal to 5 NM but shorter than 10 NM and the Reported Ground Speed is over 400 knots.
.
Longer than 1 NM but shorter than 10 NM and:
o
The calculated average speed is more than twice the Calculated Average Speed of the previous leg.
o
The Calculated average speed of the previous leg is more than twice the Calculated Average Speed.
The Reported Ground Speed is supplied with the original ETMS Flight Data while the Calculated Average Speed is derived from the time difference between two successive positions and the calculated Great Circle distance between these positions. Both speeds are expressed in knots where 1 knot = 1NM per hour.
Great Circle Distance
Calculated Average Speed =
Time Difference
Reported Ground Speeds seem to be reliable for short distances but are not always representative of the average speed over long distances. Ground Speeds over 350 knots are typical for larger aircrafts during the final stages of their climb-out, while cruising and at the beginning of their descent stages. Averaging of flight legs with slower Ground Speeds would affect the initial take-off and final approach paths.
This averaging method for short legs eliminates most kinks in flight paths (Figure 14). However, it does not handle kinks observed in slower flights with cruise speeds less than 350 knots. Therefore an additional filtering criterion is introduced based on the comparison of Calculated Average Speeds of two successive flight legs. When the averaged speed of a short flight leg of less than 10 NM is twice as fast or twice as slow as the previous leg the points of the current leg are averaged to a single point. This last criterion will affect the trajectory of some flight with many turns (Figure 15).
Every time a flight leg is averaged, the above-described criteria are reapplied to the new averaged point. There is a slight bias in the method when more than two points are averaged in that the last point will have more weight e.g. if three points are averaged the last point will have 50% weight and the first two points 25% each since the two first points are averaged first then the third point is averaged with average of the two first points. The number of flight legs where more than two points are averaged is relatively small and the bias will only have a small impact compared to other inaccuracies. 中国航空网 www.aero.cn 航空翻译 www.aviation.cn 本文链接地址：AERO2K Flight Movement Inventory Project Report(25)
