*2.1. Extended Terminal Manoeuvring Area*

The redesign of the EDDM TMA consists of four steps, which is simplified in Figure 1. First of all, the EDDM TMA was extended from a range of 50 to 125 NM. Secondly, when crossing the border from an outside sector of the E-TMA, all aircraft were distributed on direct routes leading towards the runway system. Thirdly, aircraft were distinguished by their technical functionality in their onboard FMSs. Thus, approaching traffic was sorted into two categories.

**Figure 1.** Four fundamental steps to enable green approaches: (**a**) extend the terminal manoeuvring area to 125 NM, (**b**) enable direct routes towards the airport, (**c**) distinguish aircraft by flight management system functionality and (**d**) introduce aircraft separation points and different routes to the end point.

Aircraft equipped with common FMSs, autopilots and no or only simple data links, such as Controller Pilot Data Link Communications (CPDLCs), were categorised. In this concept, these were referred to as 3D-FMS aircraft or non-equipped aircraft. They were able to perform a flight along a calculated trajectory but did not have the ability to meet a target time with less than twenty seconds of reliability, since they cannot sufficiently compensate changing wind conditions with an influence on their own airspeed. Additionally, the limited bandwidth of the data link did not allow a target time negotiation between the FMS and AMAN.

The second category had by aircraft equipped with an advanced FMS or 4D-FMS and a broadband data link. These were referred to as 4D-FMS aircraft and had the ability to perform a long-distance independent approach on a defined route with negotiated target times. With a 4D-FMS, aircraft had the capability to fly along a predefined 4D-trajectory and meet the target times at all points of the way with divergence of less than plus or minus six seconds. Deviations in route, altitude and speed due to changing wind conditions were automatically compensated by the 4D-FMS, even if this may mean a divergence from the optimal approach profile.

In the fourth step to adapt the E-TMA, aircraft separation points (ASP) were introduced. Those ASPs were located around the airport with a distance of around 20 NM to the runways and had nearly the same functionality, such as TMA entry fixes today. The difference from traditional entry fixes is that at this point, the aircraft with differing FMS equipage are split. 4D-FMS aircraft followed a direct route to the direct-only merge points (DOMP), located on the right and left sides of the final approaches. This DOMPs had the task to serve as stream collection points only for the 4D-FMS aircraft from one compass direction. 3D-FMS aircraft were guided conventionally from the ASPs—the downwind transition by the ATCOs.

To separate the inbound streams of 4D-FMS and 3D-FMS aircraft in the area between ASPs and final approaches, the downwind intercept altitude was 8000 ft. In this way, the direct approaches overtook the standard approaches at the possible crossing points. If there was more than one aircraft heading to the same ASP, the wake vortex separation was established with the traffic distribution to nearby ASPs before entering the TMA. If too many aircraft arrived at one ASP at the same time, additional speed and level clearances were advised. In the event of conflicts at the ASPs, the first arrival received its optimal trajectory. If additional aircraft arrived at the ASPs at the same time and could not be guided without conflict due to their optimal target time window, they were automatically treated as conventional aircraft and manually guided over the conventional routes. Figures 2 and 3 illustrate the designed model in three dimensions. The cyan routes correspond to the direct routes towards the airport, purple marks the E-TMA boundaries, direct approaches for 4D-FMS aircraft are displayed in green and the conventional routes for 3D-FMS aircraft are in orange. All remaining blue routes depict designed departure routes.

**Figure 2.** E-TMA: cyan = direct routes, purple = E-TMA boundaries, green = 4D paths, orange = 3D paths, blue = departure, red = direct-only merge points and routes.

**Figure 3.** E-TMA with distinction into 3D (orange lines) and 4D (green lines) arrival paths. Red = direct-only merge points and routes, purple = touchdown areas.

#### *2.2. Tactical Assistance Systems*

To support approach ATCOs sequencing the inbound traffic in the GreAT airspace structure around an airport, three systems were developed or refined as tactical support systems. In order to obtain an early picture of the target times of 4D-FMS approaches relative to 3D-FMS approaches, the label projection technique "ghosting" was used and extended for continuous decent approaches (CDA). *Ghosting* is the method of projecting an aircraft's label on a radar display on a different route in order to make it easier for the ATCOs to merge two routes at one waypoint [24]. The ghost position is located where, based on current performance, the aircraft would be if flying that route. The visualisation of the arrival slots planned by the AMAN was carried out on the centreline and the final approach with the help of *TargetWindows*. Finally, a precise numerical check of the planned and observed separations was made possible by the *Centerline Separations Visualisation Tool*.

#### 2.2.1. Time-Based Ghosting

Separation between ghost and real aircraft on different routes then showed the actual relative temporal spacing between those objects, as if both aircraft were on the same route [25]. This was originally done for two arrival streams on converging runways simulating a dependent parallel approach [26]. Two different methods can be used to calculate ghost-label positions: Time-based and distance-based ghosting. Distance-based ghosting can be used without problems for regular arrival routes, where two approach streams are merged on which the aircraft move with the same standardised approach procedure and speed [27]. The merging of approach streams with different approach procedures and speeds poses new challenges. These can be partially solved if a time-based "segmented ghosting" with dynamic approach speeds is used for the ghost-label's position calculation [28].

One of the tasks of approach ATCOs in the GreAT study was the merging of manually guided and CDA-performing aircraft with different speed profiles at the late merging point (LMP) onto the jointly used last six nautical miles of the final approach. A particular challenge for ATCOs was that they did not know the speed profiles of the 4D-FMS aircraft. Due to this reason, a time-based form of ghosting was developed, since here, aircraft with significantly different speed profiles had to be merged and therefore projected onto one route [29]. In time-based segmented ghosting, the current ghost-label position is calculated using the negotiated target time of the original aircraft at the late merging point (LMP) and calculating from this point in time back to the actual time to locate the position an aircraft would have if moving with a standard speed profile already on the final approach. The LMP represents the location where the 4D-FMS aircraft and ghost meet on the final approach. For the movement of the ghost on the final approach, a rudimentary flight simulator was implemented in the AMAN, which calculated the aircraft movements along the typical speed profile of a standard approach on the final approach and moved the ghost accordingly (see in Figure 4). Thereby, the phases of speed reduction and constant speeds were tuned for each aircraft so that ghost and real aircraft finally met at the LMP at the negotiated target time. If an aircraft deviates from its negotiated target time, it will also not meet its ghost at the LMP. However, since it will also cause a conflict on the final approach in the event of a deviation, it must then be downgraded to a standard approach and guided conventionally via the downwind and base leg to the final approach. In this case, it loses all the advantages of the direct approach. However, during the validation, we assumed that modern 4D-FMS can accurately maintain a fixed target time for a waypoint with only a few seconds of deviation even under unfavourable wind conditions. During pre-validations, it was shown that it is sufficient for ATCOs, regarding safety aspects, if the ghost label is faded out thirty seconds before reaching the LMP, instead of showing the ghost until aircraft meet at the LMP.

In this way, the approach ATCO was able to implement the distances between manually guided 3D-FMS aircraft and the 4D-FMS aircraft (callsign CDA123, CDA987) on the final approach, while being sure that they can be maintained all the way to the LMP.

**Figure 4.** The working principle of 3-segment ghosting. GRT234 and GRT456 are regularly guided aircraft; CDA123 and CDA987 are aircraft (4D-FMS) conducting an long-distance independent approach. The two CDAs are "ghosted" onto the final approach and centreline by adjusting their position calculations in the typical approach procedure of the manually guided aircraft.
