Radar/Lidar Sensors for Wind & Wake-Vortex Monitoring on Airport: First results of SESAR P XP0 trials campaign at Paris CDG Airport

Transcription

1 Radar/Lidar Sensors for Wind & Wake-Vortex Monitoring on Airport: First results of SESAR P XP0 trials campaign at Paris CDG Airport F. Barbaresco, Thales Air Systems

2 2 / Synthesis of SESAR P XP0 Sensors Trials Campaign at CDG Two configurations relative to the North runways (i.e. 09L/27R and 09R/27L) were used to benchmark the Wake Vortex and Weather sensors at Paris CDG Airport The Weather sensors benchmarked during sensors assessment campaign (XP0) were: One UHF wind profiler (PCL1300 of DEGREANE deployed under the glide), One UHF wind profiler (PCL1300 of METEO FRANCE deployed at CDG Airport), One SODAR PCS 2000 (METEK) (SODAR of METEO FRANCE still operational in CDG), One LIDAR wind profiler (WINDCUBE-70 of LEOSPHERE), Two Anemometers (of METEO FRANCE still operational in CDG) The Wake Vortex sensors benchmarked during sensors assessment campaign (XP0) were: One X band RADAR (BOR-A 550 of THALES), One LIDAR (WINDCUBE 200S of LEOSPHERE)

3 3 / SESAR P : XP0 Sensors Deployment at CDG Airport Sodar UFR Radar Wind Profiler Anemometers Visibility Wake Vortex X-band Radar Wake Vortex X-band Radar Wake Vortex & Wind 1.5 mm Lidar Scanner Meteo-France C-band Radar 1.5 mm Lidar Wind Profiler Wake Vortex 1.5 mm Radar Scanner UFR Radar Wind Profiler

4 4 / Sensors Recording Architecture in XP0 trials Meteo Centre External Weather Observations Local Meteo Sensors Anemometers Recorder UHF Wind Profiler Recorder SODAR/RASS Recorder LIDAR Wind Profiler Recorder Meteo Nowcast (Wind Profile) Mechanical scan X Band Radar Meteo Data Recorder Wake Vortex Location/Strength & short prediction Wake Vortex Decision Support System Wake Vortex Sensors WV Radar Data Recorder LIDAR scanner Recorder Lidar 1.5 um WV Lidar Data Recorder Weather Radar Recorder Air Traffic Data Recorder ATC & Airport Systems Aircraft Characteristics + 4D trajectory

5 5 / SESAR P XP0 Trials : Sensors Deployment at CDG Airport Close to Runways Wake-Vortex Sensors Deployment under the glide slope WINDCUBE-200S BOR-A radar WINDCUBE-200S BOR-A radar Wind Sensors Deployment

6 6 / X-band Wake-Vortex Radar Deployment at CDG Airport Close to runways Under the glide slope

7 7 / X-band Doppler Wake-Vortex Radar Signature in Rain Rain Doppler spectrum Wake-vortex Doppler spectrum

8 8 / Wake-Vortex tracking with X-band Radar in rainy weather A340 Take-Off B777 Landing

9 9 / X-band Radar Wake-Vortex Detection in Rain Wake-Vortex Radar Detection in Rainy Weather When the X-band radar BOR-A was located at Juilly, under the glide, it operated a vertical scan in the plane orthogonal to the glide axis. During the rainy weather condition, 34 aircrafts have crossed the scanning plane of the radar: 13 aircrafts taking-off: 4 heavy (B777, A340, A330, B767), 9 medium (B737, A319, MD80, EMB190); 21 aircrafts landing: 9 heavy (B777, MD11, A330), 12 medium (B737, A320, A321, B717, EMB170). In rain, the BOR-A was able to detect wake vortices for all of them, via the Doppler analysis of the raindrops. The observation time of the wake vortices varies from a few seconds up to 250 s. This time seems to be limited by two factors: When the rain rate is too low, it is more difficult to detect raindrops and then wake vortices; In most configurations, vortices are lost when they go out of the radar scan (in range or angle) after some time number of cases > 115 observation time [s]

10 10 / X-band Radar Wake-Vortex Detection for different aircraft categories Case Date Time (local) 16:14 15:54 9:31 11:12 15:22 9:42 9:53 Weather data: Wind profiler PCL1300 (Juilly) East wind speed [m/s] North wind speed [m/s] (direction) (South) (South) (North) (North) (North) (North) (North) Vertical wind speed [m/s] Wind direction [ ] EDR [10^-4 m2/s3] 1;2 0.3; ;4 9;51 1;6 0.4;5 Weather data: Météo France (Roissy) Temperature [ C] Relative humidity [%] Weather data: X-band radar (Juilly) Rain SNR [db] Rain rate [mm/hr] Aircraft characteristics Aircraft B777-F A330 B737 A A B B Category Heavy Heavy Medium Heavy Heavy Heavy Medium Movement Landing Landing Landing Take-Off Take-Off Take-Off Take-Off Wingspan b [m] MTOW [t] Speed V [m/s] (?) Wake vortex measures: X-band radar (Juilly) Observation time [s] Initial horizontal velocity [m/s] (North) Fall direction in the scan plane South South North North North North North Initial vertical velocity v0 [m/s] (?) Radial range [m] 510; ; ; ; ; ; ;1350 Wake vortex SNR [db] 32;41 35;44 30;35 26;30 22;35 27;37 18;24 Wake vortex estimations Vortices separation b0 = 0.8*b [m] Aircraft weight M = k*mtow [t] Time constant t0 = b0/v0 [s] Normalized observation time t/t Normalized EDR η = (EDR*b0)^(1/3)/v ; ; ; ; ;0.3 Initial circulation (1) Γ0 = M*g/(ρ*V*b0) [m2/s] Initial circulation (2) Γ0 = 2π*b0*v0 [m2/s] RCS [dbm2] -28;-25-34;-25-30;-25-29;-28-34;-26-24;-14-31;-25 Reflectivity Thales Air [dbm2/m3] Operations 29/02/ ;-70-79;-71-78;-73-78;-71-84;-76-75;-66-84;-78 Initial circulation vs aircraft configuration M g 0 0 2π b0 v0 V b0 Normalized EDR vs. normalized observation time (L: Landing, T-O: Take-Off)

11 11 / Wake-Vortex Lidar Deployment at CDG Airport

12 12 / Wake-vortex tracking by 1.5 micron Lidar in clear air

13 13 / Wake-Vortex Circulation Retrieval

14 14 / Synthesis of wake-vortex Radar/Lidar data exploitation Analysis of wake-vortex Data the X-band RADAR and the LIDAR have successfully detected wake vortices generated by aircraft of categories HEAVY and MEDIUM. Two principal sensor position options can be distinguished: Under the flight path which allows the two vortices to be separated due to the angular resolution. However it requires a large scan angle when the vortex is close to the sensor Sideways, with a vertical scan perpendicular to the corridor axis. This setup is well suited to track vortices down to the ground, the two vortices could be separated due to the range resolution Both concepts have their specific strengths and weaknesses. The optimum geometry should be chosen depending on the selected operational concept. RADAR has more restrictive limits with respect to small scan angles in order to avoid ground clutter, but This shortcoming can possibly be compensated because of the RADAR s longer range The presence of aerosols, humidity and visibility are crucial parameters which have more or less significant impact on the performance of either sensor. Ideally a sensor or a combination of sensors should yield reliable data under all weather conditions Detailed statistical conclusions about availability of data cannot be given at this stage. Such kind of assessments are foreseen for XP1 and XP2 campaigns.

15 15 / Synthesis of wake-vortex Radar/Lidar data exploitation Analysis of wake-vortex Data WINDCUBE-200S LIDAR has shown its ability to detect wake vortices up to a distance of 800 m in clear weather (equally well in upward as in sideways) Owing to the high spatial resolution (0.1 in angle and 30 m in range) and the resolution of Doppler velocity around 0.2 m/s, determination of the two vortex cores were allowed The BOR-A X-band RADAR has detected wake vortices up to 1350 m above the instrument Due to the less favourable weather conditions for Doppler-RADAR measurements in combination with the limited power budget of the BOR-A, wake vortices could only be detected during rain wake vortices were detected merely because of scattering on raindrops the variation of relative dielectric constant within the vortices. For detection in dry air, requires a higher power budget than the power which could be used in XP0. This point offers room for improvement and will be assessed during future measurement campaigns. the X-band RADAR was able to distinguish the two vortices in some cases. However its angular resolution limits its capabilities to provide accurate vortex core positions. For future campaigns, the new X-band RADAR will be designed with a narrower beam width to investigate this issue. The LIDAR nor the RADAR could be not able to detect any vortex, just behind a rain front, during a short period when the air was dry again but still clean from aerosols. In future campaigns, an optimized power setting of the X-band RADAR will be able to close the observed gap

16 16 / Synthesis of wake-vortex Radar/Lidar data exploitation Analysis of wake-vortex Data The ability of the wake vortex sensors to detect aged vortices depends on: the time the vortices stay within the area where the sensor is scanning and thus on the dimension of scanning sector and crosswind. Rapid vortex decay, e.g. due to strong atmospheric turbulence would also result in shorter vortex detection times. Other parameters like wind shear and atmospheric stability are also assumed to affect wake vortex decay and transportation, but are not included in this study. With the X-band RADAR : wake vortices could be observed up to a maximum of 250 s after the airplane wake vortices could be observed in relation to the crosswind, which transported the vortices more or less quickly out of the scanning domain. The relationship between EDR and wake vortex decay could not be analyzed Lack of algorithm to compute wake vortex circulation from a Doppler effect on raindrops. An algorithm of inversion should be developed and calibrated on Wake-Vortex simulation based Fluid mechanical model. With the LIDAR wake vortices have been observed up to 90s after the airplane detection. Variation in observation duration is due to two main factors: transport due to crosswind limitations of the algorithms to detect the wake vortex signals within a turbulent atmosphere

17 17 / Synthesis of SESAR P XP0 WV Trials Campaign at CDG Time LIDAR: 25/05/2011 9h54 for a B767 (take-off) X-band RADAR: :42 B (take-off) LIDAR: 17/05/ h06 for a B747 (take-off)

18 18 / Conclusion of XP0 Trials for Wake-Vortex Sensors Wake-Vortex Sensors Recommendations Thanks to XP0 results, it has been demonstrated that, in high altitudes, wake vortex behaviour, being affected only by the wind, is predictable. out of ground effect, wake vortex predictors will be able to compute wake vortex behaviour based on theoretical models. They need as input an accurate wind speed and direction. In these areas, no wake vortex monitoring sensor is recommended. On the opposite, close to the ground, where wake vortex behaviour is affected by IGE, a wake vortex monitoring is mandatory. sensors scanning domain must be large enough to cover both landing & take-off. the best sensors position is demonstrated to be sideways, few hundred meters upstream from the touch down area. Left Vortex Right Vortex Out of Ground Effect In Ground Effect

19 19 / Conclusion of XP0 Trials for Wake-Vortex Sensors Wake-Vortex Sensors Recommendations For wake vortex monitoring the recommendation is to deploy: an X band RADAR (electronic scanning) a 1.5µm LIDAR both located perpendicularly to the runways, a few hundred meters upstream from the touch down area. XP0 campaign has confirmed that wake vortex behaviour differs significantly depending on altitude In high altitude, out of ground effect, wake vortex behaviour is affected by the wind, but remains stable and predictable. Close to the ground. In this area low wind shear and ground effect can lead to unexpected wake vortex behaviour, (e.g. rebounds). These phenomena are very difficult to predict and to modelize.

20 20 / Conclusion of XP0 Trials for Wake-Vortex Sensors Wake-Vortex Sensors Recommendations The main results were convincing in terms of wake vortex detection: Most of wake vortices were detected in both critical areas The detection range has been demonstrated to be over the detection needs. wake vortex was detected as long as it was in the sensor scanning domain, except for some cases where detection algorithms must be tuned. Results show that RADAR and LIDAR are complementary depending on weather conditions: X-band RADAR performances are optimal under rainy conditions LIDAR performances are optimal in dry air. Nevertheless, some improvements have to be done on these sensors to reach the performances needed by an operational system: Update rate needed to scan the Wake-Vortex 3D volume should be around 10 s. This capacity is already available for LIDAR, but should be developed for RADAR by electronic scanning Both LIDAR and RADAR have evaluated the circulation of Wake-Vortex, but this part needs further algorithm development to be able to assess accurate initial circulation and decay. A gap in data availability has been observed in particular weather conditions, after a raindrop when air has been cleaned from aerosols. Thus, the RADAR power budget must be increased in order for it to detect wake vortices in the whole domain where LIDAR data are not available. These development were already planned within the project. Thus, the campaign results confirm the theoretical analysis.

21 21 / Synthesis of SESAR P XP0 Wind Sensors Trials Lidar Wind profiler WINDCUBE-70 SODAR Wind profilerpcs 2000 Radar Wind Profiler PCL-1300 Alizia 312 anemometers

22 22 / Wind sensors performances Analysis of Meteorological Data The ranges that could be covered by the various weather sensors are summarized in the table: Sensor Minimum range Maximum range 90% availability 10% availability LIDAR wind profiler WINDCUBE m 950m 1950m SODAR 20m 150m 370m Anemometers 10m 10m 10m UHF wind profiler PCL m 1600m 3950m This campaign revealed a couple of situations, which require optimisation of the sensors: LIDAR performance is linked to visibility and it s capabilities are limited by rain. The SODAR performance has to be considered as a critical item, because it is the only instrument that offers wind data from the critical height range between 10 and 100 m AGL. But availability and quality of SODAR-derived wind data are negatively affected by high ambient noise levels.

23 23 / Synthesis of Wind trials : % of availability according to altitude Lidar Wind profiler WINDCUBE-70 SODAR Wind profiler PCS 2000 Radar Wind Profiler PCL-1300

24 24 / Synthesis of SESAR P XP0 Sensors Trials Campaign at CDG Wind velocity and direction between MF UHF and DEGREANE UHF at 300m Wind velocity and direction between MF UHF and DEGREANE UHF at 2550m

25 25 / Synthesis of SESAR P XP0 Sensors Trials Campaign at CDG Weather Sensors Recommendations For weather monitoring, the recommendation is to deploy two sets of sensors: one located under the glide interception point, including : an UHF wind profiler a 1.5µm LIDAR wind profiler the other one located close to the runways and composed by; an anemometers field a high power 3D X Band RADAR a 3D 1.5µm LIDAR scanner. For wake vortex tracking and predicting, weather data are critical. The following parameters are mandatory for the system: Wind field (u, v and w components), EDR (Eddy Dissipation Rate), Temperature vertical profile

26 26 / Synthesis of SESAR P XP0 Sensors Trials Campaign at CDG Weather Sensors Recommendations Two kinds of critical areas can be defined regarding wake vortex hazards, depending on the altitude (close to the ground and at higher altitudes) In both cases, weather parameters (wind, EDR and temperature) must be provided to the system. The way to acquire these data will be: Wind field is obviously provided by weather forecast models but wake vortex tracking and predicting need accurate data in the defined critical areas, requiring direct measurements EDR is a parameter which can be : computed from wind measurements coming from different sensors. Provided by weather forecast models no specific sensor is required to acquire this parameter. Temperature vertical profile can be: measured by a RASS associated to a SODAR or a Wind Profiler. Provided by weather forecast models the recommendation is to rely on this forecast model

27 27 / Synthesis of SESAR P XP0 Sensors Trials Campaign at CDG Analysis of Meteorological Data The wind field were performed using data that were provided every 10minutes. For safety critical aviation applications however, update intervals around 1 minute are foreseen at least for the most critical parameters The Eddy Dissipation Rate, EDR, needed for Wake-Vortex Predictor is still a subject of research: EDR retrieval algorithms derived from RADAR or LIDAR wind profile measurements are under study EDR Retrieval algorithm is already available for the UHF wind profiler. Regarding the X band RADAR : a rationalization could be done by using a multifunction RADAR, able to monitor both wind and wake vortices. As the scanning domains are different, this RADAR should be an electronic scanning one. This kind of RADAR will be used for the following phases o the project.

28 28 / Sensors Requirements General Recommendations The sensors deployed for a Wake Vortex Detection, Prediction and Decision Support Tool are one of the most vitals parts of the whole system. Since the sensors are in charge of observing the reality, it is essential for the output data to be useful, reliable and on time. The sensors set used in such a system will have to be compliant with the following general recommendations Quality of the acquisitions: output data must be valid for determining the separation between each pair of aircraft, Weather and visibility conditions: even if each sensor cannot reach the required performances in all weather conditions, the sensors set as a whole shall detect wake vortices with acceptable quality for the system purposes under all possible weather and visibility conditions, Real-time measurement and processing. In order to provide operational aircraft separation support, the data acquired must be processed on a lapse of time short enough for the system output to be available in real time from the controller point of view. Electromagnetic interferences. Compromises in the positioning of sensors will have an impact on their real performance.

29 29 / Synthesis of SESAR P XP0 Sensors Trials Campaign at CDG Conclusions of XP0 XP0 confirmed the feasibility of a prototype based on existing sensor technologies. Using combinations of sensors, models and real measurements, and thanks to some improved sensors already planned within the project (e.g. more powerful RADAR), it should be possible to reach current operational needs. Considering the state of the art on sensor technology, and the needs for operational monitoring of Wake Vortices, it is recommended that further developments on sensor technology focus on: Check the performance of the selected technologies on a long-term basis, which will allow verification of the overall reliability of the sensors and their maintenance needs. Evaluate the opportunity for upgrading or replacement of some sensors with current R&D sub-systems. For future, wake vortex decision support system performances should take advantage of technology progress on sensors and associated algorithms.

30 30 / Next Phase : XP1 Trials, September/October 2012 at CDG Phase 1 Phase 2 Phase Sept Oct 2012 Full scale simulation model XP0 Trials Data acquisition: Sensors Benchmark (CDG ) WV sensors : X-band radar (mech scan) 1.5 m Lidar 2 m Lidars Weather Sensors : Ultrasonic Anemometers Lidar Wind Profiler UHF Radar Wind Profiler SODAR X-band weather radar XP1 Trials Partial prototype: «Off-Line» demonstration Time Based Separation (CDG ) WVAS System : Separation Mode Planner Wake Vortex Predictors WV Alerts Operator MMI WV sensors : X-band radar (Electronic scan) selected Lidar Weather Sensors : Selected Wind profiling sensors Model calibration & validation XP2 Trials Full scale prototype: «Shadow Mode» Weather Dependant Separation (CDG ) WVAS System : Separation Mode Planner Wake Vortex Predictors WV Alerts OperatorMMI WV sensors : X-band radar (elec scan) selected Lidar Weather Sensors : Selected Wind profiling sensors Full scale updated prototype: «Shadow Mode» Pair Wise Separation ( Frankfurt ) WVAS System : Separation Mode Planner Wake Vortex Predictors WV Alerts OperatorMMI WV sensors : X-band radar (elec scan) selectedlidar Weather Sensors : Selected Wind profiling sensors XP3 Trials Deployment of a multifunction radar with electronic scanning in September, 2012

31 31 / XP1 CDG Trials : WV & Wind Sensors Deployment Visibility Wake Vortex & Wind E-scan X-band Radar Wake Vortex & Wind 1.5 m Lidar Scanner UFR Radar Wind Profiler Runway (landing) Anemometers

32 32 XP0/XP1 / Wake-Vortex Radar Data for Radar simulator calibration Radar model in clear air UCL Wake-Vortex Radar Model Code: V1.0 : simplified model with integrated LES V2.0 : Fine LES simulation integrating turbulence matlab Interested Evolution time: s Refractivity n Log(C n2 )

33 33 / XP0/XP1 Wake-Vortex Lidar Data for Lidar simulator calibration Lidar model in clear air UCL Wake-Vortex Lidar Model Analytical model: Burnham-Hallock Clear-air conditions CNR curves versus range from XP trials Standard deviation on wind according to CNR from XP trials Code: C++ and matlab Interested Evolution time: s

34 34 / Questions