Downscaling as a way to predict hazardous conditions for aviation - PowerPoint PPT Presentation

SESAR Innovation Day 2013, Stockholm, Sweden Downscaling as a way to predict hazardous conditions for aviation activities Adil RASHEED, Karstein Sørli, Jakob Kristoffer Süld, Knut Helge Midtbø Applied Mathematics Strindveien 4, Trondheim, NORWAY adil.rasheed@sintef.no www.adilrasheed.com SINTEF ICT 1

OVERVIEW • Context • Background • Flow in complex terrain • Forecast • Computational efficiency and robustness • Validation strategy • Conclusion SINTEF ICT 2

Context: WP 12.2.2 External Weather Observations Meteo Centre INT-EXT-MET Wake Vortex Decision Support System Local Meteo Sensors Local Weather Nowcast INT-ITWS-1 Anemometers & Forecast INT-LWF-1 UHF Wind Profiler 500mX500m MHRPS SODAR/RASS INT-ITWS-2 INT-LWF-2 Turbulences Calculation 50mX50m INT-ITWS-3 Data Fusion Weather LIDAR HMI INT-LWFN-1 Local 1.5 µm LIDAR Weather INT-WVAS-2 Data Cube INT-WVAS-1 X Band Radar Supervisor Input / Output INT-WVDET-6 Separation Mode INT-WVAS-3 Planner Wake Vortex Sensors Radar INT-WVDET-1 Wake Vortex Radar Wake INT-WVDET-3 Approach Front- Predictor Processing End Monitoring & Electronic-scan Radar Alerting Wake Plots INT-WVDET-4 INT-WVAS-4 INT-WVDET-5 1.5 µm WV Lidar Tracking Wake Vortex Advisory Tower Lidar INT-WVDET-2 Lidar Wake System Front- Processing End INT-ATCS-2 INT-ATCS-1 Aircraft Characteristics + 4D trajectory ATC & Airport Systems ENAC, Master AATM4 - November 16, SINTEF ICT 2011

NON-NORWEGIAN AIRPORTS (Terrain) AMSTERDAM GENEVA SINTEF ICT PARIS FRANKFURT

NORWEGIAN AIRPORTS (Terrain) SINTEF ICT

Background: Aviation Hammerfest Airport Just before landing the wind speed veered and increased, creating a tail wind. The increase in the descent rate was compensated, but was insufficient, and the plane had a touch-down on the right main landing gear, with the leg failing and the aircraft sliding on its belly. Wideroe DH8A on May 1st 2005 The aircraft was written off and Widerøe was The Aviation Herald criticized for permitting landings under too high winds and gusts Norwegian Civil Aviation Authority imposed stricter wind regulations upon the airport. SINTEF ICT 6

Wind shear in mountainous terrain SINTEF ICT

HORIZONTAL SHEAR SINTEF ICT

Mountain waves: Qualilative Characteristics SINTEF ICT

Mountain waves: Characteristics • Maximum amplitude on the leeward-side of the hill • Successive hills might enhance or diminish the strength of the waves • The waves are more pronounced when the buoyant and inertial forces are comparable. The ratio is defined by Froude no. SINTEF ICT

Can the flow characteristics be modelled ? SINTEF ICT 11

Governing Equations SINTEF ICT 12

Mountain Waves Fr=1, stable stratification Fr=U/(NL) N2=(g/T)(dT/dz) Maximum amplitude on the leeward side of the hill SINTEF ICT

SANDNESSJØEN AIRPORT: STOKKA Stokka Tail wind on both directions of the runway SINTEF ICT

Fr=0.2, Lateral movement of air more pronounced SINTEF ICT

Fr=1, Ideal condition for the propagation of waves Waves are diminished by destructive interference SINTEF ICT

SINTEF ICT

Confirm the Pilots experidence "Tail Wind from both sides of the runway" SINTEF ICT 18

The simulations seem to confirm pilot's reports BUT….. Can we forecast flight conditions ? SINTEF ICT 19

Micro scales: Global scales: seasonal changes, terrain effects, Sea currents etc. Meso scales: effects of large mountain waves mountains, sea, forest, precipitation Each model is capable of resolving only a particular range of spatio-temporal scales The problem can be handled through nesting of different models SINTEF ICT

N UM1 E S T UM1 UM4 I N G UM1 SINTEF ICT

Værnes airport SINTEF ICT 22

SINTEF ICT 23

Hammerfest airport SINTEF ICT 24

Hammerfest SINTEF ICT 25

Is the model Computationally efficient and robust? SINTEF ICT 26

NJORD: Hardware Configurations • Mythologically NJORD is the God of the wind and fertility as well as the sea and merchants at sea and therefore was invoked before setting out to sea on hunting and fishing expeditions. He is also known to have the ability to calm the waters as well as fire. • Technically 192 nodes partitioned into 186 nodes, 4 input/ output nodes. 186 nodes are shared memory nodes with 8 dual core power 5+ 1.9GHz processors each 180 of the computational noes have 32 GB memory each The code is parallelized using MPI SINTEF ICT 27

SINTEF ICT 28

Robustness ? SINTEF ICT 29

Validation strategy ? SINTEF ICT 30

ALTA Normal Flight path PILOTS REPORT: SINTEF ICT

Realistic Boundary condition to run offline simulations SINTEF ICT 32

Turbulence Intensity Contour (3) as a function of free stream speed SINTEF ICT

SINTEF ICT 34

www.ippc.no SINTEF ICT

SINTEF ICT 36

Automatic Wind Shear and Turbulence Alert System SINTEF ICT 37

Conclusion • Operational Multiscale Model • The prediction system confirms the experiences recorded in the pilots reports and gives possible explanations • The code has been validated extensively against wind tunnel data for cubes, hills, cylinders • There is a scarcity of data for the validation of numerical codes but flight data, wind farm data, weather station data can be used together to get better insight into the flow at microscales. • The data from the different sources can be used for fine tuning and validating the model SINTEF ICT 38

NORWAY IS STILL BEAUTIFUL SINTEF ICT 39