Validation of Coastal Wind and Wave Fields by High Resolution Satellite Data Susanne Lehner, Miguel Bruck, Andrey Pleskachevsky German Aerospace Center
SWH by CWAVE from Global ENVISAT ASAR Data J. Schulz-Stellenfleth et al., JGR, X. Li et al., IEEE Folie 2 2
Spatial variability of sea state (~300m) by TerraSAR-X images 5km Pico island, Azores TerraSAR-X Stripmap VV pol. February 9, 2011 19:41 Folie 3 Institut für Methodik der Fernerkundung bzw. Earth Deutsches Fernerkundungsdatenzentrum Observation Center 3
Spatial variability of sea state (~300m) by TerraSAR-X images XWAVE algorithm Hs=3.4 Lp=326m T=14.7s Hs=3.5 Lp=300m T=13.9s Hs=3.6m Lp=310m T=14.1s Hs=2.4m Lp=320m T=14.3s Pico island, Azores TerraSAR-X Stripmap VV pol. February 9, 2011 19:41 Folie 4 Institut für Methodik der Fernerkundung bzw. Earth Deutsches Fernerkundungsdatenzentrum Observation Center 4
Horizon mask Neustrelitz for TerraSAR-X and ENVISAT Near Real Time Services for TS-X in < 20 min in European MarSur Projects MARISS with EMSA, JRC, ESA, several European satellite providers Folie 5 5
Wind Field: XMOD Algorithm Geophysical Model Function GMF (Wind Roughness of Sea Surface) Connection between Radar NRCS, Wind Speed, Wind Direction and Incidence Angle SAR Image Streaks (2D-FFT, ), Interpolation SAR Image Calibration Wind Direction NRCS Inversion XMOD Wind Einfallswinkel Wind Incidence Flight Direction waves Angle Wind-Flight Folie 6 6
Development of the X-band GMF for sea surface wind field retrieval XMOD1, linear inversion 0 ( U10,, ) x0 x1u 10 x2 sin( ) x3 cos( 2 ) x4u 10 cos( 2 ) XMOD2, non-linear inversion, comparable to CMOD GMF U U,, ( A ( U,, ) cos n 0 10 n 10 ) 10 2 n 0 wind speed; incidence angle; angle between wind direction and SAR look direction XMOD1, Y. Ren et al, IJRS, tuning dataset from X-SAR (shuttle mission) XMOD2, X. Li et al, Spie 2011 (tuning onts-x data) Folie 7 7
Comaprison of Wind Speed Algorithms Folie 8 8
Case study over Alpha Ventus Alpha Ventus is the first offshore wind farm of Germany, located at the North Sea 58 m 90 m 5MW Wind turbine Folie 9 9
Case study over Alpha Ventus, Wind Farm North Sea TS-X Stripmap Aug.7, 2011 at 17:18 UTC VV polarization, spatial res. 3 m Folie 10 10
Sea surface wind field derieved from the TS-X data over Alpha Ventus XMOD2 XMOD1 Folie 11 11
Comparison of TS-X wind measurements to DWD (global) model DWD @ 18:00UTC Higher variability in TS-X data Folie 12 12
Wind Input K-Model wind+3m/s wind wind - 3m/s JONSWAP Area GKSS (HZG) buoy Lehner, IJRS, 2011 Hs GKSS- modelled wave by K-model rider buoy measurement Folie 13 13
XWAVE: Sea State Parameters from TerraSAR-X Images sub-scene signal intensity I I I( i, j) mean m mean S FFT (m) signal modulation m image spectra S TerraSAR-X SM Elbe estuary sub scene 2-D Imagespektrum im Wellenzahlraum E k min k max S( k, ) dkd filtering, L min, L max XWAVE function Parameters: H s, L p,t p 1D Spektrum im Frequenzraum buoy spectrum integrated image spectrum 100 m 50 m 2D image spectrum Folie 14 14
H s a * a 1 4.0 E(1.0 a2 cos ) 3 Statitistics Wave Length XWAVE Empirical Algorithm Collecting data set over NOAA buoys, Image data have to be ordered for acquisition about 2 days in advance Folie 15 15
XWAVE Algorithm: estimation of Significant wave height from TS-X data Hs based on analysis of image spectra and uses fitted parameters a 1 E sin (1.0 a 2 cos( )) 4 a 3 flight θ E - total energy calculted from the image spectrum in domain λ min and λ max. θ - the incidence angle α - peak wave direction relative to azimuth (satellite flight) direction a 1..a 3 - coefficients determined by TS-X,TD-X data fitting with wave height given by buoys and DWD (German weather center) wave model data. waves α Folie 16 16
Modelling of Sea State over Helgoland with K Model and boundary conditions from TS-X North Sea, Helgoland Lehner et al. IJRS, 2011 Topography BSH Folie 17 17
SWH from K-Model with boundary conditions from TS-X The wind field: XMOD boundary spectra include two parts: - swell (obtained using XWAVE) - wind sea (JONSWAP spectra) Boundaty spectra from TSX XWIND, XWAVE Folie 18 18
Application: Sea-State spatial Variability and Wind Field estimation using TerraSAR-X Synthetic Aperture Radar Data buoy sea state at wave power plant differs from the one at the buoy location Comparison of spectral shape parameters for image and wave spectra buoy collocated measurement s Independent measurements - Spatial variability is observed by TS-X imagery - Wind field estimated Folie 19 19
Example: Pico island, Sea State near Pico Wave Farm S(m2/Hz) 60 50 40 30 20 Buoy spectrum SAR spectrum 10 2d spectrum 0 0 0.05 0.1 0.15 0.2 frequency(hz) 1d frequency spectrum SM, VV polarization, February 14, 2011 M.Bruck et al, EWTEC, 2011 2011-02-14 Buoy 19:30 TS-X 19:41 H s (m) Tp (s) T 02 (s) T -10 (s) m ( ) 4.7 16.7 11.6 14.2 347 4.3 18.1 11.8 14.6 351 Folie 20 20
Pico island: TD-X Stripmap VV polarization. February 14, 2011 19:49, ASCAT time: 22:00 Wind vectors derived by the ASCAT Wind speed derived from TS-X data Folie 21 21
Determination of Underwater Bathymetry using Dispersion Relation Rottenest Island A. Pleskachevsky et al., Ocean Dynamics Folie 22 Institut für Methodik der Fernerkundung bzw. Earth Deutsches Fernerkundungsdatenzentrum Observation Center 22
Underwatertopograpy from TerraSAR SAR and Optical Grid size 150m Folie 23 23
Determination of Underwater Bathymetry Validation A.Plekachevky et al Oceandynamics, 2011 Folie 24 24
Summary Obtained information on the highly variable wind field and sea state in coastal areas using high resolution TerraSAR data Use information from refraction to obtain underwater bathymetry Methods Estimation by XMOD and XWAVE algorithm Comparison/Validation of wind field sea state parameters derived from TS-X SAR data with in-situ buoy measurements. Characterization of sea sate using integrated sea state and spectral shape parameters Goal Real Time Satellite Services for North Sea, Baltic Sea, Mediterranean in cooperation with DWD NRT Maritime Monitoring Services in European Framework, EMSA, JRC, Folie 25 FRONTEX 25
Thank you for your attention! Folie 26 26
ENVISAT Wind Sea State Oil Land-Water Line Bathymetry Sea Ice Folie 27 27
TerraSAR-X Nominal Imaging Modes StripMap Mode Resolution: 3 m 3 m Scene Size: 30 km 50 km [Range Azimuth] Coastal NRT < 20 min SpotLight Mode Resolution: 1 m 1,5... 3,5 m Scene Size: 10 km 5...10 km [Range Azimuth] Harbour, breakwaters, platforms ScanSAR Mode Resolution: 16 m 16 m Scene Size: 100 km 150 km [Range Azimuth] Ships, sea ice Folie 28 28
Pico island: Spotlight VV polarization, 27 March 2010 19:41 λ p p 256 322 m λ p p 244 m 330 Boxsize: 770 m Folie 29 29
Wind Field Algorithm XMOD 1.0 Incident angle [ ] Wind direction [ ] NRCS by XMOD dependent on Incidence angle and wind direction (at angle of 35 ). Y. Ren et al IJRS, PORSEC 2010 Folie 30 30
Implementierung aus SAR abgeleiteter Parameter in numerische Seegangsmodelle TerraSAR-X Szene Wind und Randwerte Topographie und Seegangsmodell Seegangsmodel Ergebnisse Randspektrum aus zwei Teilen: - Dünung (XWAVE) - Windsee (JONSWAP) Windfeld: XMOD Algorithmus Seegang: XWAVE Algorithmus Folie 31 31
Graciosa island: Stripmap VV polarization February 13, 2011 50 40 Buoy spectrum SAR spectrum S(m2/hz) 30 20 10 2 D spectrum 0 0 0.05 0.1 0.15 0.2 frequency(hz) 1 D spectrum 2011-02-13 Buoy 8:39 TSX 7:55 H s (m) T p (s) T 02 (s) T -10 (s) m ( ) 4.8 15.4 9.7 12.9 341 5.2 15.4 7.4 13.5 310 Folie 32 32