Backscatter Patch Test inter-comparison of systems using shared - PowerPoint PPT Presentation

Backscatter Patch Test – inter-comparison of systems using shared reference areas for testing, calibration, and quality assessment M. Roche, T. P. Le Bas, X. Lurton, K. Degrendele, L. De Mol, V. Van Lancker, M. Baeye, J. De Bisschop, C. Vrignaud, S. Papili, O. Lopera, J.M. Augustin, N. Le Bouffant, L. Berger & D. Eleftherakis

Monitoring the seabed = EU legal obligation The European Union Marine Strategy Framework Directive (“MSFD”) Six-year review Initial assessment, of the different objectives, targets elements “Coastal waters, and indicators of the strategy 2012 (+ 6 years) including their seabed and subsoil, 2018-2021 are an integral part of the marine environment, and as such should GES 2020 also be covered by this Directive…” Implementation Monitoring of the programs Marine Strategy 2014 2016 Programs of measures 2015 dura lex , sed lex…

Combined MBES Bathy + BS = the standard technology for mapping the seabed Bathymetry Backscatter Strength NE Malta – EM3002D – acquired April 2010 on RV Hercules Ref: Aaron Micallef, Tim Le Bas , Veerle Huvenne, Philippe Blondel, Veit Hühnerbach, Alan Deidun, (2012) A multi- method approach for benthic habitat mapping of shallow coastal areas with high-resolution multibeam data, Continental Shelf Research, 10.1016/j.csr.2012.03.008.

Combined MBES Bathy + BS = the standard technology for mapping the seabed Benthic Terrain Mapping and supervised backscatter texture analysis

MBES BS for monitoring the seabed nature variation? Yes, if… TIME Kwintebank West channel KWGS area, Belgium EM3002D 2005 -15 2012 BS 2011 dB -35 2009 0 0 500 m 500 m Mean dB level trend • � Sediment / Habitat change ? Sediment / habitat limits spatial change ? • Yes, if BS measurements are enough accurate and stable…

BS required accuracy for detecting sediment change trends average difference between the mean BS levels of the different sediment types ± 2 dB • � 1 dB (1/2 average difference) = accuracy required Plateau = most discriminant part = MBES BS PARADISE •

The plateau approach to quantify the mean BS level BS versus Incidence angle for different seafloor interfaces from various sites IFREMER dataset (RV Thalia EM2040 dual receiver @ 200 kHz) Restriction to sector 35°-45° � Loss of information (specular + fall-off) • Remains the most relevant BS data derived from only oblique insonification •

BS mean level stability ? BS mean level variance evaluation by repetitive measurements during a tide cycle: Tide level BS mean level same one line • 0 -15,5 area surveyed several times -1 (heading = cst) -15,6 -2 Beam averaged -15,7 • 1x1 m mosaic -3 -15,8 Stats on BS -4 • mosaic values -15,9 from the same -5 BS mean level -15.7 area and for std 0.1 incidence angle -6 -16 interval 35°-45° ~ 13 h Average level of BS is fully stable over a short period RV Belgica EM3002D

Sources of instability (out of the sonar equation) The Very Big Question: To what extent mean BS level variation from one cruise to another really represents significant change in seabed properties and not change due to the followings? Potential sources of instability Scientific literature ( non exhaustive list, see BSWG C3 and C5) Manufacturer references ± Changes of acquisition parameters on board ? Antenna state - biofouling ? Aging of antenna (electronic and coating) Water column influence? Influence of dense near-bed and water column suspensions on the MBES BS? MBES + acoustic doppler current profiler (ADP) measurements • Benthic landers + ADPs + acoustic backscatter sensor (ABS) � near-bed sand transport • Water samples to calibrate the optical and acoustical sensors •

Sea state? Beaufort 6-7 The Captain and a world-famous The Scientists… Belgian surveyor…

Sea state? BS time series on a sandbank area covered by very large dunes: calm sea rough sea EM1002 RV Belgica Negative offset of up to 3 dB of the mean dB levels If target is to monitor the seabed, stop the survey if the weather is too rough!

MBES BS reference area Check BS stability = mandatory • Seafloor reference area = most practical approach • Basic assumption = • Reference area seafloor cst against required accuracy of BS measurement ( ~ 1 dB) The ideal reference area: • Flat • Deep and large enough • Isotropic homogenous seafloor • Not subject to seasonal change • Easily accessible • Ground truthing data • Time series • Belgian BS reference area and RV Belgica II (futuristic vision inspired from the world's largest swimming pool of San Alfonso del Mar, Algarrobo, Chile)

MBES BS reference area Similar as the approach used to calibrate the SAR scatterometer Reference surface = Amazonian rain forest Used for years by satellite radars From C. H. Buck, Alternative large-scale distributed targets for SAR elevation beam pattern characterization, Tech. rep., ENVISAT project, ESA-ESTEC, 2001. Forest canopy: High reflectivity (-10 dB) • Isotropic angular response • Excellent stability over the year • (one more point for Brazil!)

BS patch test Goal = estimate the relative stability of the BS with a reference model Goal = estimate the relative stability of the BS with a reference model Reference model based on reference area BS data: Reference model based on reference area BS data: Ultra dense survey of a reference area at Ultra dense survey of a reference area at Time series of the reference area Time series of the reference area mean of all the 35°- 45° data mean of all the 35°- 45° data Time Zero using calibrated sensor Time Zero using calibrated sensor � BS reference model with 35°- 45° data � BS reference model with 35°- 45° data � BS reference model � BS reference model Relative stability estimation: Relative stability estimation: New survey 35°- 45° BS data are compared with the reference model New survey 35°- 45° BS data are compared with the reference model Gap statistics (mean and standard deviation) Gap statistics (mean and standard deviation) • • Relative Quality Level / Acceptability Thresholds remain to be defined Relative Quality Level / Acceptability Thresholds remain to be defined • • Under study approach Under study approach • • Analog to a IHO bathymetric quality test using a reference surface Analog to a IHO bathymetric quality test using a reference surface

Reference area In/around the Bay of Brest FR - SHOM - IFREMER areas Several reference areas • Depth 25 to 60 m – Flat + wrecks • Selected by SHOM for bathymetry trials • Candidates for becoming BS reference areas • E.g. Carré Renard: Depth ± 25 m • Coarse sediment • Long time series with HF MBES (EM 3002, EM 2040) •

Reference area UK - National Oceanography Centre area Western Solent area Teledyne Reson 7101 – 240 kHz 7 km², depth from 10 to 30 m MBES Teledyne Reson 8101 or 7101 • • sand + gravels surveys for 2008 to 2014 (1 per year) • strong tidal currents Plus some sidescan and ground truth • • non-mooring area (pipelines + cables) data. •

Reference area Belgian Teams Flemish sandbank KWGS area Bathymetry BS draped m m 22 22 22 26 26 26 2 km 2 km dB dB dB 6.5 km², depth from 22 to 26 m MLLWS • -15 -15 -15 sand + gravels • strong tidal currents • EM1002 (4 surveys from 2008 to 2011) • EM3002D (9 surveys from 2009 to 2015) • EM2040 (2 surveys from 2013 to 2014) • -25 -25 -25 Sidescan sonar (Remus – SS 900 kHz) • Ground truth data: SPI, video, grab samples… •

Inter comparison on reference area HV Ter Streep RV Belgica RV Simon Stevin EM3002D 300 kHz EM3002D 300 kHz EM2040 @ 300 kHz 21/09/2012 18/09/2012 18/09/2012 -15 BS Statistics on • common area dB -35 Along corridor • cross section 0 0 200 m 200 m

Inter comparison on reference area: results dB Statistics dB Corridor cross section -14 -10 Belgica EM3002D -15 -12 -14 -16 Ter Streep EM3002D -16 -17 -18 -18 -20 -19 -22 Simon Stevin EM2040 @ 300 kHz -24 -20 -26 -21 Belgica TerStreep Simon Stevin 0 100 200 300 400 500 600 m EM3002D EM3002D EM2040 @ 300 kHz � dB level from Ter Streep and Belgica EM3002D fully comparable � EM2040 @ 300 kHz not comparable with EM3002 300 kHz ( ∆ 2-3 dB)

Sharing BS reference area: RV Belgica C1516 – 10-16/06/2015 • Collegial Survey of the 3 BS reference areas • MBES EM3002D • SAS Shadows • Video and Boxcore Shared BS reference data library • 0 0 0 100 km 100 km 100 km

Concluding message

Thank you for your attention! Special thanks to the BSWG Chairing Committee and Editors. Os meus sincero agradecimentos aos colegas brasileiros pela vossa hospitalidade em Salvador da Bahia, cidade de todos os Santos e dos grandes Jorge Amado e Dorival Caymmi.