DEMOSS. Title: Development of Marine Oil Spills/slicks Satellite monitoring System elements for the Black Sea, Caspian Sea and /Kara/Barents Seas

Transcription

1 DEMOSS Title: Development of Marine Oil Spills/slicks Satellite monitoring System elements for the Black Sea, Caspian Sea and /Kara/Barents Seas INTAS Thematic Call on Earth Sciences and Environment in cooperation with with ESA, 2006 by Stein Sandven 1, Vladimir Kudriavtsev 2 and Vladimir Malinovsky 3 1 NERSC, Bergen, Norway 2 NIERSC, St. Petersburg, Russia 3 MHI, Sevastopol, Ukraine With contribution form the other DEMOSSS partners

2 Partners 1. Nansen Environmental and Remote Sensing Center (NERSC), Norway 2. BOOST Technologies, Brest, France 3. University of Hamburg, Hamburg, Germany 4. Nansen International Environmental and Remote Sensing Center (NIERSC), St.Petersburg, Russia 5. Institute of Applied Physics Russian Academy of Sciences (IAP), Nizhny Novgorod, Russia 6. Marine Hydrophysical Institute of the Ukrainian National Academy of Sciences, Sevastopol, Ukraine 7. Arctic and Antarctic Research Institute (AARI), St.Petersburg, Russia 8. Research Center for Earth Operative Monitoring (NTs OMZ), Moscow, Russia

3 Project Objectives To develop and demonstrate components of a marine oil spill detection and prediction system based on satellite SAR and other space data in combination with models for oil slick/spill monitoring and prediction

4 Overview of Tasks Task 1 Task 2 Task 3 Task 4 Task 5 Field experiment with oil slicks in the Black Sea, IAP/MHI Radar Imaging Model Development, NIERSC Algorithm for detection & quantification of oil spills and look-alikes, NIERSC Satellite monitoring of selected areas: * Barents/Kara Sea, NIERSC * Black Sea, MHI * Caspian Sea, NTsOMZ and validation of oil slick detection, BOOST Oil spill modelling and drift forecasting, AARI

5 SAR acquisition of the study areas 1 2 Region Total* Details 3 The Barents Sea (1) and Kara Sea (2) are relatively clean areas with little ship traffic and offshore exploration has just started. The areas are expected to become much more exposed to oil pollution in the future. The Black Sea (3) and the Caspian Sea (4) have already significant tanker traffic and offshore exploitation has started from several platforms 4 Barents Kara Black Caspian WSM: 131 APM: 12 IMM: 240 WSM: 276 APM: 3 IMM: 168 WSM: 58 APM: 38 IMM: 86 WSM: 38 APM: 1 IMM: 52 * Number of image obtained from ESA rolling archive from May to December In addition, archived data from earlier years are available for the studies.

6 Radar scattering modelling DEMOSSS develops an improved model of radar scattering from a sea surface covered by oil and biogenic films to be used in detection and classification of surface film in SAR images Flow diagram of the radar scatter model for simulation of a given surface condition

7 Wind Waves Spectrum and Effect of Thin OLE Film 10 2 OmniDirectional 10 2 Up Wind Direction Oleic adic (OLE): monomolecular film B(k) 10 4 u10=6m/s OLE E=0.022 B(k,0 0 ) k, rad/m k, rad/m Contrast, db 20 Contrast, db k, rad/m k, rad/m

8 Backscatter from clean and film-covered water in tank experiments (inc. angle 30 ) Blue dots: observed backscatter from clean water Blue circles: observed backscatter from oil films (also triangles and crosses) Blue line: modelled backscatter from clean surface Green line: modelled backscatter from surface film (Ref. Gade at el.,jgr 1998, Kudriavtsev et al, JGR, 2005)

9 Spectral Contrasts for different surface films: Comparison of models with data Data from tank experiments (Ermakov et al.) 10 2 u10=7m/s Model simulations OLE Vegetable oil Crude oil Diesel oil Spectral Contrast 10 1 OLE E=0.022 VO E=0.012 CO E=0.004 Wavenumber k, rad/cm k, rad/m Contrast between wind wave spectrum for clean water and different surface films

10 Effective Oil Film Viscosity Experimental estimates by Ermakov et al. vs. Jacobs and Jenkins (1997) model Wave damping coefficients as function of film thickness 15 Hz waves 25 Hz waves Oil thickness in mm Oil thickness in mm

11 Up-wind Radar contrasts vs. oil film thickness at C- and X-band VV&HH contrast in db C band θ=20 0 u 10 =5m/s VV&HH contrast in db X band θ=20 0 u 10 =5m/s Film Thickness [m] Film Thickness [m] VV&HH contrast in db C band θ=20 0 u 10 =10m/s VV&HH contrast in db X band θ=20 0 u 10 =10m/s Film Thickness [m] Film Thickness [m]

12 Analysis of oil spill signatures in SAR images NRCS Incidence angle direction Wind speed

13 Comparison of observed (from SAR archive) and modelled C-band backscatter contrasts in oil spill signatures C band Oil Slicks Contrast, db Wnd Speed, m/s

14 Modelled backscatter of surfactants in an eddy Simulated NRCS field (in db) for an eddy current field in presence of surfactants. Wind speed (a) 5 m/s and (b) 15 m/s. Radar geometry is for ERS SAR.

15 Field experiment from an offshore tower in the Black Sea Optical system to measure short wave spectrum and surface mean slope Video system to measure wave breaking characteristics

16 Optical Spectrum Analyzer X-band radar Ka-band radar

17 Experiments with slick observations from the tower Date, start time : , 16:47 Wind direction No wind 145º (SE) 90º (E) Wind speed Z=4m 0 m/s <0.5 m/s (14:07) 2.5m/s? Wave Vector dir 260º 270º Slick observations VO (13:03) VO (14:15) OLO (15:07) Natural slick (17:12) DF (17:14) Artificial slicks: Veg.Oil, Olive oil, Dodecyl Alcohol, Diesel fuel. Total: , 10: , 15:46 110º No wind 107º (11:43) 90º (12:28) 0 m/s 1-2 m/s 2-3 m/s 1-2 m/s 270º 280º Natural slick (12:00) Natural slick (12:42) Natural slick (12:54) Natural slick (12:59) + VO Natural slick (13:13) OLO (13:43) Natural slick (15:50) Natural slick (16:03) OLO (16:09) Natural slick (16:14) Natural slicks periodic and single bands Total number of slicks: , 10:34 120º 0-2 m/s 300º Natural slick (12:37) , 15:46 150º 1-3 m/s 330º Natural slick (17:17) , 10: , 10:02 0º - 5º 70º 2-4 m/s? 2-4 m/s? 270º 250º VO (12:22) Natural slick (12:30) OLO (12:44) DA (13:00) VO (13:39) VO (15:39) VO (15:50) DA (16:09) OLO (16:23) VO (16:23) OLO (11:55) Periodic Natural slicks ASAR image

18 Contrasts in slicks observed on 05 Oct OSA 10 Radars Photo Contrast Ka-band X-band Contrast OSA Wavenumber, rad/cm Wavenumber, rad/cm Dodecyl alcohol slick (film elasticity E=50-70 mn/m) Vegetable oil slick (film elasticity E=12-15 mn/m) Wind velocity 2-4 m/s

19 SAR observation of experimental oil spills

20 NRCS profile across a slick observed in ASAR APP data Results from SAR analysis of AP images form 2003: The largest slick has a contrast of about 15 db compared to the surrounding clean water

21 Another slick observation in SAR APP image Subset of ENVISAT ASAR AP image on 23 August 2003 off Novorossisk coast: (a) VV-pol (b) HH-pol (c) Pol ratio (PR) For clean seas PR is defined by contribution of bragg scattering and wave breaking, with typical value of 5 for this inc angle. In slicks bragg waves disappear and PR becomes close to 1

22 Distribution of oil spills in the Black Sea derived from 68 SAR images SAR images from ERS-2 and ENVISAT were analyzed for a period of three years ( ), resulting in 68 images with 424 likely oil spill events. The distribution of the spills are concentrated along the main shipping lanes and in the offshore drilling area in the western Black Sea

23 Example from the Caspian Sea The ASAR Wideswath image from 04 July 2007 covers most of the Caspian Sea (left figure). A subset of the image (above) was analyzed for the area off Baku (red circle) where a spill event could be detected. The SARTool provided by BOOST Technologies was used to detect and quantify the oil spill area.

24 Oil spill event Kerch Strait 11 November 2007 C-band: RADARSAT X-band: TerraSAR L-band: ALOS PALSAR Courtesy: Scanex Courtesy: DLR Images obtained 16 November - > case study for model comparison

25 Comparison with previous SIR-C/X SAR data and field experiments

26 Oil drift modelling Oil spill input oil spill location oil spill volume and spill rate oil properties fractional composition of the oil Oil spill simulation spreading of the spillets advection evaporation advection turbulent diffusion evaporation emulsification dispersion photo-oxidation bio-oxidation Sea state input currents wind wind waves thermohaline structure bathymetry ice conditions AARI is developing an oil drift model, OilMARS, based on the components shown in the diagram. The model has been tested in the Barents and Kara Seas. Oil spill output Oil slick spatial distribution Oil mass balance

27 Oil spill modelling in Kara Sea: open water AARI uses its oil spill model OilMARS to simulate oil drift in the Kara Sea. The figures show the spreading of a spill over a period of 20 days. The red area indicate where oil reached the coast and caused pollution at the beach.

28 Oil spill modelling in Kara Sea: sea ice waters AARI uses its oil spill model Oilmars to simulate oil drift in the Kara Sea. The figures show the spreading of a spill over a period of 20 days in winter when the sea ice icecovered. Black indicates oil spill in open water, blue indicate oil spill on top of th eice and red is oil spill under the ice.

29 Summary and further work Radar scattering modelling tools is ready for use Field experiments with artificial oil spills at the tower in the Black Sea were performed in 2007, more experiments are planned in 2008 Lab experiments with radar observation of wave damping by various oil types have been conducted Build-up of SAR data base for the study regions have started, primarily with ASAR data. Will be supplemented by other SAR data (X- and L-band) Analysis of SAR data for slick and other ocean surface features, including comparison with models has started Verify hypothesis that PR can be used to identify oil and natural slicks and discriminate them from look-alikes Establish monitoring scheme using satellite data in combination with models and in situ data for validation

30 Acknowledgement The SAR data for the study is provided ESA through AOBE-2780) The research is supported by INTAS (contract no ), EU FP6 (contract no MONRUK), and national projects