Observing Dry-Fallen Intertidal Flats in the German Bight Using ALOS PALSAR Together With Other Remote Sensing Sensors

Transcription

1 Observing Dry-Fallen Intertidal Flats in the German Bight Using ALOS PALSAR Together With Other Remote Sensing Sensors Martin Gade, Institut für Meereskunde & Kerstin Stelzer Brockmann Consult

2 Outline Background & Basics DeMarine-U Field Campaigns SAR Image Examples Summary & Outlook

3 Background & Basics Intertidal Flats Distance of about 10 km offshore German and Dutch North Sea coast, S Korean coast Fall dry once during each tidal cycle Only partly vegetated Consist of fine sediments (sand, mud) Impacted by the stress of a changing world (e.g. sea level rise) Frequent Surveillance Important because of unique ecosystem and high morpho-dynamics Regulated through national and international laws Difficult task by boat, foot, or land vehicles Remote Sensing Optical data yield promising classification results But: strong dependence on cloud cover Include radar data into existing observing/classification systems

4 Background & Basics Surface depends on sediment types and water level...

5 Background & Basics ( 12 km x 12 km ) L C X 10 April 1994, 0804 UTC, 2h45 after low tide Spaceborne Imaging Radar-C / X-Band SAR (SIR-C/X-SAR) Campaigns in 1994 Different radar signatures at L-band and C-/X-band!

6 Background & Basics Integral Equation Model (IEM; Fung et al. [1992]) NRCS isolines as function of correlation lenth and rms height (statistical properties of sediment surface) Gaussian autocorrelation of surface elevation

7 Background & Basics Deriving Maps of RMS Height and Correlation Length Example: NRCS (pixel) values at VV polarisation: L band: -18 db C band: -9 db X band: -11 db Derived roughness parameters: rms height: 3.1 mm corr.length: 3.7 cm

8 Background & Basics Sediment Map Derived Using SIR-C/X-SAR Data (50 m geometrical resolution) Source: National Park Agency Dependence of sediment type and micro particles after Pröber [1981]

9 DeMarine-Umwelt (Environment) Subproject (TP) 1 Coordination, Users Office TP 2 TP 3 TP 4 TP 5 Water Quality Drift Forecast Integration Optics/SAR Data Assimilation DeMarine-Umwelt, TP 4: Integration of Optical and SAR EO Data and in-situ Data into Wadden Sea Monitoring Duration: Februar 2008 Januar 2011 Partners:

10 Concept of TP4 In-Situ Data Optical Data Radar Data Synergistic Classification System In-Situ Campaigns Validation Optimized Wadden Sea Classification & Habitat Mapping

11 DeMarine-U Regions of Interest

12 RS Data Sources Optical Sensors Sensors already used (HIMOM, OFEW) Landsat ASTER SPOT Sensors to be included Rapid Eye IRS IKONOS EnMap Radar Sensors L-Band PALSAR C-Band ERS-1/-2 SAR ENVISAT ASAR RADARSAT X-Band TerraSAR-X

13 Field Studies

14 In-Situ Protocols Parameters protocolled at every in-situ station: weather conditions sediment type / cover surface structure ripple length & orientation colour diatoms macrophytes macrofauna meta information

15 Field Campaign - Optics DeMarine-TP4 April 2008 HyMap Overpass 2006 within OFEW Project

16 Field Campaign - Radar DeMarine-TP4 April 2008 PALSAR Scene, 12 April 2008, 2143 UTC (23 min. after low tide (LT))

17 ALOS PALSAR 12 April 2008, 2143 UTC Wind 7-8 m/s (SSW), LT 2120 UTC SAR Image Examples PALSAR ASAR Difference ENVISAT ASAR 13 April 2008, 1001 UTC Wind 5-6 m/s (SW), LT 0935 UTC

18 ALOS PALSAR 12 April 2008, 2143 UTC Wind 7-8 m/s (SSW), LT 2120 UTC SAR Image Examples PALSAR ASAR Difference ENVISAT ASAR 13 April 2008, 1001 UTC Wind 5-6 m/s (SW), LT 0935 UTC

19 SAR Image Examples ENVISAT ASAR 27 July 2008, 1001 UTC Wind 2-3 m/s (ENE), LT 1022 UTC Envisat ASAR ERS SAR ERS SAR 27 July 2008, 1031 UTC Wind 2-3 m/s (ENE), LT 1022 UTC

20 Summary & Outlook Use of multi-frequency satellite SAR data to derive surface roughness parameters of intertidal flats Integration into improved monitoring and classification system Inversion of Integral Equation Model [Fung et al. 1992] Based on previous studies (1994/1998) DeMarine-U: systematic analyses of SAR data from multiple satellites routine in-situ campaigns (record surface parameters) include SAR data in (routine) monitoring of intertidal flats extend IEM inversion process towards use of data from multiple satellites

21 Thanks!

22

23 Previous Studies (3) Integral Equation Model (IEM; Fung et al. [1992]) where and Backscattering of electromagnetic waves from a dielectric surface with random roughness modulation Limiting cases: Bragg model and Kirchhoff model Special form of surface autocorrelation function important

24 Field Campaign - Radar DeMarine-TP4 April 2008 TerraSAR-X Scene, 30 August 2008, 1710 UTC (34 min. after low tide)

25 ENVISAT ASAR 18 Oktober 2007, 0955 UTC Wind m/s (NNW), LT 0955 UTC SAR Image Examples PALSAR ASAR Difference ALOS PALSAR 18 Oktober 2007, 1023 UTC Wind m/s (NNW), LT 0955 UTC

26 SAR Image Examples ENVISAT ASAR 18 Oktober 2007, 0955 UTC Wind 12 m/s (NNW), LT 0945 UTC PALSAR ASAR Difference ALOS PALSAR 18 Oktober 2007, 1023 UTC Wind 12 m/s (NNW), LT 0945 UTC

27 Combination Optics - Radar Landsat PALSAR