Change detection in cultural landscapes

Transcription

1 9-11 November 2015 ESA-ESRIN, Frascati (Rome), Italy 3 rd ESA-EARSeL Course on Remote Sensing for Archaeology Day 3 Change detection in cultural landscapes DeodatoTapete (1,2) & Francesca Cigna (1,2) (1) (2)

2 Synthetic Aperture Radar (SAR) Frequency: GHz Wavelength: 1 mm 1 m Spatial resolution: up to 0.25 m weekly monthly Repeat cycle: A, amplitude SAR amplitude ϕ, phase Strength of the radar signal backscattered to the sensor; can be derived by combining real and imaginary components of the complex radar signal A= (I 2 + Q2 ) The SAR amplitude can be transformed into physical units, e.g. σ0 (sigma nought), the strength of the radar signal backscattered from a distributed target, normalised to the ground range plane (ellipsoid model) Digital Number; i.e. the radar amplitude in GRD products 2 DN σ0 = 2 Aσ Calibration factor, f(eap, RSL, α)

3 SAR backscattering σ i0 = f (λ, pol., θ, roughness, shape, dielectric properties) wavelength incidence angle polarisation properties of the scattering surface Surface roughness A breakpoint between smooth and rough surfaces (and the corresponding dominant radar scattering mechanisms) RMS height of the microrelief Rayleigh criterion λ h < 8cosθ

4 RGB colour composition 04/02/ /04/ /06/2003 RGB a a b b CIGNAet al Amplitude change detection with ENVISAT ASAR to image the cultural landscape of the Nasca region, Peru. Archaeological Prospection 20(2),

5 Image ratioing 15/11/2005 Ratio Rσ 0 i 04/02/ < R < 1 if Rio Ingenio plain R > 1 if P4 P5 CIGNA et al Amplitude change detection with ENVISAT ASAR to image the cultural landscape of the Nasca region, Peru. Archaeological Prospection 20(2), Decreased radar backscattering σ i0 (t k ) = 0 σ i (t j ) σ i0 (t j ) > σ i0 (tk ) σ i0 (t j ) < σ i0 (tk )

6 Image ratioing 15/11/2005 Ratio Rσ 0 i 04/02/2003 Rio Ingenio plain 0 < R < 1 if σ i0 (t k ) = 0 σ i (t j ) σ i0 (t j ) > σ i0 (tk ) R > 1 if 0 If σ i is in db e.g. R = 0 db R = 10 db R = 20 db R = 30 db Rσ 0 [db] = σ i0 (tk ) σ i0 (t j ) i R = 100/10 = 1 R = 1010/10 = 10 R = 1020/10 = 100 R = 1030/10 = 1000 σ i0 (t j ) < σ i0 (tk )

7 Change detection in cultural landscapes TAPETEet al Prospection and monitoring of the archaeological heritage of Nasca, Peru, with ENVISAT ASAR. Archaeological Prospection 20(2),

8 9-11 November 2015 ESA-ESRIN, Frascati (Rome), Italy 3 rd ESA-EARSeL Course on Remote Sensing for Archaeology Day 3 Tutorial 1: Intensity change detection analysis DeodatoTapete (1,2) & Francesca Cigna (1,2) (1) (2)

9 Tutorial 1: Intensity change detection analysis OBJECTIVE The scope of this tutorial is to familiarise with how to exploit Sentinel-1A SAR data for landscape archaeology. INPUT DATA 3 GIS-ready, high resolution GRD (ground range detected) scenes, acquired over the Nasca region (Peru), in Interferometric Wide swath (IW) mode and with VV polarisation. October 2014 December 2014 March 2015

10 Tutorial 1: Intensity change detection analysis LEARNING OUTCOMES 1) how to manipulate Sentinel-1 GRD products 2) how to perform a simplified radiometric calibration to derive the radar cross-section of distributed targets and convert to the db scale 3) how to combine multi-temporal data using RGB and RC colour composites 4) how to derive ratios between pairs of SAR scenes. HOW TO FOLLOW THE INSTRUCTIONS (in the handouts) Back text:common to ArcGISand QGIS, or to be performed with other software Blue text: ArcGIS/ArcMap-specific Orange text: QGIS-specific

11 40 min 10 min 10 min Tutorial 1: Intensity change detection analysis STEPS 1) Creation and setup of GIS project (ArcGISor QGIS) 2) Coordinate system selection 3) Data unzipping into the training folder 4) Loading of Sentinel1 geotiffs in the GIS project 5) Radiometric calibration 6) Conversion to db scale 7) Clipping to area(s) of interest 8) Combination into multi-band layers 9) RGB colour composition 10) RC colour composition 11) Image ratioing 12) Adaptation of symbology 13) Interpretation of landscape changes Questions & discussion Questions & discussion Questions & discussion

12 9-11 November 2015 ESA-ESRIN, Frascati (Rome), Italy 3 rd ESA-EARSeL Course on Remote Sensing for Archaeology Day 3 Multi-temporal InSAR for structural monitoring and condition assessment DeodatoTapete (1,2) & Francesca Cigna (1,2) (1) (2)

13 Multi-temporal SAR Interferometry (InSAR) ϕ, phase Average LOS deformation rate (precision ~ mm/yr) Displacement time-series (precision ~ 1-3 mm on single measure)

14 Persistent Scatterer Interferometry (PSI) Requirement: targets on the ground with persistent reflectivity properties in time Persistent Scatterer(PS) PSInSAR technique (Ferretti et al., 2001) Pros: Suitable over urban/built-up areas Inference of ground instability, regional/local land processes Cons: No data or low PS density over vegetated areas Significant alteration of the scene constraint for PS generation Rocca (2013)

15 Recent PSI and SBAS techniques Scope: Increase the density of measurement points, allowing investigation over bare, vegetated, rural sites SqueeSAR technique (Ferretti et al., 2011) Evolution of PSInSAR technique TRE website Intermittent Small Baseline Subset (ISBAS) method (Bateson et al., 2015)

16 PSI application on archaeological sites TAPETEet al. (2012)Satellite radar interferometry for monitoring and early-stage warning of structural instability in archaeological sites.journal of Geophysics and Engineering,9, S10 S25.

17 Site-scale InSAR monitoring (urban) Processing of COSMO-SkyMed SM HIMAGE time series CIGNAet al.(2014)persistent ScattererInterferometry Processing of COSMO-SkyMedStripMapHIMAGE Time Series to Depict Deformation of the Historic Centre of Rome, Italy.Remote Sensing 6(12):

18 Site-scale InSAR monitoring (rural) TAPETE& CIGNA(2012)Rapid mapping and deformation analysis over cultural heritage and rural sites based on Persistent Scatterer Interferometry. International Journal of Geophysics, 19 pp.

19 Monument-scale InSAR monitoring (urban) CIGNAet al. (2012)Ground instability in the old town of Agrigento (Italy) depicted by on-site investigations and Persistent Scatterersdata. Nat. Hazards Earth Syst. Sci. 12:

20 Monument-scale InSAR monitoring (rural) TAPETE& CIGNA(2012)Rapid mapping and deformation analysis over cultural heritage and rural sites based on Persistent Scatterer Interferometry. International Journal of Geophysics, 19 pp.

21 9-11 November 2015 ESA-ESRIN, Frascati (Rome), Italy 3 rd ESA-EARSeL Course on Remote Sensing for Archaeology Day 3 Tutorial 2: InSAR for structural monitoring DeodatoTapete (1,2) & Francesca Cigna (1,2) (1) (2)

22 Tutorial 2: InSARfor structural monitoring OBJECTIVE The scope of this tutorial is to familiarise with Persistent Scatterers (PS) data generated by processing a long stack of SAR imagery by means of a multiinterferogram InSAR algorithm. INPUT DATA ERS-1/2 PS data covering the period over the city of Rome, processed in the framework of ESA GMES Terrafirma project.

23 Tutorial 2: InSARfor structural monitoring LEARNING OUTCOMES 1) how to detect patterns of surface deformation 2) how to identify area(s) of potential instability 3) how to correlate spatially the observed motions with: (i) distribution of the monuments, their condition and conservation history (ii) local environmental and geological setting HOW TO FOLLOW THE INSTRUCTIONS (in the handouts) Back text:common to ArcGISand QGIS, or to be performed with other software Blue text: ArcGIS/ArcMap-specific Orange text: QGIS-specific

24 Tutorial 2: InSARfor structural monitoring STEPS 10 min 1) Creation and setup of GIS project (ArcGISor QGIS) 2) Coordinate system selection 3) Addition of basemap via WMS service 4) Loading of ERS-1/2 PS data 5) Analysis of the attribute table Questions & discussion 15 min 10 min 10 min 5 min 6) Classification by velocity 7) Analysis of topography and geology 8) Correlation with monument conditions 9) Time series analysis Questions & discussion Questions & discussion Questions & discussion Questions & discussion