SAR Sensors Acquisition Geometries The Interferometric Concept Fabrizio Novali Tele-Rilevamento Europa Thursday, 2 December 2010 COPYRIGHT - Tele-Rilevamento Europa 2010 Copyright - Tele-Rilevamento Europa - 2004
SAR Sensors COPYRIGHT - Tele-Rilevamento Europa 2010 2
Satellite SAR: A Coherent Sensor A synthetic aperture radar (SAR) works Radar satellite R by illuminating the Earth with a beam of coherent microwave radiation such as a laser. Phase information is essential for SAR data processing. X-band 3 cm C-band 6 cm L-band 24 cm Earth COPYRIGHT - Tele-Rilevamento Europa 2010 3
Why Radar Remote Sensing? 1. It s an active system: it does not require external illumination sources (Sun light) like optical systems 2. It operates at microwave frequencies: they can penetrate through the clouds 3. It is a coherent system: it allows to precisely measure changes in the satellite-target distance by means of interferometry COPYRIGHT - Tele-Rilevamento Europa 2010 4
SAR vs. Optical data SAR Multi Image Reflectivity map (MIR) Optical image COPYRIGHT - Tele-Rilevamento Europa 2010 5
Milan COPYRIGHT - Tele-Rilevamento Europa 2010 6
X-band Resolution C-band What SAR Sensors are Available? Historical Analysis Monitoring Repeat Time (days) ERS 1+2 35 Envisat 35 RADARSAT-1 24 RADARSAT-2 24 COSMO SkyMed 8 TerraSAR-X 11 1992 1995 2001 2002 2008 Present Day Future COPYRIGHT - Tele-Rilevamento Europa 2010 7
Geometry of Acquisitions COPYRIGHT - Tele-Rilevamento Europa 2010 8
Acquisition Geometry 7,5 km/s orbit 23 ~ 800 km ground range 100 km azimuth COPYRIGHT - Tele-Rilevamento Europa 2010 9
Descending orbits N Rome S COPYRIGHT - Tele-Rilevamento Europa 2010
Ascending orbits N Rome S COPYRIGHT - Tele-Rilevamento Europa 2010
Satellite Images Distance from Radar (slant range) H ground range COPYRIGHT - Tele-Rilevamento Europa 2010 12
Ground Range to Slant Range A B C D E F G H I J K L M N O P Ground Range COPYRIGHT - Tele-Rilevamento Europa 2010 13
Geographic and SAR Coordinates Geographic coordinates SAR coordinates A z i m u t h Slope + Slope - A z i m u t h Slope + Slope - Ground range Slant-range foreshortening COPYRIGHT - Tele-Rilevamento Europa 2010 14
Examples of Geometric Distortions FORESHORTENING LAYOVER, SHADOWING COPYRIGHT - Tele-Rilevamento Europa 2010 15
Interferometric Concept COPYRIGHT - Tele-Rilevamento Europa 2010 16
The Basic Interferometric Principle Satellite passes over an area of interest (AOI) Radar signals are reflected back to the satellite from high reflectivity ground objects The distance from the satellite to the ground object is represented here as R1 R 2 R 1 Ground deformation occurs The satellite re-passes over the same area Reflecting object ΔR The difference between R1 and R2 is the displacement (ΔR) along the satellite line-ofsight (LOS) Ground displacement is measured COPYRIGHT - Tele-Rilevamento Europa 2010 17
InSAR for Displacement Detection The unit of length used in InSAR is the wavelength: R1 R2 R1 R2 DR = c Df Df Interferometry ΔR = 5.66 cm D r R 1 R 2 COPYRIGHT - Tele-Rilevamento Europa 2010 18
SAR Data A SAR image is a set of pixels characterized by both amplitude and phase values Amplitude Phase (known modulo 2p) The phase is related to the sensor-target distance R COPYRIGHT - Tele-Rilevamento Europa 2010 19
Single SAR image: Phase Contributions f 4p r Amplitude Y Reflectivity of the radar target 4pr/ Propagator : it depends on the sensor-radar target distance Atmospheric Phase Contribution Noise Phase COPYRIGHT - Tele-Rilevamento Europa 2010 20
Interferogram: phase contributions If: 4p Df D Dr D noise nothing has changed: D = D = 0 for high SNR: = 0 4p f Dr The interferometric phase is proportional to range variations COPYRIGHT - Tele-Rilevamento Europa 2010 21
Interferogram: Phase Contributions f f flat f elevation f displacement f atmosphere f noise Geometrical and/or temporal decorrelation B n 4p r R tan 4p B n R 0 Dq sin 4p d D COPYRIGHT - Tele-Rilevamento Europa 2010 22
InSAR for DEM reconstruction Bn COPYRIGHT - Tele-Rilevamento Europa 2010 23
Differential Interferogram Generation Interferogram - = Synthetic Interferogram generated from a DEM Differential Interferogram Df f ˆ f topo motion DEMerror noise atmosphere COPYRIGHT - Tele-Rilevamento Europa 2010 24
Limits of Conventional DInSAR Analysis Thanks to ESA archive, since 1993 a growing gallery of examples of DInSAR interferometry started being available. While more and more InSAR were generated, the presence of atmospheric artefacts and problems due to phase decorrelation (temporal and/or geometrical) became more and more evident COPYRIGHT - Tele-Rilevamento Europa 2010 25
Limits of Conventional DInSAR Analysis 4p Df D Dr D noise Amplitude data 1-day (Tandem) interferogram D 0 15-months... D 0 COPYRIGHT - Tele-Rilevamento Europa 2010
Atmospheric Effects If the propagation medium changes in the time interval between two SAR acquisitions (e.g. humidity, temperature, pressure ), an additive phase term appears. Amplitude data Interferogram COPYRIGHT - Tele-Rilevamento Europa 2010 27
Atmospheric Effects Two additive phase term appears: (1) topography dependent (2) due to turbulence phenomena 1-Aug-95 / 2-Apr-96 2-Aug-95 / 3-Apr-96 COPYRIGHT - Tele-Rilevamento Europa 2010 28
Time PSI Understanding Displacement Satellite images are analysed using a multi image approach Images are acquired each time the satellite passes over the same area of ground As interferograms are acquired, they are stacked Consistent PS are identified in the data Individual displacement time-series is calculated for each PS Time-series of PS Stack of interferograms COPYRIGHT - Tele-Rilevamento Europa 2010 29
COPYRIGHT - Tele-Rilevamento Europa 2010 30
Close up on the cemetery COPYRIGHT - Tele-Rilevamento Europa 2010 31
Measurements along the LOS Incidence angles are often very low high sensitivity to vertical displacement θ θ PS PS d PS d real d real d PS COPYRIGHT - Tele-Rilevamento Europa 2010 32
Ascending and Descending: an example Ascending Vert Descending V D >0 East LOS Desce V A <0 LOS Asce COPYRIGHT - Tele-Rilevamento Europa 2010
The importance of the reference point All PSI measurements are differential measurements with respect to the reference point COPYRIGHT - Tele-Rilevamento Europa 2010 34
Synoptic Table Advantages Limitations Regular updates over large areas Cost-effective High PS density (up to 1,000 PS/km 2 ) Fast data processing / low user interaction High accuracy Data easily imported in GIS Vegetated and forested areas prevent any PSInSAR data processing. Low-reflectivity areas (e.g. smooth surfaces and certain materials). Temporal sampling limited by satellite repeat-cycles Only slow deformation can be measured (<10 cm/yr in LOS) COPYRIGHT - Tele-Rilevamento Europa 2010 35