THE FEASIBILITY AND APPLICATION OF PSI TO DETECT A RANGE OF GROUND AND STRUCTURE MOTION PHENOMENA.

Transcription

1 THE FEASIBILITY AND APPLICATION OF PSI TO DETECT A RANGE OF GROUND AND STRUCTURE MOTION PHENOMENA. Rachel Holley, InSAR Surveying Consultant InSAR Surveying Team Fugro NPA Satellite Mapping

2 Introduction Presentation will discuss: What is a feasibility assessment? What factors do we need to consider to decide if InSAR is suitable? Look at 3 Terrafirma Case Studies: Vilnius in Lithuania, Liege in Belgium, Esbjerg in Denmark

3 PSI Persistent Scatterer InSAR (PSI): Utilises more than 30 SAR images acquired as far back as 1992 Detection of Persistent Scatterers e.g. buildings, bridges and other structures Height of PS calculated at each image benchmark Historical motion revealed and time series of motion produced

4 PSI Capabilities: Unique ground motion information service Remotely map historical ground and structure motion (archive images from 1992) Monitor ground and structure motion (programme SAR images) Millimetric precision (as shown during the Terrafirma validation) Wide-area results (up to 100km x 100km) Dense PS networks (over 400 PS/km 2 in urban areas) PS motion history (individual time series) Global Coverage Virtual GPS network (higher density and better in vertical domain) Complementary information source Constraints/Limitations: Location specific (best suited to urban and semi-urban areas) Data quantity (at least 15 SAR images, more data = more precise) Temporal distribution of data (even, frequent distribution of SAR imagery) Motion magnitude (motions ranging from mm to 10cm/year) PS point locations (location serendipitous) Non-linear motion (significantly non-linear ground motion may not be captured)

5 Market Sectors and Applications Opportunities for input into the following markets: Civil Engineering Environment & Geohazards Oil, Gas & Mining Utilities & Telecommunication Legal Media Diverse range of applications: Urban stability, tunnel settlement Flood risk and mitigation (dams, levees, dykes, embankments etc.) Volcanic and seismic events, landslides Oil / gas reservoir stability, Mine subsidence and gas storage Asset management

6 Feasibility assessments Feasibility assessments: Establish which InSAR technique is most applicable Establish if it meets its users needs Key questions for PSI feasibility Motion Direction Topography Groundcover Climate Motion Magnitude Motion Spatial Extent Data Availability

7 Motion Direction

8 Topography SAR image distortion Layover Foreshortening Shadow DEM s are used to assess the topography and decide which pass is most suitable (if any).

9 Ground cover and Land use InSAR only works under coherent conditions Different InSAR techniques may be better suited to different applications PSI certain PS density required for technique to work

10 Climate Water vapour distribution in the atmosphere can interfere with the phase signal try to remove this in PSI, but can t always as it is dependant on the number of images and the PS density Certain types of snow and ice cover can interfere the phase signal remove these SAR images Flooding, snow melt, surface water can cause coherence to be lost

11 Motion Magnitude Need to consider the revisit time of the satellite and the radar wavelength Magnitude of motion need to be consider will PSI pick it up? PSI motion > 10 cm per year cannot be measured (C-band) PSI can detect motion rates in the order of millimetres per year SAR frequency band (satellites) Minimum revisit time (days) Motion measured in one fringe cycle (mm) DifSAR (mm) Typical minimum detectable movement PSI (mm/year) Artificial Reflector InSAR (mm) DifSAR (mm) Typical maximum detectable movement PSI (mm/year) Artificial Reflector InSAR (mm) X-Band (Terrasar-X) C-Band (Envisat and ERS) < 1 < C-Band (Radarsat-1 and Radarsat-2) L-Band (ALOS) N/A N/A 100+ N/A N/A

12 Motion Spatial Extent The satellite s spatial resolution needs to be considered as motion extent may be too small to be detected. DEAD SEA SINK HOLES HOLBECK

13 Data Availability

14 Typical PS Buildings Bridges Sourced from Royal Belgium Institute of Natural Sciences Sourced from LEGMA Pylons Dams and Dykes Rocky outcrops

15 Factors that affect the presence of a PS (1) Orientation of the target If not in LOS of satellite, the PS feature will not be picked up Reflectivity of the target If not bright compared to surroundings then will not be selected as a PS This can be a combination of orientation of the target and ground cover

16 Factors that affect the presence of a PS (2) Temporal persistence of the target If want to look at something that has only been present since 1997, then PSI processing 1992 to 2000 will not include it as a persistent scatterer

17 Case Study 1: Vilnius, Lithuania (1) Why was the site chosen? USER: Capital city of Lithuania (population of 0.6 million) Built on highly compressible quaternary sediments. Susceptible to differential subsidence due to construction, water abstraction and drought Whole country dependent on ground water => water use is increasing 20% of piped water is leaking => localised subsidence Possible neo-tectonic movement due to glacial rebound Landslides are common along 2 principal rivers which meet in the centre of the city

18 Case Study 1: Vilnius, Lithuania (2) Feasibility assessment

19 Case Study 1: Vilnius, Lithuania (3) Topography

20 Case Study 1: Vilnius, Lithuania (4) Study period: 08 May 1992 to 13 August 2001 Data used: ERS-1 & ERS-2 PS Density: ~ 9 PS/km 2 Average annual motion rate of the entire processed area = mm/year (std dev 0.484) Max motion 56.2 mm/year, minimum motion mm/year 78 % of PS points were classed as stable, 16% of points subsiding and 6 % of points uplifting Problems: small data stack (32 images), highly variable weather conditions, lack of prerequisite knowledge of ground motions

21 Case Study 1: Vilnius, Lithuania (5) What did the result reveal? 1. Ground water extraction / recharge Example: The Territory of Furniture Centre Audėjas PS concentrated in this region indicate negative motion. Buildings were constructed during year period. The ground water level was m below the surface pre-construction. Post construction and the installation of drainage system dropped to 6-7 m. 2. Landslide and Erosion Example: Grigiškės area (alluvial deposits of the Neris and Voke Rivers. PS concentrated in this region showed negative motion. Settlement of houses has occurred in this region. The motion may be related to changes of the groundwater regime. 3. Compressible soils

22 Case Study 1: Vilnius, Lithuania (6) User analysis The examples of movements detected by PSI examined on site included subsidence linked to the presence of compressible deposits (organic, lacustrine etc.), mass movements, suffusion, changes of groundwater, and location of large engineering constructions. Lithuanian Geological Survey No evidence of relationship between PS result and tectonic activity which is thought to occur in the area No new information as such but some areas that require further investigation Applications identified by user: Dams Power Infrastructure Monitoring transportation networks obtaining ground and structure motion information across urban areas

23 Case Study 2: Liege, Belgium (1) Why was the site chosen? USER: City is a former mining site Mine collapse and flooding due to a rising water table. Uplift and subsidence is common Running through the former mining area is an underground high speed train tunnel Southern Liege is a karst area with sink holes and surface collapse. Building damage has led to changes to the law.

24 Case Study 2: Liege, Belgium (2) Feasibility assessment

25 Case Study 2: Liege, Belgium (3) Topography

26 Case Study 2: Liege, Belgium (4) Study period: 23 April September 2005 Data used: ERS-1 / ERS-2 / ASAR PS Density: ~48 PS /km 2 Average annual motion rate of the entire processed area = mm/year (std dev 2.580) Max motion mm/year, minimum motion -3.01mm/year 76 % of PS points were classed as stable, 18 % of points subsiding and 6 % of points uplifting

27 Case Study 2: Liege, Belgium (5) What did the result reveal? 1. Mining Example: Ground movements related to the presence of abandoned mined exploitations (subsidence) and groundwater recharge leading to an increase of the hydrostatic pressure into the mine aquifer (uplift). 2. Ground water extraction / recharge Example: Former coal mining areas are now experiencing rebound after cessation of mining and ground water abstraction as aquifer recharges 3. Compressible soils 4. Reclaimed land

28 Case Study 2: Liege, Belgium (6) User analysis Both GPS and levelling indicate motion on a similar scale to the PS data with the absolute difference lower than 1mm. 2 main movements identified: negative motion around river Meuse (alluvial plains), positive motion in the city centre and in the hills Good correspondence with karst sites Applications identified by user: better determine spatially the ground motion related to mining groundwater provide information for housing management authorities Underground construction Post mine hydrological surveys

29 Case Study 3: Esbjerg, Denmark (2) Feasibility assessment

30 Case Study 3: Esbjerg, Denmark (1) Why was the site chosen? USER: City vulnerable to flooding from the north sea storm surges harbour often experiences flooding which causes economic loss Isostatic rebound has caused vertical land movement and sea level rises have caused an increased flood risk Higher frequency of storm surges => greater risk of flooding Study potential impact of flooding on ground motion and identify zones where dykes may be at risk South of Esbjerg the land is very low and is protected by dykes

31 Case Study 3: Esbjerg, Denmark (3) Topography

32 Case Study 3: Esbjerg, Denmark (4) Study period: 13 June December 2000 Data used: ERS-1 & ERS-2 PS Density: ~19 PS /km 2 Average annual motion rate of the entire processed area = mm/year (std dev 1.954) Max motion 39.1 mm/year, minimum motion mm/year 87 % of PS points were classed as stable, 4 % of points subsiding and 9 % of points uplifting

33 Case Study 3: Esbjerg, Denmark (5) What did the result reveal? 1. Compressible sediments 2. Ground abstraction 3. Landfill (suspected) Example: Harbour: The Esbjerg Harbour is to a large degree built on landfill. This is reflected by some cases of very high vertical movement. 4. Construction induced subsidence 5. Sediment consolidation Example: Railroad: In the area west of Esbjerg, along the railroad towards Tjæreborg and Bramming. numerous cases are found where the railroad tracks appear to induce a settling of known unconsolidated postglacial freshwater and marine deposits.

34 Case Study 3: Esbjerg, Denmark (6) User analysis Some known and newly discovered defornation (railway line settlement) identified Applications identified by user: PSI provides synoptic view from regional to local scale Easier to understand processes occurring than traditional levelling where observations are primarily taken along a linear profile.

35 Conclusion PSI can be used monitor a variety of applications but it is crucial a feasibility assessment is carried out to ascertain the suitability of the technique Doing a feasibility assessment allows us to: Optimise data stack to get the maximum number of PS Look at the weather to remove any potentially problematic images Be aware that if it is a small data-stack, then there will be a low PS density which may affect processing Mechanism Vilnius Liege Esbjerg Creep Compressible soils Reclaimed Land site Dissolution Groundwater recharge Groundwater abstraction Mining Erosion Landslide Waste Disposal / Landfill