INTERFEROMETRIC SYNTHETIC APERTURE RADAR. An Introduction for Users of InSAR Data

Transcription

1 INTERFEROMETRIC SYNTHETIC APERTURE RADAR An Introduction for Users of InSAR Data

2 Table of Contents An Introduction for Users of InSAR Data Background Basics of Synthetic Aperture Radar (SAR) Satellites SAR Sensors SAR Image Acquisition Radar Frequencies and Viewing Geometry Signal Properties Satellite Orbits and Geometrical Distortions Resolution Basics of InSAR Interferometry Interferograms Contributors to Signal Phase Coherence Differential InSAR (DInSAR) Interferogram Stacking Persistent Scatterer Techniques General Concept Permanent Scatterers Calculating Displacement Precision Validation of PS Data Data Output and Presentation SqueeSAR Artificial Reflectors (AR) Strengths and Weaknesses of PS Analyses Synergistic Use of PS and GPS APPENDIX GLOSSARY PROCESS MAP Page 2

3 Table of Figures Figure 1 Satellite Radar Systems available now and into the future Figure 2: Illustration of the relationship between the orbit path of the satellite and true North South. It varies from satellite to satellite but is generally around 10 (see Section 2.6 for definition of ascending and descending orbits) Figure 3: Schematic showing how a satellite acquires a strip map of the earth s surface Figure 4: Schematic showing the orientation of signal launch toward the earth s surface... 8 Figure 5(a-b): Images showing the effect of multi image reflectivity (MIR) Figure 6: Motion measured by the sensor for different directions of terrain-motion. Red arrows represent the vector of terrain motion while blue arrows represent the LOS motion measured by the radar system Figure 7: Illustration of the method for obtaining actual motion by combining ascending and descending orbit information. In the case of vertical ground displacements (left image) the motion components on the ascending and descending directions are both negative (moving away from the sensor) while in the case of a horizontal (E-W) motion (right image) one vector is positive (moving toward the sensor) while the other is negative Figure 8: Schematic illustration showing how features on the landscape are projected on to the line of sight of the satellite radar beam Figure 9: An amplitude image of Mt. Vesuvius, in SAR coordinates. The North direction is to the left of the image (the satellite flies from left to right in a descending orbit). The coast line along the Tyrrhenian Sea is clearly visible (water appears as black). Urban areas can be identified as bright spots on the image created by the strong amplitude responses from reflections off buildings. The eastern slopes of the volcano (at the top of the image) are compressed compared to those on the western slopes (foreshortening) Figure 10: Schematic illustration showing how mountainous terrain can create noise through layover and shadow effects Figure 11: A schematic showing the relationship between ground displacement and signal phase shift. The numerical value of the wavelength is that of ERS Figure 12: An interferogram generated from two radar images one of which was acquired before the L Aquila earthquake (February 2009) and the other shortly after the event (April 2009). The fringes indicate coherence whereby displacement can be calculated in the corresponding areas. The areas with a spotty appearance are areas where decorrelation noise has occurred. Phase values range from π to +π Figure 13. The visual display of results of a PSInSAR analysis of Lake Presenzano and its surrounding area Page 3

4 Figure 14: A typical time series showing linear and non-linear patterns of movement Figure 15: Typical values of precision (1 sigma) for a point less than 1km from the reference point (P 0 ), considering a multi-year dataset of radar images Figure 16: Comparison of PSInSAR with GPS data. The x, y and z components of GPS measurements have been resolved to the equivalent LOS of the satellite data Figure 17: Optical leveling. The blue line is an optical benchmark correction curve; the red dots represent InSAR readings at the same location Figure 18: Thermal Dilation. Buildings move in response to changes in temperature and software is available to model such movement. The black line represents the results of a thermal dilation model while the red triangles correspond to InSAR readings on the same building, measured over the same time period Figure 19(a-c): These images are screen-grabs from a GIS showing how distant and close-up views of deformation phenomena can be observed using GIS platforms Figure 20(a-d): Artificial reflectors Figure 21: The Right Bank Landslide, Lake Sarez, Tajikistan showing the results of analyses of movement using both GPS and PSInSAR technologies Table of Tables Table 1: Nominal Resolution Cell Sizes of all Commercial Radar Satellites. Numbers in brackets, for COSMO-SkyMed data, refer to the use of 2 satellites of the constellation, operating in tandem Table 2: Summary of strengths and weaknesses of InSAR Table 3: Comparison of PS and GPS technologies Page 4

5 INTERFEROMETRIC SYNTHETIC APERTURE RADAR An Introduction for Users of InSAR Data 1. Background This document is a general overview of the concepts related to the measurement of surface deformation phenomena using Synthetic Aperture Radar Interferometry (InSAR). The text has been written in layman terms, avoiding radar specialist jargon. Should the reader be interested in a more detailed explanation, he/she can refer to the suggested supplementary reading, appended to this document. The two main fields of application of InSAR data are: a) reconstruction of digital elevation models of large areas; b) detection and monitoring of surface deformation phenomena and, in general c) measurement of displacement rates of objects on the ground. In this document, the focus will be on the latter topic. The first theoretical study on this subject dates back to the 1980s (Gabriel et al., 1989) when the first proof of concept using SEASAT data was published by the Jet Propulsion Laboratory s radar group. With the launch of the European Space Agency s (ESA) ERS-1 satellite, in July 1991, and ERS-2, in April 1995, an ever-growing set of interferometric data became available to many research groups. While more and more InSAR results were generated, the presence of atmospheric artifacts became more and more evident and dampened, somewhat, the enthusiasm. Research efforts were then devoted to different strategies for the combination of several interferograms or the fusion of InSAR data with prior information, in order to reduce the impact of the atmospheric disturbances. Among several research groups pursuing this challenge was the Department of Electrical Engineering of the Politecnico di Milano, in Italy. By 1999, the Institution had developed the Permanent Scatterer Technique along with its proprietary PSInSAR algorithm. In March 2000, the Politecnico di Milano and the three inventors of the technology founded TRE, to create a team specialized in InSAR data processing. Since that time, other organizations have developed similar technologies, but TRE has still the largest team of engineers specifically working on SAR interferometry. 2. Basics of Synthetic Aperture Radar (SAR) 2.1 Satellites The family of satellites that carry, or will be carrying, SAR sensors for commercial applications is illustrated in Figure 1. Other SAR-bearing satellites exist but are used exclusively for military applications. Page 5

6 Figure 1 Satellite Radar Systems available now and into the future. All satellites equipped with SAR sensors orbit the earth on a near-polar orbit at an altitude ranging from 500 to 800 km above the earth s surface, depending on the satellite platform hosting the SAR sensor. The angle between true north-south and the satellite orbit varies slightly, depending on the satellite but, in general lies in the range of 10 degrees, as shown in Figure 2. Ascending Orbit Descending Orbit Figure 2: Illustration of the relationship between the orbit path of the satellite and true North South. It varies from satellite to satellite but is generally around 10 (see Section 2.6 for definition of ascending and descending orbits). Page 6

7 Data from different satellite sources can generally be purchased or ordered without limitation, with the exception of data from the Japanese Space Agency (JAXA). Although ALOS- PALSAR data can be purchased, JAXA selected the acquisition modes of the satellite for the duration of its operating life (past and future), at the beginning of the mission, so the user cannot select the radar acquisition mode most suitable for the application at hand. Another important point to be considered is that the Italian COSMO-SkyMed constellation is a joint military/civilian mission. Whenever a conflict arises between acquisition requests, commercial projects have lower priority. 2.2 SAR Sensors Because the illuminating source of radar is microwave energy, radar signals are unaffected by darkness or clouds, in terms of visibility of the land surface. As will be discussed later, clouds impact the accuracy of InSAR but do not obstruct the passage of the signal through the medium. Therefore, SAR can function 24 hours per day, 365 days per year. The sensors emit signals with a specific central frequency. In addition, radar systems are associated with specific bands of the electromagnetic spectrum. Those commonly used in InSAR applications are L-band (1-2 GHz, ~24 cm wavelength), C-band (5-6 GHz, ~6 cm wavelength) and X-band (8-12 GHz, ~3 cm wavelength). In 1992, with the launch of ERS-1, the first SAR satellite for commercial applications, the onboard sensor offered but one acquisition mode, a single look angle, a single resolution cell size, and a single signal wavelength. The only options available related to viewing geometry, i.e. ascending or descending satellite tracks. Today, as Figure 1 indicates, the options have proliferated with the increase in numbers of satellites, offering the end-user a wide choice of look angle, repeat orbit cycle, resolution cell size, and signal wavelength. It is now possible to design a monitoring program much better suited to an end-user s needs than was possible 10 years ago. 2.3 SAR Image Acquisition As the satellite circumnavigates the earth, it launches millions of radar signals toward the earth along the radar beam s line of sight (LOS), on a continuous basis. Following impact with the earth s surface, some of the signals are reflected away from the satellite, some are absorbed in vegetation or other non-reflective materials and some reflect back to the satellite. Using the signals reflected off the earth s surface, also referred to as backscattered signals, processors on board the satellite integrate the returning signals to form a strip map. Usually, the on board memory capacity is limited so the satellite has to transmit the data to strategically located ground stations. These stations then compile images which can be used for data analysis. Figure 3 is a schematic showing the process of image acquisition by a satellite. Page 7

8 Figure 3:: Schematic showing how a satellite acquires a strip map of the earth s earth surface. 2.4 Radar Frequencies and Viewing Geometry Radar signals are transmitted in pulses (Figure 4). In any SAR system, there are three important frequencies that define its operations. The so-called called Pulse Repetition Frequency F (PRF) is the rate at which those pulses are transmitted and defines the resolution of the system in azimuth direction (i.e. the direction parallel to the satellite velocity), velocity) the central frequency (f0) defines the operating wavelength of the system and characterizes izes its propagation and penetration features,, as well as the sensitivity of the system in interferometric applications. applications Finally, radar adar pulses backscattered backscattere by the Earth surface are sampled by the radar system at another frequency (fs) defining the nominal pixel size, sometimes referred to as the resolution cell size, in the range direction, direction related to the sensor-to-target distance. Figure 4:: Schematic showing the orientation of signal launch toward the earth s surface. surface Page 8

9 Satellite sensors are mounted on their platforms with the direction of transmission at 90 to the flight direction. The earlier satellites (ERS-1, ERS-2, Radarsat-1 and Envisat) were all rightlooking satellites, meaning that microwave beam transmits and receives on the right side only of the satellite, relative to its orbital path, i.e. the system cannot rotate. Newer satellites (Radarsat-2, TerraSAR-X and COSMO-SkyMed) have both right-looking and left-looking capabilities, thus they can look to the right or the left of the craft, but not both directions simultaneously. The angle at which the sensor is pointed toward the earth s surface is referred to as the offnadir, or look, angle. The off-nadir angle of the ERS satellites was fixed at about 23, but all subsequent satellites were fitted with the means to vary the viewing angle of the sensors, ranging from values of 20 to 50 degrees. This ability to vary the off-nadir angle is important in that it is possible to adjust for hilly or mountainous terrain potential impediments to InSAR if the relationship between viewing geometry and terrain slope is not optimal. 2.5 Signal Properties Radar signals are characterized by two fundamental properties: amplitude and phase. Amplitude is related to the energy of the backscattered signal. When a signal leaves the transmitting sensor, it is broadcast at a specific energy level. On reaching an object on the ground surface, that energy level is changed depending on a number of circumstances that relate primarily to the reflective quality of the object. Metal and hard objects (natural and artificial) have a high reflective quality and, thereby, the amplitude of the reflected signal will be much higher than the background noise of the system. Softer materials, such as wood, crops, asphalt, have a lower capacity to reflect incident radar energy and so the amplitude of the reflected signal is strongly diminished. The amplitude characteristics of signals can be visualized and an individual amplitude image will appear speckled. This is because each resolution cell comprises many scattering elements, all reflecting the incoming signals back to the satellite with different signal strengths and slightly different delays (phases), creating a spotty appearance. From one image to the next, the speckle in corresponding resolution cells can be constant, and it can vary. Constant levels of reflectivity, often bright spots, are indicators of stable reflections of radar signals. When the speckle is varied, from image to image, the pixels are decorrelated across the data set and the speckling can be minimized by averaging the amplitude of all images within the stack. The result, which is referred to as a Multi Image Reflectivity (MIR) map, is a means to improve clarity of the amplitude response of the stack, highlighting those pixels that have a stable and high reflectivity in each of the images within the stack. Figure 5(a) corresponds to a single amplitude image while Figure 5(b) is the MIR map for a stack of 60 scenes. Page 9

10 Figure 5(a) An ERS-2 SAR amplitude image of Linate Airport in Milan (Italy): the speckle effect on the fields surrounding the airport is clearly visible. Figure 5(b): An MIR map of the same area at Linate Airport: the speckle effect on the fields surrounding the airport has disappeared. Figure 5(a-b): Images showing the effect of multi image reflectivity (MIR). Apart from amplitude values, radar systems record phase values, the key element in any interferometric measurement, since it is related to the sensor-to-target distance. As has already been mentioned, radar signals are characterized by a certain frequency of operation and, for interferometric applications, they can be thought of as sinusoidal waves: one complete cycle (from π to +π) corresponding to the wavelength. It is this specific property of the radar signal, and the system s ability to record both amplitude and phase information for each image pixel, that are used in estimating displacement. A signal s phase can be affected by changes in the atmosphere as it travels from the satellite to the earth and then back again. In the atmosphere, there are always layers of moisture (cloud, fog, rain etc.) through which a signal must pass. In arid areas, these layers are few compared to tropical and temperate areas. As a signal encounters a moisture-bearing layer in the atmosphere, the propagating speed of the signal changes. As a result, errors are introduced into the phase values recorded by the receiving sensor, on the satellite. Since wavelength and signal phase have a simple and direct correlation, any change in wavelength corresponds to a change in the phase of a signal. As will be explained later in Section 3, this is an important issue in measuring ground movement because there is also a direct relationship between ground displacement and signal phase. Page 10

11 2.6 Satellite Orbits and Geometrical Distortions The circumpolar orbits of all satellites mean that, for part of their trajectory, they are travelling from the North toward the South. This direction is referred to as Descending. Conversely, when the satellite is traveling from the South toward the North, it is said to be in an Ascending orbit. By acquiring imagery during both ascending and descending orbits, it is possible to view a point on the earth s surface from two different perspectives, from the East or from the West. This is an important element of measuring vertical and horizontal motion. What is actually measured in interferometric applications is the projection of a target s motion onto the LOS. If the motion direction is close to the angle of the LOS then the measured and actual motions will be similar. However, the LOS motion can often differ noticeably from the real value of motion, especially in cases where the ground motion is not vertical (Figure 6). Figure 6: Motion measured by the sensor for different directions of terrain-motion. Red arrows represent the vector of terrain motion while blue arrows represent the LOS motion measured by the radar system. Measuring true vertical and horizontal components of motion is difficult with a single geometry acquisition, unless a priori information is available on the true vector of movement (not a common situation!). However, by using ascending and descending data together, it is possible to combine the measured motion information to obtain an accurate estimate of the actual vertical motion and of the East-West component of the motion (Figure 7). Figure 7: Illustration of the method for obtaining actual motion by combining ascending and descending orbit information. In the case of vertical ground displacements (left image) the motion components on the ascending and descending directions are both negative (moving away from the sensor) while in the case of a horizontal (E-W) motion (right image) one vector is positive (moving toward the sensor) while the other is negative. Page 11

12 Figure 8 illustrates how features on an undulating landscape will be viewed by the satellite. In pixels (resolution cells) 1 and 2, the equilateral triangles on the landscape appear slightly distorted in the LOS (range, sometimes referred to as slant-range). As the signal reaches pixel 3, there is a marked change in ground slope and many more triangles appear in the pixel, even though they are all of the same size. The effect is to compress these triangles in the LOS, referred to as foreshortening. When the radar progresses to pixels 4 and 5, at which point the ground slope and LOS are parallel, the triangles now appear stretched at their base. This distortion in the appearance of land use can be seen in Figure 9, which is an amplitude image of Mount Vesuvius, in Italy, viewed in SAR coordinates (range and azimuth, corresponding to the vertical and horizontal axis, respectively). Off-Nadir Slant-range Slant-range resolution resolution Slant-range Topography G round range Figure 8: Schematic illustration showing how features on the landscape are projected on to the line of sight of the satellite radar beam. Azimuth Range Figure 9: An amplitude image of Mt. Vesuvius, in SAR coordinates. The North direction is to the left of the image (the satellite flies from left to right in a descending orbit). The coast line along the Tyrrhenian Sea is clearly visible (water appears as black). Urban areas can be identified as bright spots on the image created by the strong amplitude responses from reflections off buildings. The eastern slopes of the volcano (at the top of the image) are compressed compared to those on the western slopes (foreshortening). Page 12

13 In hilly or mountainous terrain, it sometimes occurs that the projection of steep slopes on to the LOS is reversed. Figure 10 illustrates how this phenomenon manifests itself in radar imagery. In pixel 1, the radar sees the object A normally. However, when the radar reaches pixels 2, 3 and 4, the objects E, F, and G are present in the same pixels as objects B, C, and D, the latter being masked by the former. This phenomenon is referred to as layover and generates noise. In an amplitude image, it appears as a bright white layer and, in Figure 9, it can be observed around the caldera of Mount Vesuvius. As the radar progresses from pixels 4 to 7, the slope of the ground is greater than that of the LOS and, so, the area in question cannot be seen by the sensor and, in an amplitude image, appears as a black area, as can also be seen in Figure 9. This effect is referred to as shadow. Figure 10: Schematic illustration showing how mountainous terrain can create noise through layover and shadow effects. 2.7 Resolution To most end-users of interferometric data the terms resolution and pixel are usually synonymous. They are different, however, although similar in numerical value. The explanation of the difference is rooted in signal processing theory and is beyond the scope of this User Manual. Accordingly, the resolution in the range and azimuth directions are usually referred to in nominal value terms and, for the various satellites that carry SAR sensors, these values are shown in Table 1. Page 13

14 Satellite Band Acquisition Mode Nominal Pixel Dimension: ground range x azimuth (m) Repeat Cycle (day) ERS-1 & ERS-2 C Standard Beam 20x4 35 Envisat C Standard Beam 20x4 35 Radarsat-1 C Standard Beam 20x5 24 Radarsat-1 C Fine Beam 10x5 24 Radarsat-2 C Standard Beam 20x5 24 Radarsat-2 C Fine Beam 10x5 24 Radarsat-2 C Ultra Fine Beam 3x3 24 TerraSAR-X X Standard 3x3 11 TerraSAR-X X Spotlight 1.5x Cosmo SkyMed X Standard 3x3 16 (8) Cosmo SkyMed X Spotlight 1.5x (8) ALOS PALSAR L Fine Beam 10x5 46 Table 1: Nominal Resolution Cell Sizes of all Commercial Radar Satellites. Numbers in brackets, for COSMO-SkyMed data, refer to the use of 2 satellites of the constellation, operating in tandem. The values in Table 1 relate to a land surface that is flat and horizontal. In reality, much of the earth s landscape contains land that is irregular in surface profile (hills, valleys etc.) and these features immediately impact the actual dimensions of the resolution cell. 3. Basics of InSAR 3.1 Interferometry Interferometric Synthetic Aperture Radar (InSAR), also referred to as SAR Interferometry, is the measurement of signal phase change, or interference, over time. When a point on the ground moves, the distance between the sensor and the point on the ground also changes and so the phase value recorded by a SAR sensor flying along a fixed orbit will be affected, too. Figure 11 shows the relationship between that ground movement and the corresponding shift in signal phase between two SAR signals acquired over the same area. Page 14

15 Figure 11: A schematic showing the relationship between ground displacement and signal phase shift. The numerical value of the wavelength is that of ERS. The change in signal phase ( φ) can be expressed in the form of the following simple equation: 4π ϕ = R + α λ Where λ is the wavelength, R is the displacement and α is a phase shift due to different atmospheric conditions at the time of the two radar acquisitions. As a consequence, any displacement of a radar target along the satellite line of sight creates a phase shift in the radar signal that can be detected by comparing the phase values of two SAR images acquired at different times. Apart from decorrelation effects, to be discussed in the next sections, SAR interferometry can only be applied in the following circumstances: Images have to be acquired by the same satellite using the same acquisition mode and properties (beam, polarization, off-nadir angle, etc); Images have to be acquired with the satellite in the same nominal orbit; The baseline separation ation between the master scene and any of the slave scenes must be no more than the critical baseline (a parameter that varies with the SAR sensor in use); the baseline being the distance between the satellite paths. 3.2 Interferograms An interferogram is the difference of the phase values corresponding to a certain area, i.e. it is a digital representation of change in surface characterization. It is a matrix of numerical values ranging from π to +π (since they correspond to phase variations) and it can be converted to a map the easiest way to observe whether or not motion has occurred over a certain area. Page 15

16 Figure 12 is an interferogram of the L Aquila earthquake that occurred in Italy, in April The colored bands, referred to as fringes, indicate areas where movement can be measured. The highly speckled areas indicate where some form of decorrelation arose. Here the noise level (mostly due to vegetation) prevents the application of InSAR and no useful information can be extracted. Data were acquired by the ENVISAT satellite (see Table 1 for the properties of this satellite) for which one phase cycle corresponds to 28 mm of ground deformation along the line of sight (neglecting atmospheric effects). The analysis of a SAR interferogram is not a trivial task to perform for non specialists. Apart from noise and decorrelation effects, interferometric phase values are a blend of different signal contributions, as will be discussed in the next section. Figure 12: An interferogram generated from two radar images one of which was acquired before the L Aquila earthquake (February 2009) and the other shortly after the event (April 2009). The fringes indicate coherence whereby displacement can be calculated in the corresponding areas. The areas with a spotty appearance are areas where decorrelation noise has occurred. Phase values range from π to +π. 3.3 Contributors to Signal Phase Interferometric phase ( φ) is impacted by four contributions: topographic distortions arising from slightly different viewing angles of the two satellite passes (t), atmospheric effects (α) arising from the wavelength distortion that occurs when signals enter and leave a moisturebearing layer, any range displacement of the radar target ( R), and noise; range being the distance between the sensor and the target. More precisely: ϕ = 4π λ R + α + t + noise It is then clear that the difficulties related to the estimation of surface deformation signals from a single SAR interferogram are essentially due to the presence of decorrelation effects Page 16

17 (contributing to the noise level), the impact of local topography on phase values and the presence of atmospheric phase components superimposed on the signal of interest. In Figure 12, most of the fringes visible in the interferogram are due to co-seismic deformation induced by the earthquake: in fact, the impact of the local topography has been removed, and atmospheric disturbances are not evident in this image. 3.4 Coherence Interferometric fringes can only be observed where image coherence prevails. When an area on the ground appears to have the same surface characterization in all images under analysis, then the images are said to be coherent. If the land surface is disturbed between two acquisitions (e.g. an agricultural field has been ploughed, tree leaves have moved positions, etc.), those subareas will decorrelate in an InSAR analysis, resulting in noise and no information being obtainable. Coherence and correlation have the same meaning in this context. The term noise is frequently used in this context and it is another word for non-coherence, or decorrelation. The fringes visible in Figure 12 reveal areas with high coherence while the speckled areas represent very low coherence and noise. The coherence of an interferogram is affected by several factors, including: Topographic slope angle and orientation (steep slopes lead to low coherence) Terrain properties The time between image acquisitions (longer time intervals lead to lower coherence) The distance between the satellite tracks during the first and second acquisitions, also referred to as the baseline (larger baselines lead to lower coherence) Typical sources of decorrelation are: Vegetation. Leaves grow and die and they also move. From one scene to the next, these changes are sufficient to change the appearance of the surface characterization. This is a particular problem for X-band and C-band sensors. L-band sensors can overcome this limitation in many situations, because their significantly longer wavelength is able to see through foliage and reflect off objects beneath the vegetation and back through the foliage. Construction. At a construction site, the appearance of the land surface is changing constantly. This is a problem that is common to X-band, C-band, and L-band sensors. Erosion. Whether prompted by rain, snowmelt or wind, surface erosion will also change the surface characterization of land and, thereby, can decorrelate those areas where erosion is prevalent. Rapid Movement. Landslides and earthquakes precipitate rapid motion of an area of land. Quite often, the rapid motion causes destruction and, with it, a total change in the land surface s appearance. With earthquakes, it is sometimes possible for rapid motion to occur without changes to surface characterization and, in those situations, interferometry can be successful. If the total movement occurring between successive image acquisitions exceeds one-half of the signal s wavelength, decorrelation is likely to occur. Page 17

18 Coherence is measured by an index which ranges from 0 to 1. When an area is completely coherent, it will have a coherence value of 1. Correspondingly, if an area completely decorrelates, its coherence index will be 0. In general, interferometry is successful and accurate deformation is measurable when the coherence index lies between 0.5 and 1.0. Interferometry can still produce meaningful results with coherence levels below 0.5 but as the index gets lower, so the results will display increasing levels of noise and may show erratic deformation patterns, from scene to scene, although movement trends are visible and generally reliable. Wherever fringes occur, it is possible to calculate deformation by calculating the number of fringes and multiplying them by half of the wavelength. In the case of L Aquila, C-band SAR was used and, therefore, each fringe should be multiplied by 28 mm (one-half of the wavelength) to calculate the total apparent displacement. 4. Differential InSAR (DInSAR) When a pair of images is subjected to interferometric analysis with a view to identifying movement and, thereafter, quantifying that movement, the process is referred to as Differential InSAR. Since change detection is now the goal, topographic effects are compensated for by using a Digital Elevation Model (DEM) of the area of interest, creating what is referred to as a differential interferogram (the word differential here refers to the subtraction of the topographic phase contribution from the SAR interferogram). The equation in section 3.3, can then be represented as follows: ϕ = 4π R + α + ε + noise λ Where ε is the contribution to phase arising from possible errors in the DEM that was used to remove the topographic effects. Whenever the noise is low (i.e. decorrelation effects are negligible) and the phase contribution due to the local topography is accurately compensated for (i.e. ε is negligible as well), the interferometric phase can be simplified to the following equation: 4π ϕ = R + α λ Where φ is the differential interferometric phase, R is the incremental distance the signal travels from the sensor to the ground and back, and α is the atmospheric contribution to phase shift. Once the differential interferogram has been prepared, a deformation map can be created for all areas that are coherent. In the mid-90 s, after extensive application of the DInSAR technology, the atmospheric contribution to phase shift was found to be significant, particularly in tropical and temperate areas. Unfortunately, there is no method for removing the α component, so users have to be aware of its effects. Thus, DInSAR should only be used on the understanding that deformation measurements are prone to errors arising from atmospheric circumstances. However, DInSAR, Page 18

19 while not the tool for accurate displacement measurements, still has a use in identifying footprints of progressing movement. It can only measure total displacement between two points in time. Accordingly, it cannot distinguish between linear and non-linear motion. 5 Interferogram Stacking Following the realization that atmospheric effects on signal phase values were significant, a method emerged in the late 1990 s that sought to mitigate this effect by averaging data within multiple interferograms. This process was referred to as Interferogram Stacking. By averaging the data in a stack of interferograms, the signal to noise ratio (SNR) values are enhanced and, theoretically, it is easier to extract information on displacement over longer periods of time than are realistic for single interferogram DInSAR. However, for this process to work, certain assumptions are made: Although different versions of this technique exist, the displacement rate of the area of interest is assumed to be constant in time. In reality, such an assumption has limited validity. Multiple interferograms usually describe ground movement over time lines measured in years. Apart from tectonic deformation, linear movement over such time periods is not common. The data are heavily filtered, spatially, before the stacking procedure is implemented. Not only does this reduce the resolution but also prompts the loss of potentially valuable data contained in isolated pixels with high SNR values, and it also smoothes out abrupt changes in displacement, e.g. seismic faults. The atmospheric contribution to signal phase is not estimated. Thereby, no assessment is possible on the quality of the filtering procedure. Atmospheric disturbances are characterized by specific statistical features, and the separation of motion and atmospheric phase components should take into account the peculiarities of the "noise" to be filtered out. Typically, stacking procedures are only applied using interferograms with an orbital baseline less than 300m, because of the spatial filtering. As a result, substantial quantities of information that can be found from within interferograms whose baselines are as high as 1300m are overlooked, the latter being a common baseline upper limit for PSI technologies. While interferogram stacking provides the user with better information than can be obtained from single differential interferograms (DInSAR) the approach is far from optimal, particularly because deformation cannot be considered constant in time. Moreover, for the estimation of atmospheric noise, the procedure usually adopted to produce a weighted average, i.e. to assign different "importance" to different interferograms, is based on visual inspection of multiple interferograms. Finally, as already mentioned, the estimation of errors is usually not performed. Page 19

20 6. Persistent Scatterer Techniques 6.1 General Concept Persistent Scatterer Interferometry (PSI) is the collective term used within the InSAR community to distinguish between single interferogram DInSAR and the second generation of InSAR technologies, of which there are but a few. The first of these to appear, in 1999, was the PS Technique, the base algorithm of which is PSInSAR. It is proprietary to the Politecnico di Milano (Polimi) and licensed exclusively to TRE for commercial development. TRE has no specific knowledge of the competing algorithms; however, in concept they are all likely to be similar in approach, although probably different in their analytical capability. The following description of PSI technology is based on the PSInSAR model. All PSI technologies are advanced forms of DInSAR. In other words, the interferogram is at the core of PSI. The fundamental difference is that PSI technologies develop multiple interferograms from a stack of radar images. As a minimum, 15 radar scenes are usually required for PSI methods, including PSInSAR, even though there are circumstances when an analysis can be conducted with fewer images (typically in urban areas). However, it should be noted that the more there are radar scenes available, the more accurate will be the results of PSInSAR, and the same holds true for other PSI methods. The main driver for the development of PSInSAR was the need to overcome the errors introduced into signal phase values by atmospheric artifacts. By examining multiple images, usually a minimum of 15 scenes, many interferograms (in this case 14 interferograms) are generated by selecting one of the scenes as a master to which the other 14 scenes become slaves. The process by which removal of atmospheric effects is achieved involves searching the imagery and interferograms for pixels that display stable amplitude and coherent phase throughout every image of the data set. They are referred to as Permanent - or Persistent Scatterers. Thus a sparse grid of point-like targets characterized by high signal to noise ratios (SNR) is identified across an area of interest on which the atmospheric correction procedure can be performed. Once these errors are removed, a history of motion can be created for each target. Having removed the atmospheric artifacts, the interferometric data that remain are displacement values (resolved along the satellite LOS) plus noise, dependent on the quality (SNR) of the reflector. 6.2 Permanent Scatterers A Permanent Scatterer (PS) is defined as a radar target, within a resolution cell, that displays stable amplitude properties and coherent signal phase, throughout all of the images within a data stack. Sometimes a target may behave with a stable amplitude characteristic but its phase is erratic, or non-coherent. Further, some targets behave as if they are PS but only within a portion of the images within the data stack. Such targets are not PS. Page 20

21 Objects that make good PS are varied and can be natural or man-made. Among the natural forms are: rock outcrops, hard un-vegetated earth surfaces, and boulders. Among the manmade objects are: buildings, street lights, transmission towers, bridge parapets, above-ground pipelines, appurtenances on dams and roof structures, and any rectilinear structure that can create a dihedral signal reflection back to the satellite. Figure 13 shows the results of a PSInSAR analysis of a man-made reservoir, in Italy. The colored dots represent the location of a PS, the color reflecting the displacement rate measured at that point. Figure 13. The visual display of results of a PSInSAR analysis of Lake Presenzano and its surrounding area. 6.3 Calculating Displacement All measurements are made in the LOS of the satellite s radar beam and are relative to a point that is pre-selected as being stable and not moving (P 0 ). The selection of the reference point is best made conjunctively with the client, the latter having better local knowledge on which subareas are stable within an AOI. Once the data have been cleaned up, it is possible to develop the history of movement across the AOI. This is achieved by sequentially calculating the relative displacement between an individual radar target and the reference point, throughout the entire period of the analysis. Thus, the deformation is relative in time and space. A typical time series of movement of a PS is shown in Figure 14. Page 21

22 Figure 14: A typical time series showing linear and non-linear patterns of movement. It should be noted that the PSInSAR algorithm generates a standard deviation map for the AOI, as well as providing error bar data for each PS, within the data base. A priori information is always helpful before commencing a PS analysis. If an area is known to be subsiding, then measurement can be satisfactorily made using a single viewing geometry, also referred to as acquisition mode. However, if the hazard is a landslide, where significant horizontal movement might occur, the use of data acquired by satellites in both the ascending and descending orbits will enable true vertical movement and the East-West component of horizontal movement to be computed. At the present time, it is not possible to determine the horizontal component of movement in the North-South direction. However, research is underway to try to solve this problem.. Such computations will requiree the use of at least 3 data sets with differing viewing geometries and look angles. 6.4 Precision Error bars of measurement of a PS are calculated as the deformation pattern is developed. However, precision of the displacement calculations is an important element in validating PS data. The most important factors impacting on data quality are: Spatial density of the PS (the lower the density, the higher the errorr bar) Quality of the radar targets (signal-to-noise ratio levels) Climatic conditions at the time of the acquisitions Distance between the measurement point (P) and the reference (P 0 ) Figure 15 is a chart showing precision values obtained from many analyses of data from the ERS, Envisat, and Radarsat-1 satellites. Page 22

23 . Figure 15:: Typical values of precision (1 sigma) for a point less than 1km km from the reference point (P0), considering a multi-year year dataset of radar images. images Comparable values for the satellites launched during 2007/8 are not yet available since the volume of data from these satellites that has been been processed to date is still quite low. However, it is expected that precision will be improved because a) the sensors on the newer satellites are more sophisticated, and b) the resolution cell sizes are smaller than those of the earlier satellites. 6.5 Validation of PS Data PS data have been compared with measurements obtained by other recognized measurement methods. However, it must be remembered that InSAR methods determine relative displacement, not absolute movement. Notwithstanding, it is possible to develop some comparisons and Figure 16, 16 Figure 17 and Figure 18 show how PS data performed against Differential GPS and optical leveling surveys, as well as with thermal dilation modeling of buildings. The interested readers should refer to to the technical papers published by POLIMI and TRE. Figure 16: Comparison of PSInSAR with GPS data. The x, y and z components of GPS measurements have been resolved to the equivalent equiv LOS of the satellite data. Page 23

24 Figure 17: Optical leveling. leveling The blue line is an optical benchmark correction curve; the red dots represent InSAR R readings at the same location. location Figure 18: Thermal Dilation. Dilation Buildings move in response to changes in n temperature and software is available to model such movement. The black line represents the results of a thermal dilation model while the red triangles correspond to InSAR readings on the same building, measured over the same time period. 6.6 Data Output and d Presentation The results of an InSAR analysis are best understood if they can be visualized and, in this regard, geographic information systems (GIS) are excellent tools. The digital data are provided in ESRI shapefile format, which includes a database file, f readable in most spreadsheet software, software and can be used as input to downstream am modeling exercises. Figure 19 shows a display of the database in which the location coordinates and displacement history of each PS is listed, along with other data about the particular PS, such as coherence, average velocity, acceleration, and height of the PS centroid relative to that of the reference point used in the PS analysis. Visualization ualization is possible using several forms: overlays on a digital orthophoto on a GIS overlays on an engineering drawing on a GIS overlays on a Google Earth platform on line hosting on a webgis Page 24

25 All of these options allow the viewer to obtain close up and remote remote observation and, with the PSInSAR service, a software tool is provided to enable the viewer to point the cursor to any PS, click on it, and to view the pop up window showing the history of movement of that PS. Figure 19(a-c) represents a sequence of screen grabs from a GIS showing the zooming features. Figure 19(a):: GIS area, showing PS. Figure 19(b):: GIS close up of AOI. Figure 19(c): 1 PS superimposed on topographical map. Figure 19(a-c): These images are screen-grabs screen grabs from a GIS showing how distant and close-up close views of deformation phenomena can be observed using GIS platforms. TRE cannot confirm whether the viewing options described above are available from all providers of PSI services but are standard procedure with the PSInSAR process. 7. SqueeSAR In 2010, the new SqueeSAR algorithm was developed, which is an advance on the PSInSAR algorithm. SqueeSAR is a second generation PSInSAR analysis: exploiting both point wise PS and spatially distributed scatterers (DS). The new algorithm provides information in low-reflectivity reflectivity homogeneous areas by identifying DS previously unidentified with PSInSAR. Page 25

26 DS are typically identified from homogeneous ground, scattered outcrops, debris flows, noncultivated lands and desert areas. Figure 20 shows a schematic of the breakdown of the distribution of PS and DS over a typical AOI. PS (as identified with the previous algorithm PSInSAR ) usually correspond man-made objects. DS, as described above, are only identified with the latest SqueeSAR algorithm and correspond to homogeneous areas of ground. Satellite signals are not returned over heavily vegetated areas. Figure 20: Schematic showing the distribution of PS and DS over a typical AOI. PS are identified as single objects returning a strong signal to the satellite. DS are homogeneous areas or scattered outcrops. Areas heavily covered by vegetation do not return the satellite signal. SqueeSAR is the only algorithm that TRE offers due the redundancy of the PSInSAR algorithm, providing a significantly increased coverage of ground points, especially over nonurban areas. Figure 21 shows a comparison between the number of ground points identified using the previous PSInSAR algorithm and the latest SqueeSAR algorithm. Figure 21: Comparison between the number of ground points identified using PSInSAR (previous algorithm identifying only PS) and SqueeSAR (latest algorithm identifying both PS and DS). There is a significant increase in the number of identified ground points Page 26

27 SqueeSAR exploits both PS and DS, providing a significantly higher density of ground points and hence coverage of ground displacement over the AOI. A summary of the benefits of SqueeSAR are given below: both PS and DS ground measurement points identified high density of ground points supplied time-series provided for each ground point identified millimetre accuracy on ground displacement values time-series standard deviation reduced compared to previous algorithm i.e. coherence increased and noise decreased increased confidence on ground behaviour due to increased coverage of points especially significant for generic areas with low reflectivity Since its introduction in 2010, as the replacement to the widely accepted PSInSAR algorithm, SqueeSAR has challenged the industry standard by identifying many more ground points, and hence increasing overall understanding of ground displacement occurring in an AOI. 8. Artificial Reflectors (AR) In most cases, PS can be found for land use situations, although PS density and distribution will vary from one application to another. Where PS density can be expected to be too low to provide a reliable definition of ground stability, artificial reflectors (AR) can be deployed. Inherently, it is not possible to reconstruct historical data using artificial reflectors, but it can be used for monitoring uses. Figure 22 shows four different kinds of reflector that have been used in InSAR analyses. The single view reflectors are designed to reflect signals from a single viewing geometry, i.e. ascending or descending. The dual geometry reflector is designed to maximize the radar cross section for both ascending and descending geometries. The rotational orientation of the reflecting surfaces has to be set quite precisely as there is limited room for error. Artificial reflectors are specifically designed to reflect a high proportion of the satellite s signal directly back towards the satellite. This produces a strong and consistent amplitude signal throughout all satellite images, providing coherent data stacks and allowing high accuracy displacement values to be calculated. Page 27