displacement, a parametric model like a linear rate of change or a seasonal oscillation is assumed. Preliminary estimation of displacement parameters has the clear disadvantage that a functional model has to be postulated, chosen either from a priori knowledge or by statistically testing the performance of different models (van Leijen and Hanssen, 2007). Whereas preliminary displacement estimation is included in most PS processing chains, it is disapproved by Hooper et al. (2004, 2007) who only estimate per PS an absolute height error δh (or look angle error, respectively) and a spatially uncorrelated contribution of the master image. Assuming the relative displacement of nearby PS to be small, their approach does not rely on the preliminary choice of a specific functional model. However, in contrast to other approaches, large spatial displacement gradients may be critical for successful unwrapping. The most critical issue inherent to all unwrapping approaches is the validity of the assumption that phase differences of PS adjacent in time or space are smaller than π. Taking into account the unseizable noise contribution, it is rather desirable that systematic signal components differ by distinctly less than π. Preliminary mitigation of atmospheric and orbital signals is usually neither performed nor required, because their local gradients are mostly sufficiently small. However, even though this assumption applies to the great majority of applications, it might prove invalid in some cases with very large orbit errors, where only a preliminary estimation enables successful unwrapping.
The relativistic phase and time offsets from this paper are not only of high importance for DEM generation with a formation flying SAR cross-track interferometer. Formations with multiple satellites have also been suggested for a wide range of further remote sensing applications, ranging from along-track interferometry for moving object and ocean current measurements over sparse aperture ambiguity suppression and super resolution for enhanced high-resolution wide-swath SAR imaging up to single-pass SAR tomography for vertical struc- ture measurements –. Due consideration of relativistic effects from varying along-track baselines is again of essential importance for these advanced bistatic and multistatic SAR systems to avoid mutual range and phase offsets between the received SAR signals. The phase accuracy requirements for the combination of the different receiver signals are typically in the order of 1° or a few degrees. For comparison, an along-track baseline of 100 m causes in an X-band system a relativistic phase shift in the order of several tens of degrees. Future multistatic SAR satellite missions should therefore take into account relativistic effects in the design of the radar synchronization system and/or the SAR processor to avoid a possible performance loss.
With respect to measurements of sea surface wind field in TCs using scatterometer or SAR, two major sources may limit the accuracy of retrieval for high winds: (1) deficiencies of the Geophysical Model Function (GMF) for high winds, as presented in [ 7 ]. Improvement of GMF, such as CMOD5 [ 8 ] is dedicated for retrieval of high wind using scatterometer or SAR data, has somewhat reduces this error sources for inversion of sea surface wind field in hurricane scale [ 9 ]. However, one still faces the problem of speed ambiguity when applying CMOD5 for retrieving high winds [ 10 ] and saturation or damping of radar signal under severe weather conditions [ 11 ]. (2) Effects of heavy rains on radar signal. Microwave signals are likely to suffer effect of heavy rains which are permanent features in TCs and therefore errors are induced of deriving sea surface wind speed, e.g., studies presented by Quilfen et al. [ 7 ] and Weissman et al. [ 12 ]. Yueh et al. [ 13 ] proposed an updated GMF for retrieval of sea surface wind field considering the rain rate as a parameter, which is applied to the hurricane Floyd with maximum wind speed reaching 60 m/s showing a good agreement NOAA Hurricane Research Division (HRD) wind reanalysis.
Having defined the rationale and the objectives of the research, this chapter will present the background relevant to the project. Section 2.1 is dedicated to a brief introduction to SAR remote sensing. Subsequent sections are dedicated to monostatic SAR (Section 2.2) and bi-static SAR (section 2.3). The physics related to SAR imaging are presented in section 2.4 (SAR image interpretation). Section 2.5 presents SAR interferometry describing the basic principles and introducing all the key concepts that provide the essential background to the analysis carried out in the chapter dealing with possible fields of applications. The notion of signal coherence is analysed in section 2.6 while the following section is dedicated to the principle of SAR image processing. Multi- static SAR configurations are discussed in section 2.8 while section 2.9 and the remaining part of the chapter present a literature review on Earth tides, tropospheric and ionospheric effects on SAR imaging.
In Europe, one of the first published GEO SAR concepts was by Prati et al.  in 1998. They described a bistatic passive radar reusing L-band broadcast signals. Such a system could achieve 120-m spatial resolution using an antenna with a diame- ter 4.8 m. The orbit inclination is small (satellite motion of only 25 km from the geostationary position is assumed). However, a long integration time of up to 8 h is required to form a satisfac- tory image. Imaging effects of clutter and partially stable tar- gets, as well as measuring the atmospheric phase screen (APS) are noted. Research on other GEO SAR concepts (mainly con- ventional monostatic) has continued with contributions from Cranfield –, Milan –, and Barcelona ,  in particular. These recent studies have made significant contribu- tions in the areas of system design and APS estimation/phase compensation. For the low inclination orbits and modest an- tenna sizes, which these authors have assumed, integration times are relatively long, and thus, atmospheric phase com- pensation is needed. There has been particular interest again in applications for short repeat period interferometry related to geohazards.
The performance of the position acquisition systems is particularly relevant for dynamic pro- cesses in 3D space. The systems are compared on the basis of the quality of SAR images using a simple measurement scenario shown in Fig. 3 and realistic trajectories as shown in Fig. 4. In order to avoid interferences from buildings or trees (multipath effects), the measurements were carried out on a large open area. The minimum and maximum distance between the total station and the prism was 4 m and 35 m, respectively. As reference 20 reflectors have been placed on the surface as shown in Fig. 3. The UAS was manually steered to perform 12 stripmap
try (DInSAR) has shown its capability in monitoring ground displacement caused by the freeze-thaw cycle in the active layer of permafrost regions. However, the unique landscape in the dis- continuous permafrost zone increases the difficulty of applying DInSAR to detect ground displacements. In this study, datasets from two radar systems, X-band TerraSAR-X and L-band ALOS PALSAR, were used to evaluate the influencing factors and ap- plication conditions for DInSAR in the discontinuous permafrost environment based on a large number of analyzed interferograms. Furthermore, the impact of different DEMs on the application of DInSAR was illustrated by comparing the high-resolution LiDAR- DEM, TanDEM-X DEM, and SRTM DEM. The results demon- strate that temporal decorrelation and strong volume decorrelation in areas with developed vegetation highly constrains the application of X-band data. In terrain with more developed vegetation (such as shrubs and spruce), the X-band differential phase becomes linked to the canopy rather than the topography, whereas L-band data show promising results in retrieving topography-related displace- ment. By comparing the displacement velocity maps of the two sensors and referencing in situ measurements, we demonstrated that the ALOS PALSAR results capture the permafrost-induced terrain movement characteristics and values in the correct range. Moreover, the influence of soil moisture and vegetation phenol- ogy on the accuracy of displacement retrievals using the L-band data are illustrated and discussed. The analyses confirm that the L-band has strong advantages over the X-band in monitoring dis- placements in discontinuous permafrost environments.
In this contribution a polarimetric side-looking syntheticapertureradar (SAR) mounted on a unmanned aerial vehicle (UAV) is presented and discussed with respect to the detec- tion and localization of landmines. As an example for an anti-personal mine a PFM-1 which contains an elongated aluminium rod was considered. Such anisotropic geometries exibit a polarization dependend radar cross section (RCS). Through a special configuration of three antennas, polarimet- ric SAR measurements involving a back projection algorithm could be implemented. This concept allows for the detection and furthermore the classification of such anisotropic objects. First field tests using a tachymeter for localization of the UAV over a snow covered meadow successfully demonstrated the performance by the detection of small metal rods depend- ing on their orientation with respect to the flight path of the UAV. These experimental results were supported by simula- tions expressing the necessity of polarimetric measurements in combination with a distinct flight trajectory for a robust detection of certain landmines.
ﬂäche, aber im Falle von vergrabenen oder versteckten Minen versagt diese Art von Sensor. Die Autoren von [ 15 ] verwenden Fluxgate-Vektor-Magnetometer, die auf einem UAS montiert sind, um UXOs zu erkennen. Die Herausforderung für Magnetometer und Metalldetektoren ist vor allem die zuverlässige Detek- tion von Plastikminen mit geringem Metallanteil. In [ 16 ]–[ 19 ] wird ein nach unten gerichtetes GPR zur Minendetektion untersucht. So kann das Radar die Tiefe eines Ziels direkt messen. Der Flächendurchsatz, also die zu untersuchen- de Fläche pro Zeit, ist jedoch gering. Das Untersuchungsgebiet muss parallel dazu in geraden Linien (B-Scans) gescannt werden, damit ein hochauﬂösendes 3D-Bild (C-Scan) erzeugt werden kann. Zudem können aufgrund der starken Bodenreﬂexion bei senkrechtem Einfall der elektromagnetischen Welle schwa- che Ziele, die sich auf oder nur wenige Zentimeter unter der Oberﬂäche beﬁnden, eventuell nicht detektiert werden.
A closer investigation showed that a lot of irregular sea surface features are due to atmospheric phenomena in particular at very low wind speeds. This eﬀect was investigated using collocated ECMWF wind speeds for comparison. Fig. 8.3 (A) shows a scatterplot of the inhomogeneity parameter as a function of wind speed with the threshold introduced above indicated by a horizontal line. The plot in Fig. 8.3 (B) gives the respective percentage of non-homogeneous imagettes (solid line). One can clearly see, that the ratio increases strongly as the wind speed drops below 5 ms −1 . The phenomenon is attributed to the fact that due to surface tension the response of Bragg waves, which dominate the radar return (compare Chapter 4), to the wind is not continuous at very low wind speed. In other words a minimum wind speed is required to produce centimetre waves and these waves disappear abruptly as the wind speed drops below a certain limit. The inhomogeneous image features are thus in many cases believed to be due to small spatial variations of the wind speed around these thresholds. The dashed line in Fig. 8.3 (B) indicates the percentage of collocated ERS-2 scatterometer measurement, which are ﬂagged as unusable. As the dashed line is below the solid line, it seems that a signiﬁcant number of scat- terometer measurements is taken as usable, although the NRCS of the sea surface is strongly inhomogeneous. SAR data thus have the potential to improve the quality of scatterometer measurements as well.
RT2 is a radiative transfer model simulator developed by ESA under task 4, contract 10644/93/NL/NB (Knight, 1997), with the scientific goal of assisting the study of radar signatures from the interaction with soil and canopy. The program is run from a Sun Operating System and consists of three modules: rtsetup, RT2, and read_and_write. The first module, rtsetup, is used to create a model input file (extension .rtm) to be used by the RT2 module. The graphical user interface of rtsetup is shown in Figure 4-31. This allows the user to define the radar, soil and surface scattering model parameters. As the study concentrates on bare soils, the canopy layers parameters are not used. In addition to the ground roughness and correlation length, it is possible to define the volumetric soil moisture and the soil morphological structure, or soil texture, defined by percentage values of three soil components: clay, silt and sand. This parameter designates the proportionate distribution of the different sizes of mineral particles in a soil, excluding any organic matter. According to their size, these mineral particles are grouped into separates which can be classified following the US Department of Agriculture system of nomenclature as shown in Figure 4-32. Since various sizes of particle have quite different physical characteristics, the nature of mineral soils is determined to a remarkable degree by the particular separate that is present in larger amounts (USDA, 2007). Together with the volumetric soil moisture, soil texture is in fact used to compute the susceptibility characteristics of the soil which affect both radar backscatter amplitude and phase.
b SAR image formed using a 2 m sub-aperture, showing the target off-centre from the central broadside position
The output from the artefact localisation model is superimposed onto the SAR images in Fig. 4. The model has yielded an accurate localis- ation of the vibration artefacts in the experimental SAR images. Conclusion: A detailed analysis of the localisation of the phenomena produced when imaging vibrating targets in SAR imagery has been presented. It was shown how synthetic micro-Doppler is equivalent to the micro-Doppler described in the literature. This equivalence allowed the measurement of vibrating target paired echoes correspond- ing to arbitrary syntheticradar platform speeds, using a low-speed lab- oratory scanner.
The detection of sea surface phenomena by SAR is de- pendent on the availability of wind , which is present in oceanic scenarios most of the time. Even very low wind speeds, about 2 m/s, are sufficient to create rip- ple waves, centimetre-sized distortions of the otherwise smooth water surface . The characteristics of the radar echo averaged over a subscene can directly be re- lated to wind speed and local variation of intensity can be related to wave height by image spectra analysis . When swell waves, reaching wavelengths of up to sev- eral hundred meters, and ripple waves are present, the swell waves are visible on the SAR image as regular brightness modulations. Too strong local wind speeds, however, may cause white capping and wave breaking, which is seen as strong smearing in the SAR image, so the individual wave crests can no longer be discri- mated. Hence, detecting wavelengths is possible for low to medium wind speeds only. Very shallow waters lead to increased wave steepening and finally breaking, which means depths below about 10 m are also not suitable for this type of analysis.
the sparse reconstruction process in tomography with syntheticapertureradar. Two hyperparameter-free approaches are introduced into the framework of SL1MMER (Scale-down by L1 norm Minimization, Model selection, and Estimation Reconstruction). By means of numerical simulations, we evaluate their performance regarding mean and standard deviation of elevation estimates, as well as detection rate. Preliminary results with real data are also provided.
Subsequent to the first demonstration of SAR tomography [Pasq 95,Home 96,Reig 00], several extensions and alternatives have been put forward in order to attain low side- lobe and ambiguity levels with a reduced number of irregular passes. The use of adap- tive spectral estimators was introduced in [Gini 02,Lomb 03,Lomb 09a,Lomb 09b] and further developed in [Saue 11, Frey 11a] (see also [Frey 11b]). In addition, subspace- based spectral estimators, such as the multiple signal classification (MUSIC) algorithm, have been recently employed [Guil 05, Nann 09, Frey 11a, Frey 11b, Huan 12, Lomb 13]. In [Forn 03], the authors formulated the tomographic inversion under the framework of linear inverse problems, thus exploiting the truncated singular-value decomposi- tion (TSVD). Also, a maximum a posteriori estimator was developed in [Zhu 10b]. Other publications have addressed irregular geometries by means of interpolation techniques (see, for example, [Lomb 08a, dAle 12]). Alternatively, an extension of SAR interferometry [Baml 98] from a parametric perspective was proposed in [Teba 10a]. In a nutshell, this last work employs covariance matching estimation techniques in order to estimate the effective scattering center of different scattering mechanisms (SMs), along with their backscattered power (see also [Lomb 98a, Cors 99, Bess 00]). Moreover, the author in [Clou 06] introduced the concept of polarization coherence tomography (PCT). Basically, the method exploits the variation of the interferomet- ric coherence with polarization to estimate ground topography and height of vege- tation layers. Then, it uses these parameters to represent a backscatter profile as a Fourier–Legendre series. Finally, sparsity-based inversion techniques were introduced in [Budi 09, Zhu 10a, Budi 11, Zhu 12, Agui 12a]. In essence, the authors applied and further developed the relatively new compressed sensing (CS) theory to achieve super- resolution imaging of vertically-sparse targets.
 J.B. Billingsley, Low-angle Radar Land Clutter: Meas- urements and Empirical Models, William Andrew Pub., 2002.
 S. Hobbs, C. Seynat, and P. Matakidis, “Videogram- metry: A practical method for measuring vegetation mo- tion in wind demonstrated on wheat,” Agricultural and Forest Meteorology, vol. 143, no. 3-4, pp. 242–251, 2007.
Such ambiguities are usually resolved during phase unwrapping, which exploits spatial correlations between the height values of natural topography , , . The accuracy of this absolute phase (or height) reconstruction process depends on several factors like the signal-to-noise ratio, the surface and volume decorrelation, the ground resolution, and, most important, the actual terrain itself. The latter may strongly limit the useful baseline length for rough terrain like deep valleys, isolated peaks, tall forests, or mountains with steep slopes. On the other hand, large baselines are desired to achieve a sensitive radar interfer- ometer with a good phase-to-height scaling. This dilemma becomes especially pronounced for future radar sensors, which will provide a high range bandwidth and enable coherent data acquisitions with long interferometric baselines. To illustrate this problem, we choose for the aforementioned satellites just 10% of the critical baseline. The corresponding baseline lengths vary between several hundred meters for TerraSAR-X and several kilometers for Tandem-L. The spatial decorrelation will be small in this case and can be removed by range filtering . The corresponding heights of ambiguity are shown in Fig. 6 on the right. It becomes clear that the ambiguous heights are rather low in this case, which may cause irresolvable height errors in areas with rough terrain. It is hence in general not possible to take full advantage of the opportunity for large baseline acquisitions in these high- bandwidth radar systems.
To investigate the influences of motion errors or evaluate the performances of moti- on compensation algorithms, realistic and complete motion parameters of the vehicle dynamics are required. One way to obtain the motion parameters is to purchase them from professional vehicle dynamics measurements company, such as Drivability Tes- ting Alliance. However, the exact measurement of the vehicle performance in dri- ving tests is a nontrivial problem. As depicted in Figure A.1, high accuracy dynamics measurements of ground vehicle need various kinds of sensors, such as acceleration sensor, fiber optic gyroscopes, wheel vector sensor, laser sensor, GPS-based speed and position sensor, etc., thereby being expensive to carry out. Furthermore, such measu- rements are specific for a certain vehicle under certain circumstances instead of the whole spectrum of the vehicle dynamics, which is preferred for a guidance research. On the other hand, in the standards ISO 2631  and VDI 2057 , the guidelines of the effects of exposure to vibration on humans have been presented, which cover a very wide frequency bandwidth and amplitude range. Therefore it is reasonable to use such standards for designing motion parameters for the simulations with regard to evaluating the influences of motion errors and performances of the corresponding motion compensation algorithms.
The penetrating capability through foliage and even ground allows underneath dielectric mate- rials to be visible in SAR imagery. The microwave with a long wavelength is able to penetrate into the vegetation and ground to a certain extent. The longer the wavelength, the deeper the microwave penetrates. For example, the X-Band sensor in TerraSAR satellite with wavelength between 2.40cm and 3.75cm attenuates quickly on the surface of ground objects, and there is little penetration. Backscattering is relatively simple because most of it happens on the surface. However, L-Band sensor of PALSAR [ 26 ] can penetrate deep into the vegetation and occasionally into the ground. In this case the backscattering is complicated, which may be contributed from many objects on the wave track. Multiple reflections can occur inside the vegetation or ground. The dielectric properties, conductivities and vegetation densities also influence the penetration. Some advantages of SAR imagery result from the special imaging mechanisms of SAR sensor. The interaction of microwave and ground targets are recorded in the images. SAR imagery is appropriate for object detection because backscattering is determined by the surface conditions of targets. The scattering of microwave on the ground provides valuable information of the geometrical and physical characteristics of ground objects. SAR imagery contains information about the geophysical structures of the observed scene. The conductivity of objects in the scene influences their reflections of radar waves. For example, the soil moisture affects its conductivity. The differences in soil moisture can result in varied image evidence. Similarly, the sensitivity of radar waves to vegetation growth can be exploited for crop monitoring. Furthermore, the sensitivity to moving targets enables the velocity measurement and monitoring from the air. Polarimetric SAR and interferometric SAR provide many promising products which are unavail- able in optical images. Microwave polarization alteration during the interaction depends on the geophysical parameters of the scatterers. Scattering mechanisms can be analyzed using polari- metric decomposition. DEM and surface displacement can be extracted from interferograms of InSAR data.