This PDF file contains the front matter associated with SPIE Proceedings Volume 10003, including the Title Page, Copyright information, Table of Contents, and Conference Committee listing.

High-resolution Synthetic Aperture Radar (SAR) data represent an essential resource for the extraction of Ground Control Points (GCP) with sub-metric accuracy without in situ measurement campaigns. Conceptually, SAR-based GCP extraction consists of the following two steps: (i) identification of the same local feature on more SAR images and determination of their range/azimuth coordinates; (ii) spatial 3D positioning retrieval from the 2D radar coordinates, through spatial triangulation (stereo analysis) and inversion methods. In order to boost the geolocation accuracy, SAR images must be acquired from different line of sights, with intersection angles typically wider than 10 degrees, or even in opposite looking directions. In the present study, we present an algorithm specifically designed for ensuring robustness and accuracy in the fully automatic detection of bright isolated targets (steel light poles or towers) even when dealing with opposite looking data takes. In particular, the popular Harris algorithm has been selected as detector because it is the most stable and robust-to-noise algorithm for corners detection on SAR images. We outline the designed algorithmic solution and discusses the results derived over the urban area of Pisa (Italy), where more than ten COSMO-SkyMed Enhanced Spotlight (ES) stereo images are available, thus resulting an optimal test site for an assessment of the performances of the processing chain. The experimental analysis proofs that, assumed timing has been properly recalibrated, we are capable to automatically extract GCP from CSK ES data takes consisting in a very limited number of images.

Ground Control Points (GCP), automatically extracted from Synthetic Aperture Radar (SAR) images through 3D stereo analysis, can be effectively exploited for an automatic orthorectification of optical imagery if they can be robustly located in the basic optical images. The present study outlines a SAR/Optical cross-matching procedure that allows a robust alignment of radar and optical images, and consequently to derive automatically the corresponding sub-pixel position of the GCPs in the optical image in input, expressed as fractional pixel/line image coordinates. The cross-matching in performed in two subsequent steps, in order to gradually gather a better precision. The first step is based on the Mutual Information (MI) maximization between optical and SAR chips while the last one uses the Normalized Cross-Correlation as similarity metric. This work outlines the designed algorithmic solution and discusses the results derived over the urban area of Pisa (Italy), where more than ten COSMO-SkyMed Enhanced Spotlight stereo images with different beams and passes are available. The experimental analysis involves different satellite images, in order to evaluate the performances of the algorithm w.r.t. the optical spatial resolution. An assessment of the performances of the algorithm has been carried out, and errors are computed by measuring the distance between the GCP pixel/line position in the optical image, automatically estimated by the tool, and the “true” position of the GCP, visually identified by an expert user in the optical images.

The development of SAR technology during the last decade has made it possible to collect a huge amount of data over many regions of the world. In particular, the availability of SAR images from different sensors, with metric or sub-metric spatial resolution, offers novel opportunities in different fields as land cover, urban monitoring, soil consumption etc. On the other hand, automatic approaches become crucial for the exploitation of such a huge amount of information. In such a scenario, especially if single polarization images are considered, the main issue is to select appropriate contextual descriptors, since the backscattering coefficient of a single pixel may not be sufficient to classify an object on the scene. In this paper a comparison among three different approaches for contextual features definition is presented so as to design optimum procedures for VHR SAR scene understanding. The first approach is based on Gray Level Co- Occurrence Matrix since it is widely accepted and several studies have used it for land cover classification with SAR data. The second approach is based on the Fourier spectra and it has been already proposed with positive results for this kind of problems, the third one is based on Auto-associative Neural Networks which have been already proven effective for features extraction from polarimetric SAR images. The three methods are evaluated in terms of the accuracy of the classified scene when the features extracted using each method are considered as input to a neural network classificator and applied on different Cosmo-SkyMed spotlight products.

Fires exacerbated during El Niño Southern Oscillation are a serious threat in Indonesia leading to the destruction and degradation of tropical forests and emissions of CO2 in the atmosphere. Forest structural changes which occurred due to the 1997-1998 El Niño Southern Oscillation in the Sungai Wain Protection Forest (East Kalimantan, Indonesia), a previously intact forest reserve have led to the development of a range of landcover from secondary forest to areas dominated by grassland. These structural differences can be appreciated over large areas by remote sensing instruments such as TanDEM-X and LiDAR that provide information that are sensitive to vegetation vertical and horizontal structure. One-point statistics of TanDEM-X coherence (mean and CV) and LiDAR CHM (mean, CV) and derived metrics such as vegetation volume and canopy cover were tested for the discrimination between 4 landcover classes. Jeffries-Matusita (JM) separability was high between forest classes (primary or secondary forest) and non-forest (grassland) while, primary and secondary forest were not separable. The study tests the potential and the importance of potential of TanDEM-X coherence and LiDAR observations to characterize structural heterogeneity based on one-point statistics in tropical forest but requires improved characterization using two-point statistical measures.

Nowadays a well-established tool for Earth remote sensing is represented by Spaceborne synthetic aperture radars (SARs) operating at L-band and above that offers a microwave perspective at very high spatial resolution in almost all-weather conditions. Nevertheless, atmospheric precipitating clouds can significantly affect the signal backscattered from the ground surface on both amplitude and phase, as assessed by numerous recent works analyzing data collected by COSMO-SkyMed (CSK) and TerraSAR-X (TSX) missions. On the other hand, such sensitivity could allow detecting and quantifying precipitations through SARs. In this work, we propose an innovative processing framework aiming at producing X-SARs precipitation maps and cloud masks. While clouds masks allow the user to detect areas interested by precipitations, precipitation maps offer the unique opportunity to ingest within flood forecasting model precipitation data at the catchment scale. Indeed, several issues still need to be fully addressed. The proposed approach allows distinguishing flooded areas, precipitating clouds together with permanent water bodies. The detection procedure uses image segmentation techniques, fuzzy logic and ancillary data such as local incident angle map and land cover; an improved regression empirical algorithm gives the precipitation estimation. We have applied the proposed methodology to 16 study cases, acquired within TSX and CSK missions over Italy and United States. This choice allows analysing different typologies of events, and verifying the proposed methodology through the available local weather radar networks. In this work, we will discuss the results obtained until now in terms of improved rain cell localization and precipitation quantification.

We evaluate the usefulness of X-band, airborne (helicopter) data for tomography over forestry regions and discuss the use of compressive sensing algorithms to aid X-band airborne tomography. This work examines if there is any information that can be gained from forest volumes when analysing forestry sites using X-band data. To do so, different forest scenarios were simulated and a fast SAR simulator was used to model airborne multipass SAR data, at X-band, with parameters based on Leonardo's PicoSAR instrument. Model simulations considered varying factors that affect the height determination when using tomography. The main parameters that are considered here are: motion errors of the platform, the spacing of the flight paths, the resolution of the SAR images and plant life being present under the canopy (an understory). It was found that residual motion errors from the airborne platform cause the largest error in the tomographic profile.

We present a general method for retrieving time-series of three component surface velocity field vector given a set of continuous synthetic aperture radar (SAR) acquisitions collected from multiple geometries. Our algorithm extends the single-line-of-sight mathematical framework developed for time-series analysis using interferometric SAR (InSAR) to three spatial dimensions. The inversion is driven by a design matrix corresponding to a dictionary of displacement functions parameterized in time. The resulting model minimizes a cost function using a non-regularized least-squares method. We applied our method to Rutford ice stream (RIS), West Antarctica, using a set of 101 multi-track multi-angle COSMO-SkyMed displacement maps generating azimuth and range pixel offsets.

Multi-temporal InSAR (MTI) applications pose challenges related to the availability of coherent scattering from the ground surface, the complexity of the ground deformations, the atmospheric artifacts, the visibility problems related to the ground elevation. Nowadays, several satellite missions are available providing interferometric SAR data at different wavelengths, spatial resolutions, and revisit time. A new interesting opportunity is provided by Sentinel-1 mission, which has a spatial resolution comparable to previous ESA C-band missions, and revisit times reduced to up to 6 days. It is envisioned that, by offering regular, global-scale coverage, improved temporal resolution and freely available imagery, Sentinel-1 will guarantee an increasing use of MTI for ground displacement investigations. According to these different SAR space-borne missions, the present work discusses current and future opportunities of MTI applications to ground instability monitoring. Issues related to coherent target detection and mean velocity precision will be addressed through a simple theoretical model assuming backscattering mechanisms related to point scatterers. The paper also presents an example of multi-sensor ground instability investigation over the site of Marina di Lesina, Southern Italy, a village lying over a gypsum diapir, where a hydration process, involving the underlying anhydride, causes a smooth uplift pattern affecting the entire village area, and the formation of scattered sinkholes. More than 20 years of MTI SAR data have been used, coming from both legacy ERS and ENVISAT missions, and last-generation Radarsat-2, COSMO-SkyMed, and Sentinel-1A sensors.

In volcanic areas, where it could be difficult to gain access to the most critical zones for carrying out direct surveys, remote sensing proved to have remarkable potentialities to follow the evolution of lava flow, as well as to detect slope instability processes induced by volcanic activity. By exploiting SAR and optical data a methodology for observing and quantifying eruptive processes was developed. The approach integrates HR optical images and SAR interferometric products and can optimize the observational capability of standard surveillance activities based on in-situ video camera network. A dedicated tool for mapping the evolution of the lava field, using both ground-based and satellite data, was developed and tested to map lava flows during the 2011-2015 eruptive activities. Ground based data were collected using the permanent ground NEtwork of Thermal and VIsible Sensors located on Mt. Etna (Etna_NETVIS) and allowed to downscale the information derived from satellite data and to integrate the satellite datasets in case of incomplete coverage or missing acquisitions. This work was developed in the framework of the EU-FP7 project “MED-SUV” (MEDiterranean SUpersite Volcanoes).

The precision and accuracy of Digital elevation model and deformation measurement, from SAR interferometry (InSAR/DInSAR) depend mainly on the quality of the interferogram. However, the phase noise, which is mainly due to decorrelation between the images and the speckle, makes the step of phase unwrapping most delicate. In this paper, we propose a filtering method that combines the techniques of decomposition into sub-bands and nonlinear local weights. The Spectral / Contextual filter that we propose, inspired from to Goldstein filter is applied to the sub-bands from the wavelet decomposition. To validate the results, we applied to interferometric products tandem pair ERS1/ERS2 taken in the region of Algiers Algeria.

This work has been focused on the integration of microwave data coming from different sensors (SMAP, Sentinel-1, AMSR2), in order to obtain an improved estimation of hydrological parameters and in particular of the Soil Moisture (SMC). The failure of radar sensor in SMAP satellite induced to look for other available microwave frequencies, both from active (e.g. Sentinel-1, C band) and passive sensors (e.g. AMSR2, from C to Ka bands). The contribution of higher frequencies, which are more suitable for monitoring vegetation and other surface features, has been evaluated for compensating the vegetation and roughness effects on the SM retrieval accuracy. In particular, the contribution of C band SAR from Sentinel-1 and of the multi-frequency AMSR2 radiometer (C to Ka bands) have been considered. Moreover, a disaggregation technique, based on the Smoothing filter based intensity modulation (SFIM), enabled achieving a more suitable ground resolution for comparing SAR and radiometric data. Disaggregated microwave data were used as inputs of a retrieval algorithm based on Artificial Neural Network (ANN), able to exploit the synergy between active and passive acquisitions. The algorithm was defined basing on the data available from the SMEX02 and SMAPVEX12 campaigns, and on data simulated by electromagnetic forward models. The algorithm validation on these datasets returned encouraging results, while the disaggregation technique enabled adapting the algorithm to work with Sentinel, SMAP and AMSR2 synergic acquisitions

In mountain regions and their forelands, glaciers are key source of melt water during the middle and late ablation season, when most of the winter snow has already melted. Furthermore, alpine glaciers are recognized as sensitive indicators of climatic fluctuations. Monitoring glacier extent changes and glacier surface characteristics (i.e. snow, firn and bare ice coverage) is therefore important for both hydrological applications and climate change studies. Satellite remote sensing data have been widely employed for glacier surface classification. Many approaches exploit optical data, such as from Landsat. Despite the intuitive visual interpretation of optical images and the demonstrated capability to discriminate glacial surface thanks to the combination of different bands, one of the main disadvantages of available high-resolution optical sensors is their dependence on cloud conditions and low revisit time frequency. Therefore, operational monitoring strategies relying only on optical data have serious limitations. Since SAR data are insensitive to clouds, they are potentially a valid alternative to optical data for glacier monitoring. Compared to past SAR missions, the new Sentinel-1 mission provides much higher revisit time frequency (two acquisitions each 12 days) over the entire European Alps, and this number will be doubled once the Sentinel1-b will be in orbit (April 2016). In this work we present a method for glacier surface classification by exploiting dual polarimetric Sentinel-1 data. The method consists of a supervised approach based on Support Vector Machine (SVM). In addition to the VV and VH signals, we tested the contribution of local incidence angle, extracted from a digital elevation model and orbital information, as auxiliary input feature in order to account for the topographic effects. By exploiting impossible temporal transition between different classes (e.g. if at a given date one pixel is classified as rock it cannot be classified as glacier ice in a following date) we here propose an innovative post classification correction based on SVM classification probabilities. Optical data, i.e. Landsat-8 and Sentinel-2, have been employed, when available, for training sample collection. Detailed field observations from two glaciers located in the Ortles-Cevedale massif (Eastern Italian Alps) have been employed for validation.

The aim of this work is to exploit the potential of AMSR2 for hydrological applications on a regional scale and in heterogeneous environments characterised by different surface covers at subpixel resolution. The soil moisture content (SMC) estimated from Advanced Microwave Scanning Radiometer 2 (AMSR2) through the ANN-based “HydroAlgo” algorithm is firstly compared with the outputs of the Soil Water Balance hydrological model (SWBM). The comparison is performed over Italy, by considering all the available overpasses of AMSR2, since July 2012. The SMC generated by HydroAlgo is then considered as input for generating a rainfall product through the SM2RAIN algorithm. The comparison between observed and estimated rainfall in central Italy provided satisfactory results with a substantial room for improvement. In this work, the ANN “HydroAlgo” algorithm [1], which was originally developed for AMSR-E, was adapted and re-trained for AMSR2, accounting for the two C band channels provided by this new sensor. The disaggregation technique implemented in HydroAlgo [2], devoted to the improvement of ground resolution, made this algorithm particularly suitable for the application to such a heterogeneous environment. The algorithm allows obtaining a SMC product with enhanced spatial resolution (0.1°), which is more suitable for hydrological applications. The AMSR2 derived SMC is compared with simulated data obtained from the application of a well-established soil water balance model [3]. The training and test of the algorithm are carried out on a test area in central Italy, while the entire Italy is considered for the validation. The last step of the activity is the use of the HydroAlgo SMC into the SM2RAIN algorithm [4], in order to exploit the potential contribution of this product at enhanced resolution for rainfall estimation. [1] E. Santi, S. Pettinato, S. Paloscia, P. Pampaloni, G. Macelloni, and M. Brogioni (2012), “An algorithm for generating soil moisture and snow depth maps from microwave spaceborne radiometers: HydroAlgo”, Hydrology and Earth System Sciences, 16, pp. 3659-3676, doi:10.5194/hess-16-3659-2012. [2] E. Santi (2010), “An application of SFIM technique to enhance the spatial resolution of microwave radiometers”, Intern. J. Remote Sens., vol. 31, 9, pp. 2419-2428. [3] L. Brocca, S. Camici, F. Melone, T. Moramarco, J. Martinez-Fernandez, J.-F. Didon-Lescot, R. Morbidelli (2014), “Improving the representation of soil moisture by using a semi-analytical infiltration model”, Hydrological Processes, 28(4), pp. 2103-2115, doi:10.1002/hyp.9766. [4] Brocca, L., Ciabatta, L., Massari, C., Moramarco, T., Hahn, S., Hasenauer, S., Kidd, R., Dorigo, W., Wagner, W., Levizzani, V. (2014). Soil as a natural rain gauge: estimating global rainfall from satellite soil moisture data. Journal of Geophysical Research, 119(9), 5128-5141, doi:10.1002/2014JD021489.

In this work, a comparison between soil moisture products derived from satellite and land model data was performed; in particular, the soil moisture retrievals of SMOS and ASCAT were compared with those of the ERA-Interim/Land model, produced by the ECMWF in a timeframe of 3 years. Subsequently, for a limited period of time, the product from the SMAP radiometer was joined to SMOS, ASCAT and ERA-Interim model data as a fourth dataset. In both cases, the whole H-SAF region of interest, which includes Northern Africa and Europe, was analysed. In order to validate the products, the Triple Collocation technique was applied to estimate the independent error standard deviation of three systems that observe the same target parameter. When more than three datasets were available, the Quadruple Collocation technique was used to jointly estimate the error standard deviation of four sources. Moreover, when the SMOS and SMAP radiometer products were considered, the Extended Collocation was adopted in order to evaluate the error variances of the systems, taking into account the possible presence of an error cross-correlation between the radiometer retrievals.

This paper focuses on the generation of Digital Elevation Models (DEMs) with COSMO SkyMed Spotlight data in providing DEMs. In particular, the peculiarity of Spotlight data (affected from Doppler centroid drift) is investigated, and the use of the processing chain included in the Delft Object-oriented Radar Interferometric Software (DORIS [1]). The effects of not correctly handled Doppler drift is shown. The standard interferometric processing, without Doppler drift handling, has been applied to Spotlight image pairs, resulting in interferometric coherence loss in interferograms as we move away from scene center. So, the standard processing chain has been modified to take in account the Doppler centroid drift affecting Spotlight data and very high resolution and accuracy DEMs have been obtained. Some Spotlight image pairs have been processed and the obtained DEMs have been shown and analyzed proving the high details and product accuracy.

Evolution of the damping ratio for Bragg wavenumbers in the range 32-43 rad/m is evaluated for oil slicks of different composition released in the open ocean and allowed to develop naturally. The study uses quad-polarimetric L-band airborne synthetic aperture radar data acquired over three mineral oil emulsion releases of different, known oil-to-water ratio, and a near-coincident release of 2-ethylhexyl oleate that served as a biogenic look-alike. The experiment occurred during the 2015 Norwegian oil-on-water exercise in the North Sea during a period of relatively high winds (~12 m/s). NASA’s Uninhabited Aerial Vehicle Synthetic Aperture Radar (UAVSAR) was used to repeatedly image the slicks over a period of eight hours, capturing the slicks’ early development and providing a time series from which to track the evolution of the slicks’ size, position, and radiometric characteristics. Particular emphasis is given in this analysis to identification of zones of higher damping ratio within the slicks (zoning) as potential indicators of thicker oil, and to comparison of the evolution of emulsion and plant oil damping ratios. It was found that all mineral oil slicks initially exhibited zoning apparent in VV, HH, and HV intensities, and that the areas of higher damping ratio persisted the longest for the highest oil content emulsion (80% oil by volume). In contrast, zoning was not unambiguously evident for plant oil at any time from 44 minutes to 8.5 hours after release.

Ground based Synthetic Aperture Radar Interferometry (GBInSAR) can set the right baseline according to the monitoring area, which is more freedom and convenient maneuverability and has been used on deformation monitoring of buildings, Bridges, etc. However, in the process of GBInSAR monitoring, atmospheric disturbances is a key factor affected the accuracy of the monitoring. According to the statistical characteristics in monitoring area images, the paper selected the Permanent Scatters (PS) points, established Delaunay triangulation and acquired the atmospheric correction of GBInSAR monitoring data for by linear interpolation. As the same time, the paper used the fixed point method to calculate the atmospheric disturbance. Then, the paper compared the both methods’ corrected results with high precision vertical data. The results showed that the PS method was better than the fixed point method. The PS method can better correct the atmospheric disturbances elements in the GBInSAR monitoring sequence, which provided reference value for further popularization and application of GBInSAR.

Rheticus® is an innovative cloud-based data and services hub able to deliver Earth Observation added-value products through automatic complex processes and, if appropriate, a minimum interaction with human operators. In this paper, we outlines the capabilities of the “Rheticus® Displacement” service, designed for geohazard and infrastructure monitoring through Multi-Temporal SAR Interferometry techniques.

Structural health monitoring (SHM) of engineered structures consists of an automated or semi-automated survey system that seeks to assess the structural condition of an anthropogenic structure. The aim of an SHM system is to provide insights into possible induced damage or any inherent signals of deformation affecting the structure in terms of detection, localization, assessment, and prediction. During the last decade there has been a growing interest in using several remote sensing techniques, such as synthetic aperture radar (SAR), for SHM. Constellations of SAR satellites with short repeat time acquisitions permit detailed surveys temporal resolution and millimetric sensitivity to deformation that are at the scales relevant to monitoring large structures. The all-weather multi-temporal characteristics of SAR make its products suitable for SHM systems, especially in areas where in situ measurements are not feasible or not cost effective. To illustrate this capability, we present results from COSMO-SkyMed (CSK) and TerraSAR-X SAR observations applied to the remote sensing of engineered structures. We show how by using multiple-geometry SAR-based products which exploit both phase and amplitude of the SAR signal we can address the main objectives of an SHM system including detection and localization. We highlight that, when external data such as rain or temperature records are available or simple elastic models can be assumed, the SAR-based SHM capability can also provide an interpretation in terms of assessment and prediction. We highlight examples of the potential for such imaging capabilities to enable advances in SHM from space, focusing on dams and cultural heritage areas.

For centuries the promontory of Capo Colonna in Calabria region, southern Italy, experienced land subsidence and coastline retreat to an extent that the archaeological ruins of the ancient Greek sanctuary are currently under threat of cliff failure, toppling and irreversible loss. Gas extraction in nearby wells is a further anthropogenic element to account for at the regional scale. Exploiting an unprecedented satellite Synthetic Aperture Radar (SAR) time series including ERS-1/2, ENVISAT, TerraSAR-X, COSMO-SkyMed and Sentinel-1A data stacks acquired between 1992 and 2016, this paper presents the first and most complete Interferometric SAR (InSAR) baseline assessment of land subsidence and coastal processes affecting Capo Colonna. We analyse the regional displacement trends, the correlation between vertical displacements with gas extraction volumes, the impact on stability of the archaeological heritage, and the coastal geohazard susceptibility. In the last 25 years, the land has subsided uninterruptedly, with highest annual line-of-sight deformation rates ranging between -15 and -20 mm/year in 2011-2014. The installation of 40 pairs of corner reflectors along the northern coastline and within the archaeological park resulted in an improved imaging capability and higher density of measurement points. This proved to be beneficial for the ground stability assessment of recent archaeological excavations, in an area where field surveying in November 2015 highlighted new events of cliff failure. The conceptual model developed suggests that combining InSAR results, geomorphological assessments and inventorying of wave-storms will contribute to unveil the complexity of coastal geohazards in Capo Colonna.