Science Service System

Summary of Proposal LAN1858

TitleInfrastructure control with TerraSAR-X interferometry along the Danube River in Hungary
Investigator Strozzi, Tazio - Gamma Remote Sensing, N/A
Team Member
Dr. Urs, Wegmüller - Gamma Remote Sensing, N/A
Dr. Grenerczy, Gyula - Institute of Geodesy, Cartography and Remote Sensing, Satellite Geodetic Observatory
SummaryThe proposed activities will be concerned with the investigation of the tailing and river dams associated with various industrial infrastructures in Hungary using a large stack of TSX images acquired since 24/10/2008. The most significant industrial complex lays in the western bank of the Danube River, the artery of Central Europe. We also plan to investigate dykes of the Danube located near the industrial complex. Dam stability investigation related to industrial infrastructure is highly important regarding the red mud reservoirs as well, where waste product of bauxite refining used to be deposited. Burst of one of such red mud reservoirs in 2010 led to the biggest fatal industrial and ecological disaster of the region. The PSI study(Grenerczy and Wegmüller, 2011, 2013) carried out after the disaster with C-band ENVISAT data suggested that proper monitoring of the dam motion may have highlighted the risk of the structural failure and that the disaster might have been prevented. We demonstrated that if such monitoring is omitted, PSI is capable of revealing deformation history posteriorly and can help preventing such occurrences. The large archive of 120 TSX images available since 2008 over this area represents a very valuable source of information specific investigations. These investigations are foreseen within the RESOURCE(REmote Sensing based Operational service for infrastrUctuRe Controland managEment) proposal that was recently submitted as collaborative project to the "Space-based applications at the service of European Society / Pre-operational validation of GMES services and products" Space Call 6 with Consiglio Nazionale delle Ricerche(CNR) Napoli, Italy, as project coordinator. Final aim of the RESOURCE project is to achieve at a robust operational infrastructure monitoring system primarily based on X-band InSAR. Ground-based surveys will be performed for validation.
Detailed reportIntroduction This report describes the study performed with TerraSAR-X Persistent Scatterer Interferometry (PSI) for the detection and assessment of ground movements between 2008 and 2012 for two regions along the Danube River in Hungary including refineries, hydrocarbon storage and distribution facilities, power plants, industrial ports, various linear infrastructures related to transportation, and river banks. The study was performed by Gamma Remote Sensing, Gümligen (Switzerland) in collaboration with the Institute of Geodesy, Cartography and Remote Sensing and the Geological and Geophysical Institute of Hungary (MFGI), both in Budapest (Hungary). Study areas and TerraSAR-X data The two study areas of about 12 km x 12 km each are shown in Image 2 together with the footprint of the TerraSAR-X frame. In the first study region to the south the most significant industrial complex lays in the western bank of the Danube River, the artery of Central Europe. Near the industrial complex there are dykes of the Danube, relevant for stability and waste deposit. In the second study region to the north there are known landslide/riverbank problems. Our analysis was performed with 107 TerraSAR-X images acquired along the descending orbit from 2008 to 2012. The perpendicular baselines and the acquisition time intervals with respect to the geometric reference image of 2010.08.15 are indicated in Image 1. Persistent Scatterer Interferometry (PSI) The Single Look Complex (SLC) TerraSAR-X images were co-registered to the reference geometry of the 2010.08.15 scene and Multi-Look Intensity (MLI) images were computed with 5 looks in slant-range and 5 looks in azimuth. With use of the GAMMA DIFF&GEO (Differential Interferometry and Geocoding) software a topographic reference was prepared with the SRTM DEM. The topographic reference was considered for geocoding and removal of the terrain-induced phase changes in the SAR interferograms. An Interferometric Point Target Analysis (IPTA) was conducted with a stack of 106 TerraSAR-X images, excluding the image of 2010.01.07 because of snow-cover. On the basis of the SAR data, point target candidates were first extracted with use of the GAMMA IPTA software. For the point target candidates interferometric phase were then computed and analyzed. The deviation from a linear (uniform) movement was used as a statistical quality feature. For linear motions, the average movement rate during the observation period could be determined with a statistical accuracy of 1 mm/year. Reference points were selected in areas estimated as stable. IPTA results consist of linear deformation rates and displacement histories in the satellite Line-Of-Sight (LOS) direction. The displacement values indicate the component of the movement in the observation direction of the SAR sensor, i.e. in a direction of about 45° to the zenith and of about 9° to the east. If a displacement occurs in the vertical direction, the observed component is smaller by a factor of 0.71 (cos 45°). In this case, the observed value must be scaled by 1 / 0.79 = 1.41 in order to obtain the full value. Results for study area 1 The IPTA average displacement rate image derived for the study area to the south is shown in Image 3. In this image negative values indicate an increase in the distance from target to satellite or, in general, a lowering of the surface. Significant movements of more than 2 mm/year were detected along roads and railways and on several isolated structures. The coverage with valid information is restricted to the built-up areas. Focusing on the industrial complex a significant movement of several containers could be recorded (Image 4). On the other hand, points around dikes show a pattern of general stability over time, although in certain cases seasonal variability of the deformation was observed (Image 5). Validation of the results for study area 1 The following statements were made after a on-site investigation of 4 containers where 13-15 mm displacement were detected. - Three 500 m3 containers have been refurbished after 2012 (the end of the investigation interval), this included the rainwater drainage collar, the basement, rainwater collector and the concrete pavement around the containers. These all improved the water drainage around the base of the container. Visible problems were not detected and the containers operate normally. - One 5000 m3 container shows clear signs of motion. The pavement around it subsided, the rainwater collection cannot play its role as it is now stands higher than the concrete pavement (Image 6). It is necessary to perform local geodetic stability measurement for the container and it is recommended that the drainage around the base of the container must be fixed as soon as possible. Evaluation of the motion measurements from the Danube Refinery An evaluation of the motion measurements was performed by DS Production MOL. The following statements were made by the Maintenance Engineering of the company: - The data sets from InSAR analysis is applicable for raising awareness of slow and long term changes. - Based on this it is possible to plan engineering geodetic measurement monitoring shorter changes in time. - For the evaluation of the data on-site investigation is always necessary. - Motion of the containers is highly dependent on the level of filling. Results for study area 2 The IPTA average displacement rate image derived for the study area to the north is shown in Image 7. Significant movements of more than 2 mm/year were detected on a former riverbank, which was known for a former landslide event. The coverage with valid information is restricted to the upper part of the slope, while the lower part close to the Danube is without valid information because of vegetation cover. Also more to the south along the river, where the landslide inventory indicates other former events, the coverage with PSI data is very limited because of the vegetation cover. Conclusions Our study performed with TerraSAR-X PSI indicated significant motion of more than a couple of mm/year for parts of roads, railways, industrial complexes, and the Danube riverbank, but no significant motion anywhere except the normal expected signals could be recorded. TerraSAR-X PSI has indeed the capability to be employed for infrastructures monitoring; however, in our study we we could not ensure commercial continuation of this new technology.

Back to list of proposals

© DLR 2004-2016