With the introduction of mobile mapping technologies, geomonitoring has become increasingly efficient and automated. The integration of Simultaneous Localization and Mapping (SLAM) and robotics has effectively addressed the challenges posed by many mapping or monitoring technologies, such as GNSS and unmanned aerial vehicles, which fail to work in underground environments. However, the complexity of underground environments, the high cost of research in this area, and the limited availability of experimental sites have hindered the progress of relevant research in the field of SLAM-based underground geomonitoring.
In response, we present SubSurfaceGeoRobo, a dataset specifically focused on underground environments with unique characteristics of subsurface settings, such as extremely narrow passages, high humidity, standing water, reflective surfaces, uneven illumination, dusty conditions, complex geometry, and texture less areas. This aims to provide researchers with a free platform to develop, test, and train their methods, ultimately promoting the advancement of SLAM, navigation, and SLAM-based geomonitoring in underground environments.
SubSurfaceGeoRobo was collected in September 2024 in the Freiberg silver mine in Germany using an unmanned ground vehicle equipped with a multi-sensor system, including radars, 3D LiDAR, depth and RGB cameras, IMU, and 2D laser scanners. Data from all sensors are stored as bag files, allowing researchers to replay the collected data and export it into the desired format according to their needs. To ensure the accuracy and usability of the dataset, as well as the effective fusion of sensors, all sensors have been jointly calibrated. The calibration methods and results are included as part of this dataset. Finally, a 3D point cloud ground truth with an accuracy of less than 2 mm, captured using a RIEGL scanner, is provided as a reference standard.
Multi-temporal landslide inventories are important information for the understanding of landslide dynamics and related predisposing and triggering factors, and thus a crucial prerequisite for probabilistic hazard and risk assessment. Despite the great importance of these inventories, they do not exist for many landslide prone regions in the world. In this context, the recently evolving global-scale availability of high temporal and spatial resolution optical satellite imagery (RapidEye, Sentinel-2A/B, planet) has opened up new opportunities for the creation of these multi-temporal inventories.To derive such multi-temporal landslide inventories, a semi-automated spatiotemporal landslide mapper was developed at the Remote Sensing Section of the GFZ Potsdam being capable of deriving post-failure landslide objects (polygons) from multi-sensor optical satellite time series data (Behling et al., 2016). The developed approach represents an extension of the original methodology (Behling et al., 2014, Behling and Roessner, 2020) and facilitates the integration of optical time series data acquired by different satellite systems. The goal of combining satellite data originating from variable sensor systems has been the establishment of longest possible time series for retrospective systematic assessment of multi-temporal landslide activity at highest possible temporal and spatial resolution. We applied the developed approach to a 2500 km² study area in Southern Kyrgyzstan using an optical satellite database acquired by the Landsat TM/ETM+, SPOT 1/5, IRS1-C LISSIII, ASTER, and RapidEye sensor systems covering a time period between 1986 and 2013. A multi-temporal landslide inventory from 2009-2013 derived from RapidEye satellite time series data is available as separate publications (Behling et al., 2014; Behling and Roessner, 2020).The resulting systematic multi-temporal landslide inventory being subject of this data publication is supplementary to the article of Behling et al. (2016), which describes the extended spatiotemporal landslide mapper in detail. This multi-sensor approach prioritizes most suitable images within the available multi-sensor satellite time series using parameters, such as spatial resolution, cloud coverage, similarity of sensor characteristics and seasonality related to vegetation characteristics with the goal of establishing a robust back-bone time series for initial detection of possible landslide objects. In a second step, this initial analysis gets more refined in order to achieve the best possible approximation of the date of landslide occurrence. For a more detailed description of the methodology of the extended spatiotemporal landslide mapper, please see Behling et al. (2016).In general, this landslide mapper detects landslide objects by analyzing temporal NDVI-based vegetation cover changes and relief-oriented parameters in a rule-based approach combining pixel- and object-based analysis. Typical landslide-related vegetation changes comprise abrupt disturbances of vegetation cover in the result of the actual failure as well as post-failure revegetation which usually happens at a slower pace compared to vegetation growth in the surrounding undisturbed areas, since the displaced landslide masses are susceptible to subsequent erosion and reactivation processes. The resulting landslide-specific temporal surface cover dynamics in form of temporal trajectories is used as input information to identify freshly occurred landslides and to separate them from other temporal variations in the surrounding vegetation cover (e.g., seasonal vegetation changes or changes due to agricultural activities) and from permanently non-vegetated areas (e.g., urban non-vegetated areas, water bodies, rock outcrops).The data are provided in vector format (polygons) in form of a standard shapefile contained in the zip-file 2020-002_Behling_et-al_2016_landslide_inventory_SouthernKyrgyzstan_1986_2013.zip and are described in more detail in the associated data description.
Im Rahmen eines von Prof. Buchroithner in Zusammenarbeit mit Doz. Haeusler (Univ. Wien) beim Oesterreichischen Fonds zur Foerderung der wissenschaftlichen Forschung (FWF) beantragten interdisziplinaeren Forschungsprojektes wurden mittels multisensoraler Fernerkundungsdaten und GIS geogene und anthropogene Oberflaechenveraenderungen erfasst (Environmental Monitoring). Das Projekt wurde im Juni 1996 offiziell beendet. Die Dissertation des Projektwissenschaftlers Mag. D. Leber wird Anfang 1997 abgeschlossen sein.
Die weltweite Beobachtung des Meereises mit aktiven und passiven Fernerkundungsverfahren hat wegen seiner grossen Bedeutung fuer das globale Klima der Erde grosses Interesse geweckt. Besonders Sensoren im Mikrowellenbereich eignen sich wegen eines starken Kontrasts zwischen Eis und Wasser in diesem Frequenzbereich zur Detektion des Meereises. Auch lassen sich Messungen in der Polarnacht problemlos durchfuehren, da diese Geraete die emittierte Strahlung der Erde empfangen. Auf den Gebieten der in-situ Messung von Meereis und empirischen Retrieval-Verfahren einerseits und auf der modellierenden Seite (Emissivitaet deterministischer Eisstrukturen und Strahlungstransport durch die Atmosphaere) andererseits liegen so viele Erkenntnisse vor, dass jetzt beide Teilgebiete durch ein geschlossenes Modell zugsammengefasst werden sollen. Der Einfluss verschiedener Parameter des Meereises wie Salzgehalt und Temperatur wird untersucht. Zur Validierung des Modells wurden zusaetzliche Daten, welche im Rahmen der Arktisexpedition ARK XII vom Mitarbeiter in diesem Projekt aufgenommen wurden, genutzt, und ein Algorithmus zur Bestimmung der Schneedicke aus SSMiI Daten entwickelt.