Search

The dataset is composed of Neo HySpex (VNIR/SWIR) hyperspectral imagery acquired during the GFZ/DIMAP Geoarchive airborne campaign on June 6th, 2015 covering the Omongwa salt pan located in the South-West Kalahari, Namibia. The dataset includes 9 merged cloud-free flight lines with 408 spectral bands ranging from VNIR to SWIR wavelength regions (0.4-2.5 μm). The dataset also includes Level 2A EnMAP-like imagery simulated using the end-to-end Simulation tool (EeteS). The overall goal of the campaign was to test the potential of advanced optical hyperspectral remote sensing, or imaging spectroscopy, for the analysis of surface processes in the Omongwa salt pan and for the quantification of surface sediments. Specifically, the mineralogical composition of exposed evaporites such as halite, gypsum and calcite were investigated at the airborne and spaceborne scale, associated with comprehensive field campaigns, ich which spectral reflectance and ground-truth chemical data of field samples have been collected. The data are highly novel and can be used as testbeds for the development and validation of retrieval algorithms based on air- and space-borne hyperspectral imagery for estimation of surface sediment properties in a highly saline and dynamic environment.

Recent developments in high‐resolution ecosystem mapping using unmanned aerial vehicles (UAV) open up promising perspectives for the monitoring of fine‐scale vegetation patterns and ecological functioning. In this study, we examine the potential of UAV imagery to track the structural composition and related phenological traits of the dwarf shrub Calluna vulgaris on a former military training area. On a European Natura 2000 heathland site, habitat management is shown to be evaluated on the basis of flowering dynamics and reproductive recovery as a proxy of functional changes that are generated after fire destruction. In particular, we utilize true color camera information and digital surface models to determine the spatiotemporal evolution of Calluna life‐cycle phases and flowering phenology 2 years after controlled burning. A stepwise methodological framework is presented that extracts Calluna pixels, spatially separates juvenile and mature life‐cycle phases, differentiates between generative and vegetative shoot extension and finally quantifies proportions of flowers, fruits and vegetative growth in two consecutive peak flowering periods. We show that Calluna life‐cycle phases can be spatially differentiated in pioneer, building and mature phases in UAV imagery. In the juvenile phase, regeneration from germination is low (

The dataset is composed of Neo HySpex (VNIR/SWIR) hyperspectral imagery acquired during airplane overflights on June 6th, 2015 covering the Omongwa Pan located in the South-West Kalahari, Namibia. The dataset includes three cloud-free flight lines with 408 spectral bands ranging from VNIR to SWIR wavelength regions (0.4-2.5 µm). The dataset also includes Level 2A EnMAP-like imagery simulated using the end-to-end Simulation tool (EeteS). The overall goal of the campaign was to acquire imagery over the Omongwa Pan and use the spectral reflectance for the analyses of surface sediments, specifically the mineralogical composition of exposed surface evaporites / salts on the airborne and spaceborne scale. The data are highly novel and can be used to test estimation of surface sediment properties in a highly saline and dynamic environment.

The combined effects of climate change and human land use activities are increasing the occurrence of extreme events, such as floods and landslides worldwide. Between mid-March and the beginning of April 2019, almost the whole Iran was affected by intense record rainfall leading to widespread flash flooding. Starting in the northern province of Golestan and moving to the South and the West of the country, heavy rains have poured on 28 out of the 31 provinces. Approximately, 1900 cities and villages across the country were damaged with half a million people displaced at the peak of the disaster. The intense rainfall not only caused more than 140 rivers to burst their banks, but also triggered thousands of landslides. The presented work makes use of high resolution - spatial and temporal - optical and radar satellite remote sensing with the goal of performing large-area and at the same time detailed spatiotemporal analysis of landslide occurrence related to the main hydro-meteorological triggering events. Besides detection of rainfall and flood triggered landslides, pre- and post-failure landslide evolution has been analyzed including the first observed onset of landslide activation in February 2019 resulting from already increased precipitation in late fall and winter of 2018/19. Large-area landslide detection has been performed in a semi-automated way using time series of optical Planet Scope and Sentinel-2A/B data. For assessing the kinematics of pre- and post-failure slope motions, radar remote sensing based image correlation and differential interferometric synthetic aperture radar (DInSAR) methods have been applied using Sentinel-1A/B data. The obtained satellite remote sensing based results have been evaluated by field surveys conducted in cooperation between the GFZ Potsdam and the Forest, Range and Watershed Management Organization of Iran (FRWM) being responsible for landslide hazard and risk assessment as well as the design and implementation of mitigation measures. In this context, emphasis has been put on the potential of multi-sensor satellite remote sensing for improving large-area process monitoring and understanding, especially in relation to extreme triggering events of so far unprecedented magnitude which are likely to occur more frequently in the future because of ongoing climate change.

Unmanned aerial vehicles (UAV) open up new perspectives for the repetitive spatial monitoring of vegetation stands and possibly even for fine scale analyses of individual plants. The potentials of recent UAV camera systems for integrating high spatial resolution (< 10 cm per image pixel) and area wide mapping are undergoing research. In particular, robust algorithms are needed for a spatially explicit characterization of individual plants, their structural composition and related functional traits. Such information can be used to examine ecological interdependencies that determine habitat establishment, and in general, to evaluate the habitat quality for nature conservation purposes [1]. In our study we developed a UAV-based methodological procedure to analyse open heathland areas that are protected in the European Natura 2000 network, and therefore, need to be managed to preserve a favourable conservation status [2]. We mapped the flowering phase of regenerated dwarf shrub heath (Calluna vulgaris) stands 2 years after burning and of unmanaged old stands using a standard RGB drone camera. From the UAV imaging point cloud a digital orthophoto and a digital surface model was generated with 2 cm pixel resolution. A method was developed that extracts every individual plant on the basis of RGB-colour value classification, texture filtering and local maxima estimation on a normalized digital surface model. For every plant the maximum plant height as well as the RGB colour distribution could be extracted and related to field surveys of plant traits. On that basis, a colour model was constructed that statistically predicts the phenological heath plant status with regard to fractional cover of flowers, fruits, vegetative shoots and senescent plant components. Our study reveals that the flowering phase of Calluna vulgaris is spatially much more divers than expected. Colour models from simple RGB imagery show that phases of completely vegetative, withered, fruit and full flowering plants spatially coexist in heterogeneous alternations of individual plants, in fact, independent of growth age after burning (height < 40cm). Only unmanaged old stands (height > 40 cm) exhibit significantly lower fractions of fruit and flowers while senescence is increased. However, the high phenological diversity after fire management shows that driving factors of shoot extension and flower productions, thus, are triggered by additional mechanisms that are not affected by habitat management practice. UAV-based imagery can help to observe such effects for a more detailed evaluation of management effects. REFERENCES: 1. Neumann C, Weiss G, Schmidtlein S, Itzerott S, Lausch A, Doktor D, Brell M., 2015. Gradient-based assessment of habitat quality for spectral ecosystem monitoring. Remote Sensing. 7(3), 2871-98. 2. Habitat Directive, 1992. Council Directive 92/43/EEC of 21 May 1992 on the conservation of natural habitats and of wild fauna and flora. Official Journal of the European Union. 206, 7-50.

On November 12, 2017, a large earthquake (Mw = 7.3) hit Kermanshah province of Iran near the border of Iran and Iraq. At least 600 people were killed, 9000 injured and thousands left homeless. In this study we integrated observations from radar and optical remote sensing, seismology and field mapping to investigate source parameters, coseismically triggered slope failures and secondary faulting related to the Mw 7.3, Sarpol-e Zahab earthquake in Iran. Coseismic surface deformation was constrained by interferometric synthetic aperture radar (InSAR) analysis from SAR images obtained by L-band ALOS-2 and C-band Sentinel-1 satellites. ALOS-2 wide-swath (WS) interferograms were derived from ascending and descending orbits covering 09.08.2016-14.11.2017 and 04.10.2017-15.11.2017 time periods, respectively. Sentinel-1 TOPS coseismic interferograms were derived from ascending and descending orbits covering 11.11.2017-17.11.2017 and 12.11.2017-18.11.2017, respectively. Source parameters and slip models of the earthquake were then obtained by the Bayesian inversion of interferometric results using elastic dislocation modeling, considering the seismic parameters as a priori information. Preliminary results of geodetic source modeling suggests that the 12 November 2017 Sarpol-e Zahab earthquake was generated by blind ENE oblique thrust faulting with an average slip of approximately 4 m at a depth between 17 and 22 km. The 2017 Sarpol-e Zahab event also triggered a lot of coseismic slope failures including landslides and rock falls, and secondary faulting. Their potential locations were initially assessed by identifying local phase changes and areas affected by loss of coherence in the interferograms. Moreover, large-scale analysis of horizontal surface displacements was performed by pixel-offset tracking using Sentinel-1 SAR images. The results were evaluated by expert interpretation based on optical Sentinel-2 data in combination with geological information and field investigations. Largest horizontal deformation was observed for a co-seismically triggered landslide which was further analyzed using optical PlanetScope data acquired one day apart right before and after the earthquake (12th and 13th of November). Pixel-based change detection allowed exact spatial delineation of the displaced block covering an area of about 4 km2. Cross-correlation analysis resulted in a spatially detailed derivation of horizontal movement vectors indicating displacements in the range from about 5 to 30 m. The results of this study have shown that a combination of various remote sensing methods with modelling and field investigations is required for gaining an improved understanding of surface processes caused by the Mw 7.3, Sarpol-e Zahab earthquake. Complementary analysis of radar and optical remote sensing has led to a more comprehensive assessment of various types of co-seismically induced surface changes occurring at different spatial scales. Such information can contribute to an improved spatial characterization of co-seismically activated geological structures. This is especially important in the context of seismic hazard assessment for scarcely investigated areas like the one affected by the Sarpol-e Zahab earthquake in the border region between Iran and Iraq.

Large areas in southern Kyrgyzstan are subjected to high and ongoing landslide activity; however, an objective and systematic assessment of landslide susceptibility at a regional level has not yet been conducted. In this paper, we investigate the contribution that remote sensing can provide to facilitate a quantitative landslide hazard assessment at a regional scale under the condition of data scarcity. We performed a landslide susceptibility and hazard assessment based on a multi-temporal landslide inventory that was derived from a 30-year time series of satellite remote sensing data using an automated identification approach. To evaluate the effect of the resulting inventory on the landslide susceptibility assessment, we calculated an alternative susceptibility model using a historical inventory that was derived by an expert through combining visual interpretation of remote sensing data with already existing knowledge on landslide activity in this region. For both susceptibility models, the same predisposing factors were used: geology, stream power index, absolute height, aspect and slope. A comparison of the two models revealed that using the multi-temporal landslide inventory covering the 30-year period results in model coefficients and susceptibility values that more strongly reflect the properties of the most recent landslide activity. Overall, both susceptibility maps present the highest susceptibility values for similar regions and are characterized by acceptable to high predictive performances. We conclude that the results of the automated landslide detection provide a suitable landslide inventory for a reliable large-area landslide susceptibility assessment. We also used the temporal information of the automatically detected multi-temporal landslide inventory to assess the temporal component of landslide hazard in the form of exceedance probability. The results show the great potential of satellite remote sensing for deriving detailed and systematic spatio-temporal information on landslide occurrences, which can significantly improve landslide susceptibility and hazard assessment at a regional scale, particularly in data-scarce regions such as Kyrgyzstan.

The consequences of climate change represent a major threat to sustainable development and growth in Southern Africa. Understanding the impact on the geo- and biosphere is therefore of great importance in this particular region. In this context the Kalahari salt pans (also known as playas or sabkhas) and their peripheral saline and alkaline habitats are an ecosystem of major interest. They are very sensitive to environmental conditions, and as thus hydrological, mineralogical and ecological responses to climatic variations can be analysed. Up to now the soil composition of salt pans in this area have been only assessed mono-temporally and on a coarse regional scale. Furthermore, the dynamic of the salt pans, especially the formation of evaporites, is still uncertain and poorly understood. High spectral resolution remote sensing can estimate evaporite content and mineralogy of soils based on the analyses of the surface reflectance properties within the Visible-Near InfraRed (VNIR 400-1000 nm) and Short-Wave InfraRed (SWIR 1000-2500 nm) regions. In these wavelength regions major chemical components of the soil interact with the electromagnetic radiation and produce characteristic absorption features that can be used to derive the properties of interest. Although such techniques are well established for the laboratory and field scale, the potential of current (Hyperion) and upcoming spaceborne sensors such as EnMAP for quantitative mineralogical and salt spectral mapping is still to be demonstrated. Combined with hyperspectral methods, multitemporal remote sensing techniques allow us to derive the recent dynamic of these salt pans and link the mineralogical analysis of the pan surface to major physical processes in these dryland environments. In this study we focus on the analyses of the Namibian Omongwa salt pans based on satellite hyperspectral imagery and multispectral time-series data. First, a change detection analysis is applied using the Iterative-reweighted Multivariate Alteration Detection (iMAD) method to identify and investigate surface changes based on Landsat archive imagery covering the period 1984-2015. For the complete Landsat time series, a total of 130 bi-temporal change maps have been derived and are compiled to produce a change magnitude map, which indicates different regions of activity and stability of the pan surface. Then radiometric geometric and atmospheric correction of Hyperion data were performed using the pre-processing chain of Rogass et al. (2014). Based on field spectroscopy and XRD analysis of soil samples, one of the main evaporite minerals of the pan soil could be identified as gypsum. Different approaches based on spectral features were tested and validated against reference samples to find the most suitable approach for estimating the gypsum content from the Hyperion data. Major challenges were the influence of water vapour absorption bands close to major gypsum absorption feature as well as low signal-to-noise ratio of Hyperion in the SWIR wavelength. A best method was determined that allow to determine Gypsum content with R² of 0.7 and relative RMSE of 0.14. Results reveal a variable spatial distribution of different mineralogy and in particular gypsum content within the pan, which seems to be associated with varying dynamic of the pan surface.

The main principle in landslide hazard and risk assessment is that conditions of past and present landslide activity indicate future landslide occurrence. Hence, the probabilistic assessment of landslide hazard and risk requires a profound knowledge about spatiotemporal landslide activity over longer time spans and large areas. However, for most parts of the world such information is largely missing, because the identification of landslide activity still mainly relies on time-consuming and resource-intensive conventional methods (i.e. visual interpretation of optical data supported by comprehensive field surveys) and because all efforts to automate this landslide mapping procedure are hitherto limited to small areas and/or very short time periods or even single events. This thesis presents the development of an automated approach for efficient multi-temporal identification of landslides based on optical satellite-based remote sensing time series data. The developed approach allows for retrospective analysis of long-term landslide occurrence and for monitoring recent landslide activity for large areas. For this purpose, a comprehensive optical remote sensing database has been created. To achieve best temporal resolution, data of multiple optical sensors are used simultaneously. In total, the database consists of 729 datasets acquired by Landsat-(E)TM, SPOT 1 & 5, IRS-1C (LISS3), ASTER, and RapidEye between 1986 and 2013 for a landslide-affected area of 12000 km² in Southern Kyrgyzstan, Central Asia. The developed approach comprises automated multi-sensor pre-processing as well as knowledge-based and uncertainty-related multi-temporal change detection methods to enable efficient and robust spatiotemporal identification of landslides in a highly heterogeneous multi-sensor time series database. The change detection builds on the analysis of temporal NDVI-trajectories, representing footprints of vegetation changes over time. Landslide-specific trajectories are characterized by abrupt vegetation cover destruction and longer-term revegetation rates resulting from landslide-related disturbance and dislocation of the fertile soil cover. In combination with DEM-derivatives the developed approach enables automated identification of landslides of different sizes, shapes and in different stages of development under varying natural conditions. The approach is applied to two scenarios. Firstly, the recent landslide activity (2009 – 2013) is identified by a RapidEye-based application of the approach in a 7500 km² area. Secondly, a long-term analysis (1986 - 2013) is performed on the basis of the multi-sensor database in a highly landslide-affected region of 2500 km². In both cases, the number of automatically mapped landslides exceeds the existing landslide records of the Kyrgyz authorities by more than a factor of ten. In total, almost 2000 landslides are mapped, whereas the size of the landslides ranges from 50 m² to 2.8 km². The identified landslide occurrence shows clear spatial patterns with highest activity along the foothills of the Tien-Shan mountain ranges. Temporally, the long-term analysis reveals a peak of landslide activity for the years 2003 and 2004. In these years, the annual landslide rate was more than five times higher than the identified long-term average rate of 57 landslides per annum. These spatiotemporal activity patterns are evaluated against the morphological setting (predisposing factor) and the temporal variations of the precipitation (triggering factor), exemplarily showing the suitability of the achieved results to determine the individual or combined influence of specific landslide-causing factors in an analyzed region. For the derivation of the recent landslide activity, the approach takes advantage of the high spatial (5 m) and temporal resolution (acquisition intervals of up to several days/weeks) of the RapidEye data, allowing identification of small landslides that are often pre-cursors of subsequent more hazardous landslides. Thus, the approach can also provide valuable information for early warning applications in the context of a regular landslide monitoring system. Overall, the presented approach identifies spatiotemporal landslide activity patterns to improve the regional landslide process understanding. This builds an important step to realize probabilistic landslide hazard and risk assessments in order to contribute to the mitigation of the landslide consequences for the local population of remote mountainous regions such as Southern Kyrgyzstan. Moreover, the multi-sensor applicability and uncertainty-related landslide change detection of the approach provides a great potential to be applied to other global landslide hotspots (e.g. South-East Asia and South America) as well as to new optical sensors (e.g. Sentinel-2), opening up new opportunities to establish a widely applicable large area landslide monitoring system. Zur Gefahreneinschätzung von Hangrutschungen greift man auf das grundlegende Prinzip zurück, dass zukünftige Hangrutschungen den gleichen Bedingungen unterliegen wie vergangene und gegenwärtige. Eine Voraussetzung für solche probabilistischen Gefährdungseinschätzungen sind daher fundierte Kenntnisse der raumzeitlichen Hangrutschungsaktivität über möglichst lange Zeiträume und große Gebiete. Allerdings fehlt solches Wissen in den meisten Regionen der Erde, da die Erfassung der Hangrutschungsaktivität nach wie vor meistens durch zeit- und arbeitsaufwendige Methoden erfolgt (d. h. visuelle Interpretation von optischen Fernerkundungsdaten und umfassenden Geländearbeiten) und weil Ansätze diese Hangrutschungsidentifizierung zu automatisieren sich bisher auf kleine Untersuchungsgebiete und/oder auf kurze Zeiträume beschränken. In dieser Dissertation wird eine automatische Methode zur effizienten Hangrutschungsidentifizierung auf Basis von Zeitreihen optischer Satellitenfernerkundungsdaten entwickelt. Diese Methode erlaubt großflächig eine rückwirkende Analyse langzeitlichen Hangrutschungsvorkommens und ein Monitoring rezenter Hangrutschungsaktivität. Zu diesem Zweck wurde eine umfangreiche Datenbank optischer Fernerkundungsdaten erstellt, die, um eine möglichst hohe zeitliche Auflösung zu erreichen, Daten mehrerer optischer Sensoren einschließt. Insgesamt enthält die Datenbank 729 Datensätze sieben verschiedener optischer Satellitensysteme (Landsat-(E)TM, SPOT 1 & 5, IRS-1C (LISS3), ASTER und RapidEye). Die Aufnahmen erstrecken sich über einen Zeitraum von 1986 bis 2013 und über ein 12000 km² großes von Hangrutschungen betroffenes Gebiet im südlichen Kirgisistan (Zentralasien). Die entwickelte Methode umfasst automatische multisensorale Vorverarbeitungsschritte sowie wissensbasierte und Unsicherheiten einbeziehende Algorithmen der multitemporalen Veränderungsdetektion um eine effiziente und robuste raumzeitliche Identifizierung von Hangrutschungen auf Basis einer stark heterogenen multisensoralen Datenzeitreihe zu ermöglichen. Die Algorithmen zur Veränderungsdetektion basieren auf der Analyse von zeitlichen NDVI-Trajektorien, welche ein Abbild zeitlicher Vegetationsveränderungen darstellen. Hangrutschungstypische Muster sind gekennzeichnet durch abrupte Vegetationszerstörung und einen langsamen Wieder-bewuchs, da durch die Hangrutschungen fruchtbarer Oberboden gestört oder gar verlagert worden ist. In Verbindung mit reliefbeschreibenden Parametern erlaubt der entwickelte Ansatz somit die automatische Identifizierung von Hangrutschungen unterschiedlicher Größe, Form und Entwicklungsstadien unter verschiedenen natürlichen Bedingungen. Die Methode wird auf zwei Szenarien angewendet. Zunächst wird die rezente Hangrutschungsdynamik auf Basis der RapidEye-Daten für ein Gebiet von 7500 km² untersucht. Anschließend erfolgt eine langzeitliche Analyse (1986-2013) eines 2500 km² großen stark von Hangrutschungen betroffenen Gebietes mittels der multisensoralen Datenbank. In beiden Fällen übersteigt die Anzahl der automatisch identifizierten Hangrutschungen jene offizieller kirgisischer Berichte um mehr als das Zehnfache. Insgesamt sind annähernd 2000 Hangrutschungen mit einer Größe von 50 m² bis 2.8 km² identifiziert worden. Das erfasste Hangrutschungsvorkommen zeigt deutliche räumliche Muster mit höchster Aktivität entlang der Hänge des Tien-Shan Gebirgsvorlandes. Zeitlich betrachtet hat die langzeitliche Analyse ein Aktivitätsmaximum in den Jahren 2003 und 2004 ergeben. Diese Jahre waren durch eine mehr als fünffach erhöhte Hangrutschungsaktivität gegenüber dem langjährigen Mittel von 57 Hangrutschungen gekennzeichnet. Diese raumzeitlichen Hangrutschungsaktivitätsmuster werden in der Dissertation der morphologischen Beschaffenheit (ursächlicher Faktor) und den zeitlichen Variationen des Niederschlages (auslösender Faktor) gegenübergestellt. So wird exemplarisch die Eignung der erzielten Identifizierungsergebnisse gezeigt, einzelne oder zusammenhängende Einflüsse von Hangrutschungen verursachenden Faktoren für ein Untersuchungsgebiet quantitativ zu bewerten. Für die Ableitung rezenter Hangrutschungsaktivität wird auf die RapidEye-Daten zurückgegriffen, die durch ihre hohe räumliche (5 m) und zeitliche Auflösung (Datenwiederholungsrate von bis zu wenigen Tagen und Wochen) eine Identifizierung von kleinen Rutschungen als Vorboten von größeren gefährlicheren Hangrutschungen zu ermöglichen. Dadurch kann die Methode innerhalb eines regelmäßigen Monitorings auch für Frühwarnsysteme nützliche Informationen liefern. Die entwickelte Methode erlaubt die Identifizierung von raumzeitlichen Hangrutschungsaktivitätsmustern, welche zur Verbesserung die des regionalen Hangrutschungsprozessverständnis beitragen. Damit ist ein wichtiger Beitrag geschaffen zur Realisierung von Hangrutschungsgefährdungseinschätzungen, und damit zur potentiellen Reduzierung des Hangrutschungsrisikos für die lokale Bevölkerung von abgelegenen Gebirgsregionen wie im Falle Südkirgistans. Des Weiteren hat die Methode durch ihre sensorübergreifenden und Unsicherheiten einbeziehenden Analysealgorithmen ein großes Potential auf weitere hangrutschungsgefährdete Gebiete (wie bspw. Südostasien oder Südamerika) sowie auf neue optische Sensoren (z. B. Sentinel-2) angewandt zu werden, was Möglichkeiten für ein vielseitig und großflächig anwendbares Monitoringsystem eröffnet.