Search

Canny edge detection is a multistage processing method that can be employed to detect edges in an image while suppressing noise. This edge detection operator was proposed by Canny in 1986 and is nowadays used for the real-time analysis of images in several application areas, including facial expression recognition, traffic management, tracking of unmanned aerial vehicles, public safety, and more.Ground Penetrating Radar (GPR) systems are most often used to detect buried targets. Small circular-section targets (such as tubes and cables, tree roots, landmines, concrete rebar, etc.) translate into radargrams as hyperbolic-shaped signatures. Considering the large quantity of GPR data that can be acquired during a field campaign, the manual detection and localization of hyperbolas in radargrams can be a lengthy task, which becomes impracticable in large-scale surveys. Moreover, the manual detection may be unfeasible in real-time applications. For these reasons, various research teams have been recently working at the development of quick and automated algorithms for the detection of hyperbolas in radargrams.In this work, we investigate the applicability of the Canny edge detection operator in the GPR area. In particular, we use Canny edge detection as the first step of a processing algorithm for the real-time detection of hyperbolic reflections in GPR images. Particular efforts were done to reduce the time needed to detect edges in a radargram. Several criterions were proposed and analysed, to eliminate unwanted edge pixels and extract those points that actually belong to the sought hyperbolic reflections. The optimized edge image is passed to the subsequent steps of the algorithm, which estimate the position of the relevant targets. The proposed processing algorithm was implemented in Matlab and it was tested on synthetic and experimental radargrams. The open-source finite-difference time-domain simulator gprMax was used to generate the synthetic radargrams, whereas the experimental radargrams were obtained by field survey. [ABSTRACT FROM AUTHOR]

Mass is an important parameter in both the physical processes which govern CME motion and is directly linked to CME observations, as most remote sensing measurements of a CME are white-light images where intensity is correlated with density. As such, mass is both a crucial parameter to measure and one of the few that can be studied directly. Most attempts to determine masses from remote sensing data have focused on coronagraph data, where the CME is bright and still a roughly self-similarly expanding flux rope that has undergone few dynamic changes since the initial eruption. However, to understand the flux rope evolution once the CME begins to interact with the upstream solar wind and determine how it may change in the heliosphere, it is necessary to probe the data and constrain the mass at higher heights. Attempts to extend the coronagraph methods and study the mass at higher heliographic distances have been difficult given the low signal to noise ratio of CME observations in the STEREO SECCHI HI-1 FOV. This study uses improved image processing techniques to better isolate the ejecta mass from the background. Obtaining this signal without resorting to running-difference techniques reduces the amount of useful signal lost due to image processing while still reducing the unwanted contribution from stars and other background features. By comparing the obtained masses with both COR2 measurements as well as in-situ densities we can better understand the challenges in determining mass farther away from the Sun and quantify the systematic error such an approach introduces. Directly studying more local evolution of the mass inside the CME, we can directly address the degree to which the CME evolves non-uniformly and how momentum is transferred between the CME and the upstream solar wind. [ABSTRACT FROM AUTHOR]

This presentation examined a machine learning with image processing techniques for classifying land cover conditions in UAV (Unmanned Aerial Vehicle) images of a riverine landscape. The UAV images were taken in a river course of Kurobe River, Japan in both November 2017 and 2018, where it had a braided channel with well-vegetated gravel beds. The UAV image analyzed was composed of a RGB, NDVI (Normalized Difference Vegetation Index), and a DSM (Digital Surface Model) of the river geomorphology made from an SfM (Structure from Motion) image processing of the UAV images. In addition, as a preprocessing of a machine learning, the DSM was decomposed into the low/high wavenumber components by a wavelet transform, and further its edges were extracted by a Laplacian filter for effectively utilizing the height difference information in DSM. These characteristic values from UAV images were all used in combination to examine their effectiveness in machine learning for the riverine land cover classification. The machine learning techniques examined in this presentation was RF (Random Forest), and the true values of trees, grasses, bare gravel/sand bed, and water surface in part of the river course were given as a training dataset. The result of the machine learning with the image preprocessing treatment showed that the F-measure, which is consist of precision and recall rates, had high enough above 0.91 in the dataset including all the characteristic values from RGB, DMS, and NDVI into the machine learning algorithm. As for the comparison of effectiveness in each characteristic value in riverine land cover classification, the combination of DSM high wavenumber and DSM edge values made the F-measure higher in trees/grasses classification. Furthermore, the applicability of the machine learning with the image processing techniques examined here were discussed in detail in terms of the calculation time and the quality of input images for accurate classification. [ABSTRACT FROM AUTHOR]

In Brazil the use of drones has become popular in commercial and scientific sectors. In geosciences, however, this tool is yet little applied, being restricted to the initiative of some research institutes. One example is the brazilian Instituto de Pesquisas Tecnológicas of the State of Sao Paulo (IPT) that has been using small multi-rotor UAVs for mapping and monitoring of geological-geomorphological surface processes and for natural disaster and risk management.Drones are an important tool for monitoring the evolution of geological-geomorphological process, allowing quick response, damage assessment and stability analysis, in places with limited access or that pose danger to the authorities. A working routine was created by the Institute, based on versatile tools, with low cost and easy-to-use. Popular commercial low-cost UAVs were applied, as well as free applications and softwares for flight planning and image processing, using mainly cloud services, allowing agility in data processing and in the obtention of its main products (orthomosaics and tridimensional models).We use the DJI UAVs Phantom 3 and DJI Inspire 1, with conventional RGB cameras. For flight planning and execution, we use DJI Go application. As a protocol, we make a preliminary reconnaissance flight on the field above the area of interest to recognize its main characteristics, especially regarding topographic differences and vegetation heights, anticipating possible collisions and distortions on the products obtained. Images are then processed in a cloud web service (Precision Mapper), through which the orthomosaics and digital elevation models are obtained.This routine allows the analysis and monitoring of geological-geomorphological processes, such as landslides, debris-flows and gullies evolution, allowing estimates of the volume of mobilized material, number of endangered houses and risk analysis associated with anthropic activities, such as mining, pipelines and landfills.Up to now the main results obtained through this routine consist in the monitoring of two major landslides and the evolution of a gully, located in the state of São Paulo. Regarding the landslide cases, based on the orthomosaic and the DEMs, we obtained the scar area and perimeter, extension of damages, volume of the mobilized material and its possible evolution. As for the gully monitoring, we obtained the volume of mobilized material and identified the morphologies associated with the process evolution for posterior analysis and monitoring.The use of drones in emergency and geomorphological disaster management has so far been very satisfactory, allowing the support of field teams, agility on data acquisition, damage analysis, evaluation of process evolution and the monitoring of potential instability conditions, allowing more fast and assertive decision making.However, there are limitations and challenges that need to be overcome, such as the presence of dense vegetation, of special importance in the humid tropics, pointing out the need of replacing conventional RGB cameras for sensors that penetrate the vegetation, like LiDAR. Also, the limited availability of free data and the cost of sensors in Brazil are other important limitations to the diffusion of this technology, even though its adoption is inevitable considering the various advantages that it presents. [ABSTRACT FROM AUTHOR]

Geodetic Astronomy provides high precision deflection of the vertical (DoV) data usingDigital Zenith Camera Systems (DZCS) with an accuracy of 0.1 − 0.2 arc seconds, whichcontribute to the accurate geoid modelling studies. The concept of processing DZCSobservations depends on the data coming from various sensors such as CCD camera, GPS,and tiltmeters. The star images are captured using CCD camera near zenith direction,and the image solutions (astronomical coordinates) are combined with GNSS datato calculate DoV components. The DoV observations can be used to derive theslope of the geoid, and to produce geoid undulations using astronomical levellingtechnique. The astro-geodetic studies in Turkey using DZCS has started in 2012 with the ACSYSV.1. Later on, beginning from 2015, the modernized version of the system has started its firstobservations in Istanbul using more precise sensors. Observed DoV components have beenused to determine geoid height differences which are compared with the ones acquired usingGNSS/Levelling data in Istanbul. The accuracy of the DZCS is not only dependent on the precision of the sensors but alsoto the image processing methods used for astrometric reduction. Geodetic astronomy thatuses CCD images follows the same procedure of astrometric reduction process; however,opposing the conventional astronomical observations, the DZCS does not track the celestialobjects. Leaning on this background this study investigates a new method for increasing thenumber of stars in the images by extending the exposure time up to 60 seconds, which yieldstrailed images of the reference stars. The test data was collected as star imagesin various exposure time in order to model the best solution to acquire accurateastronomical coordinates. With this method of image processing it is possible toimprove the accuracy of the astronomical coordinates using fitting functions of the startrails. [ABSTRACT FROM AUTHOR]

High-alpine regions react sensitively to climate change. The most obvious consequence of warming temperatures is glacier retreat, a direct effect of negative mass balances over a number of years. Continuous monitoring of glaciers is important for the quantification of indicators such as changing snow cover patterns. An easy-to-handle and inexpensive way to do so is the analysis of terrestrial photos taken by automatic cameras. Photos from two cameras, operated by the Climate Research Department of the Central Institution for Meteorology and Geodynamics (ZAMG) in Vienna, Austria are installed at the high-alpine observatory on Hoher Sonnblick in the Austrian Alps. They are used to continuously monitor two glaciers (Kleinfleisskees and Goldbergkees) at high spatial and temporal resolution. Photos for the period 2014–2018 are available.The aim of the present work is to investigate on the spatio-temporal evolution of the snow cover patterns on these two glaciers, thereby providing and evaluating a prototype of an automated image processing framework facilitating the work flow. We explore four research questions: (i) How did the patterns of snow cover evolution on the analyzed glaciers change in the period 2014–2018? (ii) Which level of automation can be achieved in processing terrestrial photos for the detection of glacier changes, in particular snow-free areas on the glacier? (iii) What are the limitations and the key challenges in doing so? (iv) How large is the loss of accuracy with increasing degree of automation?The image processing framework developed operates in two steps: (i) The existing georectification tool PRACTISE is applied to automatically georectify all the terrestrial photographs. (ii) The newly developed routine GLACIERIZER, combining ArcGIS, Python, and the R software, is employed for image classification. A supervised classification method is applied, building on training areas mapped manually for a number of user-defined points in time. The entire framework is flexible enough to be applicable on other cameras, too. The results allow for quantitative statements on, for example, the evolution of snow-free areas on the glacier during each ablation period covered by the automated photographs.The derived snow cover patterns on the investigated glaciers show highly dynamic intra- and inter-annual variations, which can be clearly linked to the recorded current and antecedent weather patterns. Reduced melting in 2014 was reflected in increased 2015 firn snow cover. Even though we clearly demonstrate that the automated image processing framework leads to scientifically relevant results, we would like to emphasize that the manual definition of training areas remains imperative. We further identify a clear trade-off between the accuracy of the results and time efficiency: the accuracy – using manually mapped patterns as reference – increases with the number of points in time at which training areas are defined. This pattern is seen as an invitation for further research towards improving time efficiency while maintaining a high accuracy of the results. [ABSTRACT FROM AUTHOR]

In many landslide studies, the possibility to predict future behaviour is still a major concern. The most popular methods to achieve this prediction rely on the analysis of monitoring data of displacement for the detection of accelerating areas, as they can directly related to precursory movements that typically precede collapse. Methods deducing reliable failure predictions have been largely applied at local scale, where in situ monitoring systems can be installed. So far, the lack of systematic information on ground displacement acquired at regional scale was a serious limit hampering the application of failure prediction methods at wide scale. Such limitations can be partially solved through the exploitation of spaceborne platforms. Despite this, the low data sampling frequency of most of the satellite systems hampered the possibility to retrieve the necessary details of tertiary creep characterized by accelerating deformation.The launch of Sentinel-1 mission opened a new opportunity for InSAR monitoring applications thanks to the increased acquisition frequency, the regularity of acquisitions and the policy on data access. The potential of the Sentinel-1 for the detection of pre-failure accelerating creep has been recently proved through some retrospective InSAR analysis. For instance, the post-event application of the Fukuzono method to the Sentinel-1 deformation time series of the catastrophic Maoxian landslide (China), occurred on 24 June 2017, pointed out that an accurate estimation of the time of collapse was already possible at the begin of June.Despite the operational readiness of the Sentinel-1 constellation, most of the applications so far have been aimed at assessing the use of this data source to spot unstable areas or to capture pre-failure signs after major collapse, rather than providing new streamlines of information for monitoring solutions. We provide an example of the potential of multi-temporal InSAR analyses applied to Sentinel-1 data for continuous monitoring of ground deformation induced by hydrogeological processes. We exploit Sentinel-1 images to implement an operational service based on advanced interferometric products suitable for risk mitigation at the regional scale. This service relies on the systematic processing of Sentinel-1 images to create continuously updated ground deformation data. In fact, as soon as a new acquisition is available, the new image is used to update all displacement time series of all measurement points identified in the area of interest. Time series are then automatically analysed through a post-processing procedure, highlighting any anomalous trends and/or acceleration affecting the area of interest and providing possible alerts.We present and discuss this monitoring system through the case study of the Tuscany and Valle d'Aosta Region (Italy), specifically selected due to their peculiar geological settings, highly prone to slope instability phenomena. These results show that satellite radar data, systematically acquired over large areas with short revisiting time, could be used not only for mapping unstable areas, but also for actual landslide warning, at least for some typologies of sliding phenomena. This means that the transition from historical analysis of ground deformation to a continuous monitoring with prediction capabilities at regional scale using satellite radar data is now possible. [ABSTRACT FROM AUTHOR]

On August 12, 2018 the NASA Parker Solar Probe (PSP) Mission was launched into orbit about the Sun. For the first time ever, a man-made object will explore, in situ, the hot solar corona at heliocentric distances well below Mercury's orbit. The PSP S/C carries several in-situ instruments and one remote sensing instrument suite: the Wide-field Imager for Solar Probe (WISPR). WISPR is a heliospheric imager placed on the ram side of the S/C, which consists of two telescopes akin to the SECCHI heliospheric imagers onboard the STEREO mission. Their field of view ranges from 13.5 to 53.5 deg elongation for the inner telescope, and from 50 to 108 deg for the outer one. The science data acquisition is carried out while the S/C is below 0.25 AU (i.e., during the solar encounters around perihelion). The orbital path of PSP throughout this never-visited region introduces a myriad of observational challenges. For instance, the high eccentricity and small perihelion of the orbit (which translates into a very high speed of the S/C during the data science acquisition), makes the FOV of the WISPR telescopes to comprise different solar distances in short time periods and hence a continuous change of the background scenery. This precludes the use of conventional image processing techniques employed on imagers at about 1 AU aimed at revealing the coronal signal of interest for Space Weather purposes (e.g., the use of a straight running- or base-difference approach to remove the background brightness and hence reveal the K-coronal dynamic events). Moreover, WISPR observes both the K-coronal pseudo-static features and transients while moving across them, its interpretation becoming tricky without proper modeling. In this presentation, we present these and other challenges as we learned from them during the first solar encounter lived so far by the PSP mission. We gratefully acknowledge support from the NASA Parker Solar Probe project office. [ABSTRACT FROM AUTHOR]

Scanning Electron Microscopy is one of the most prevalent methods used to image and quantify the pore size distribution of shale rock, critical in understanding unconventional petroleum production. Generally, digital greyscale SEM images of shale are currently processed either by a manual drawing method, manual threshold method, automatic threshold method, edge detection or watershed methods, all of which have some limitations that will impact the quality of obtained image pore extraction results. A new, Edge-Threshold Automatic Processing (ETAP) method is reported here to enable robust extraction and quantification of pore data in shale images. Image pre-treatment makes the greyscale of regions brighter than that of kerogen set to the peak value of kerogen greyscale. The pore image is subsequently obtained using an edge detection method. Then, a discriminant function has been designed to determine the best threshold of the greyscale image to obtain the pore image. Finally, combination of both processed pore images gives the final pore image. Our new method overcomes the impact of kerogen, mineral, roughness and artificial debris caused by pre-treatment of samples, which potentially introduce errors in the processing of pore image extraction using alternative methods. We compare our new method to a systematic manual drawing method. The processing results through ETAP provide reliable results. [ABSTRACT FROM AUTHOR]

After more than three decades of research, the role of polar stratospheric clouds (PSCs) in stratospheric ozone depletion is well established. However, important questions remain unanswered that have limited our understanding of PSC processes and how to accurately represent them in global models, calling into question our prognostic capabilities for future ozone loss in a changing climate. A more complete picture of PSC processes on vortex-wide scales has emerged from the CALIOP (Cloud-Aerosol Lidar with Orthogonal Polarization) instrument on the CALIPSO satellite that has been observing PSCs at latitudes up to 82 degrees in both hemispheres since June 2006. A new CALIOP Version 2.0 (v2) PSC algorithm has been developed that addresses known deficiencies in previous algorithms and includes additional refinements to increase the robustness of the inferred PSC composition. In this paper, we present a state-of-the-art PSC reference data record and comprehensive climatology constructed by applying the v2 algorithm to the more than 13-year CALIOP spaceborne lidar dataset. In addition to 4-D (latitude, longitude, altitude, and time) information on the occurrence, composition, and variability of PSCs in both hemispheres, it also includes the estimated particulate SAD and VD to facilitate comparisons with in situ data and PSC measurements by other remote sensors, as well as with theoretical models relating PSCs to heterogeneous chemical processing and ozone loss. Finally, we compare the post-Pinatubo CALIOP PSC data record with the 1979-1989 SAM II (Stratospheric Aerosol Measurement II) solar occultation PSC record to investigate possible long-term variability in PSC occurrence. [ABSTRACT FROM AUTHOR]

The main purpose of a geophysical survey is to solve an engineering or exploration problem by obtaining the clearest possible image of the subsurface. For this purpose, data measurement methodologies and capabilities of measuring equipments are continuously improved. On the data processing side, various inversion schemes, usage of different smoothing and/or constraining functionals, plenty of alternative filters, migration operators and many more principal or auxiliary procedures are suggested to achieve the clearest possible geophysical image. Particularly in potential field methods, various filtering processes are applied to the data to highlight some elements in the geophysical image. In any case, the final geophysical model or image is presented by a color coded graphic. User-dependent settings used to create the graphics, such as the selection of color map, amplitude limits, contour interval etc. will affect the detail level of the final image. Choices related to the mentioned parameters may result in the disappearance of details in the model/section/map or the appearance of artificial anomalies.In this work, a group of image processing methods was applied to ground penetrating radar (GPR) data for the aim of obtaining an enhanced view of the subsurface. The processes are classified into main categories such as de-noising, sharpening, lighting and shading, edge and linearity detection. These processes were realized by the application of convolution filters, transforms (e.g. Hough transform) or gradient tools. The codes that perform the mentioned processes are written in MATLAB language and supported by a user-friendly interface. Thus, the interpreter can choose the best image of the subsurface among the images rendered with all possible processing options. In addition, the data can also be processed within a predetermined workflow. All of the image processing methods mentioned here have been directly applied to numerical data sets instead of a soft copy of an image. The overall methodology aims to minimize losses and errors based on the user choices about the visualization options. The interpreter will be able to emphasize various structures on the map or the sections by interactively changing the filter or transform parameters. The GPR data used for the demonstration of the study were measured during the extensive archaeo-geophysical surveys carried out in Side and Patara, the two most important cities of the ancient Lycia and Pamphylia regions. The GPR surveys were supported with electrical resistivity tomography (ERT) measurements at anomalous regions. The GPR data were processed with usual steps such as static correction, dewow, band-pass filtering, gain correction, background removal and migration. Time slices of processed GPR data were provided to the available image processing routines in the code. Most options enhanced the final image compared to the original one. This semi-automated process helps the interpreter to highlight the anthropogenic structures seen in the time/depth slices and export the preferred image in different available formats. [ABSTRACT FROM AUTHOR]

The Vienna Basin is the largest petroleum basin in Austria. As part of the ongoing seismic exploration for hydrocarbon resources in the basin, a large-N (N=10,530) dataset of continuous seismic recordings of the ambient seismic field was acquired in early 2018. Data were recorded on geophones (10-Hz corner frequency) and each location was made up of 12 or 24 densely clustered geophones, which were stacked to increase signal-to-noise ratio. Data were collected by these geophone clusters for up to 6 days, recorded only during day time. The station locations are distributed on a grid of roughly 40m by 400m and cover an area of about 500km2.We are in the process of imaging the upper part (top few kilometers) of the Vienna Basin by surface-wave tomography using surfaces waves extracted from ambient noise. For this, we study the properties of the ambient seismic field at higher frequencies (more than 5Hz) to ensure that the ambient seismic field contains surface waves that are appropriate for our purposes.Here, we present the first insights into the ambient seismic field in the Vienna Basin retrieved from this dataset. We perform beam-forming to detect noise sources outside of the study area, as well as Matched Field Processing to detect and locate noise sources at the surface inside of it. Generally, the field is dominated by cultural noise. Inside the network, the ambient seismic field at these higher frequencies is mainly generated by cars on streets, trains on railways, activity in town centers, and also wind turbines. Outside the network, we map the city of Vienna as a major contributor to noise. The ambient seismic field is comprised of body and surface waves and careful processing will be necessary to extract high-quality surface-wave measurements from the data. [ABSTRACT FROM AUTHOR]

The complexity of soils extends from the ecosystem-scale to individual microenvironments, where micro-scale interactions between microbiota, organic matter (OM) and mineral particles are thought to control the long-term fate of soil carbon and nitrogen. It is known that such biogeochemical processes show disproportionally high reaction rates within nano- to micro-meter sized isolated zones ('hot spots', e.g. rhizosphere) in comparison to surrounding areas. However, the majority of soil research is conducted on large bulk (> 1 g) samples, which are often significantly altered prior to analysis and destructively analyzed. Thus it has previously been impossible to study elemental flows (e.g. C and N) between plants, microbes and soil in complex environments at the necessary spatial resolution within an intact soil system. Especially the high sensitivity and lateral resolution of nano-scale secondary ion mass spectrometry (NanoSIMS) offers one smart approach to study the composition and fate of organic matter at biogeochemical interfaces of biological, organic and mineral soil microscale structures. Especially the possible imaging of stable isotopes (e.g. 13C, 15N) at high resolution enables the direct tracking of microscale processes via experiments using stable isotope enrichment. However, NanoSIMS can only provide meaningful results if applied in concert with other imaging (e.g. focused ion beam (FIB-SEM), micro-CT) and classical analytical techniques (isotopic and elemental analysis) to both integrate different spatial scales and multiple layers of information. Most notably classical bulk analyses are the backbone to connect the imaging of microscale soil features and phenomena active at Pedon and ecosystem scale. Exemplary a range of studies on properties and processes at soil microenvironments with increasing complexity will be presented, reaching from microaggregates to intact rhizosphere and soil cores. [ABSTRACT FROM AUTHOR]

In Poland, as well as in other European countries, societies are slowly undergoing an ageing process. The number of senior citizens has been increasing successively in the last dozen years. According to the forecast by the Central Statistical Office of Poland, by 2030 the number of citizens aged 65 and more will have exceeded 23% of the population. The trend will be accompanied by a number of logistical issues: where will those people reside? What kind of real estate will they choose? Will they opt for living with their families, in an old people's home, or will they rent their own apartments? The aim of the article is to identify the needs of both senior citizens as well as of persons over 50 who will become senior citizens in the 13-15 years in the context of preparing the urban space for the changes occurring in the society. The aim is to be achieved in two stages. The first stage consists of a survey conducted on a representative group of population (aged over 60) on the following: current residence, future residential preferences, architectural restraints, current trends (senior residential estates), and the choice of determinants of most relevance for senior citizens in this aspect. The second stage consists of application of the Voronoi diagrams for the visualisation of the current state of urban area in respect of availability of selected public amenities essential for senior citizens. The purpose of this is to show the existing sites and lack of such in selected urban areas in the context of making investment decisions related to residential needs of the analysed population. The issue should be of high importance for the local governments. Their role is to conduct sustained spatial policy, which is supposed to be based on the comprehensive picture of processes and phenomena constantly occurring in a town or city, so that urban space is planned in a way best serving the citizens, while taking into account their needs. In conclusion, the aim of the research is to be used in diagnosis of the current state of the readiness of local governments for the changes in the structure of the society and in identification of the most convenient location of amenities essential for senior citizens. [ABSTRACT FROM AUTHOR]

Since 1980, severe thunderstorms have caused nearly USD 3 billion in damages annually inEurope. Hailstorms account for a significant portion of thunderstorm-related losses, yet theclimatological aspects of large hail are not well understood. While the European SevereWeather Database (ESWD) contains the most comprehensive archive of hail observations inEurope, spatial and temporal inhomogeneities and discontinuities preclude the database frombeing used to create a representative long-term hail climatology. We therefore propose amethod to generate a hail event set for Central Europe by: 1) objectively identifying recenthail swaths from high-resolution radar data, and 2) combining the identified hail swathproperties with a probabilistic convective hazard model and ESWD hail observations tosimulate individual hailstorms. Such a hail event set will facilitate the evaluation of theclimatological aspects of large hail and time-dependent changes in the frequency ofhailstorms. In addition, the hail event set will be combined with exposure and vulnerabilitydata to produce detailed loss estimates and better understand the risk of large hail inEurope. Hail swaths occurring in Germany over a 4-year period (2015–2018) were identifiedusing the vertically integrated ice (VII) product from the Deutscher Wetterdienst (DWD).Specifically, we isolated continuous regions where 6-hour accumulated VII satisfied aminimum intensity (25 kg m−3) criterion, and then used Python image processing to generatehail swath ellipses and analyze their spatial characteristics. Next, for each 6-hour intervalduring the 1979–2018 period, we applied the convective hazard model to ERA-Interimreanalysis data to simulate the occurrence/non-occurrence of hail in each reanalysis grid cell.If hail was simulated in a given grid cell, we also predicted the number of hail swaths in thegrid cell during the 6-hour period. For each hail swath, we simulated the length,area, and orientation based on the spatial properties of the previously identified hailswaths. In addition, we simulated the maximum hail diameter based on the statisticaldistributions of observed hail diameter under various instability and deep-layer shearscenarios. Results from the VII hail detection algorithm suggest that large hail in Germany occursprimarily after 12 UTC, with a maximum frequency during the afternoon and early evening(12–18 UTC). While the majority of identified hail swaths affected areas less than 100km2, hail swaths exceeding 500 km2 were not especially uncommon. Hail swathsgenerally followed trajectories between 225˚ (southwest-to-northeast) and 315˚ (northwest-to-southeast) and were typically associated with right-moving thunderstorm cellswhen the synoptic-scale environment was moderately-to-strongly sheared. The simulated hailswaths exhibit a meridional gradient in the predicted frequency of large hail, with thelowest frequencies near the North and Baltic Seas, and the highest frequencies overthe eastern Alps. The number of simulated hail cases peaks in July and August,whereas large hail as reported in the ESWD is most prevalent between May and July. [ABSTRACT FROM AUTHOR]

ASIM was launched on April 2nd 2018 and integrated in the Columbus external Payload Facility. The full system is operational since June 2018. The aim of this presentation is to report the TGF imaging results for the first six months of ASIM operational phase. We discuss the correlation between the TGF footprint positions found and simultaneous events detected by MMIA cameras and GLD360 and WWLLN lightening data when available. Two imaging cases are analysed in detail to outline the imaging process. Preliminary correlations between TGF position in the MXGS FOV and CZT light curves morphology and are presented. Apparent angular source size, instrumental TGF spectral index and TGF position relationships at the MXGS FOV are explored. [ABSTRACT FROM AUTHOR]

Tall volcanic edifices are subject to gravitational instability and destruction. However, after sector collapse and large-scale destruction many volcanoes begin to regrow, gradually refilling the escarpment. Often, this process begins with the formation of lava domes in newly formed collapse scars. In the case of continued activity and the predominance of constructional over destructive processes, the collapsed volcanic edifice may develop into a conical morphology. However, details on regrowth from dome- to cone-building volcanoes have been suspected but barely investigated. Here, we present a comprehensive analysis of morphologic observations of edifice regrowth at Shiveluch, Bezymianny, and Avachinsky volcanoes in Kamchatka, Russia. We applied photogrammetric processing of aerial images obtained since 1967, and compare to the current tri-stereo Pléiades satellite images from 2017 and 2018. The processing was performed in the Erdas Imagine 2015 and Photomod 5 software, yielding a time series of three-dimensional point clouds for each of the volcanoes. Using these datasets we investigate morphologic, structural, and volcano-tectonic changes of the three studied edifices. We also derived volumes of eruptive material by calculating differences between consecutive point clouds. Thus, our detailed photogrammetric analysis enabled to identify different stages and locations of volcanic regrowth after sector collapse. We find that the regrowth of the conical edifice in the sector collapse escarpment develops in a systematic way, from initial endogenous to exogenous lava dome formation. We suggest a continuum model explaining the transition in volcanic regrowth and discuss possible mechanisms for it. [ABSTRACT FROM AUTHOR]

The aim of this presentation is to overview some applications of hyperspectral imaging for sediment core analysis in paleo-environmental studies. The increase of the sensibility and the resolution of sensors, as well as the development of performant data processing methods allow the analyze of natural sample in a very informative (chemical, physical and biological) and time effective way. Hyperspectral imaging is a method at the interface of spectroscopy and imaging where a spectrum is obtained for each pixel in the sample image. In laboratory conditions, many spectroscopic range can use this kind of method, from the X-ray to the infrared, also fluorescence and Raman. The dataset obtains is called a datacube due to his three dimensions (two spatials and one spectral), a fourth temporal dimension can be added. For that, both methods of image processing and spectral processing methods (chemometrics) can be used independently or together to improve sample knowledge. Quantitative modelling allows to establish a relationship between the spectra and destructive analysis (Particle Size Distribution, RockEval). These supervised models can then theoretically use on other samples without any sampling. Known structures can be discriminating accurately with classification methods, for example lamina or floods, with the labeling of some pixels to a known group. Unknown ones defined by chemical, physical or biological compounds can be determinates with similar methods and then explained by environment hypothesis. For these chemometrics methods, discriminant spectral wavelengths are identified and therefore sediment properties (chemical, physical or biological). Image processing of the core image or of the quantitative or qualitative map can be used to reveal surface structures (instant event, lamina) or local anomalies (coal, grain). The image resolution allows to see structures that the eye cannot see precisely, as infra-millimeter lamina. The use of spectro-spatial methods together and a time model allow to have a chronicle of structural sediment compositions and events.In this study, we use visible and near infrared hyperspectral imaging (pixel sizes: 60 and 240 micrometers) and X-Ray Fluorescence spectroscopy (240 micrometer) to analyze the lake Le Bourget sediment core (Savoie, France). Quantitative prediction models have been made for LOI550, Particle Size Distribution and fractions, some RockEval variables with correlation coefficients of the model upper than 0.8 and consistent uncertainties. Lamination counting was also realized with spectral classification methods and image processing, and this is in agreement with eye counting. With these two methods, seasonal variations can be separated in the eutrophic part of the core and compare. Hyperspectral imaging can also be used to find the optimal sampling areas for routine destructive analysis and to create predictive models. The use of several sensors, with data fusion methods, has shown that the increase of the amount of spectro-chemical information allowed to improve the robustness of predictive models. Hyperspectral imaging present already many possibilities to improve our knowledge on natural sample. [ABSTRACT FROM AUTHOR]

In geophysical prospecting with GPR, namely in archaeological environment, the background noise in radargrams impairs its interpretation if it is not possible to remove it effectively. This happens quite frequently when applying processing operations whose parameterization is limited, preventing their application in a personalized way to the data concerned.Similar to the methodology used to process noised images, for its improvement, using spatial filters, the noise in radargrams can be corrected through two approaches: in the spatial domain and in the frequency-wavenumber domain. The filters applied in the frequency-wavenumber domain consists in the modification of the image in its Fourier transform, allowing to act directly in its structure, whereas in the spatial domain the filters act in each pixel. The filter design is planned from the simultaneous analysis of magnitude spectrum and wavenumber distribution of the data. After the filter design, it is applied to the input data in the frequency-wavenumber domain using the multiplication. This simple operation requires a small computational effort, even involving a big amount of data.In this work, applying the proposed image processing approach to GPR-3D data, we produced results that correspond only to the useful information contained in the input data, allowing a better interpretation of the results and allowing a better evaluation about the existence of buried objects. The GPR-3D model produced after the noise estimation and its removal from the radargrams, allowed to detect structures that are not visible in the models obtained exclusively using commercial programs.We conclude that the presented approach improves the noise removal and consequently allows obtaining of better interpretations of the subsurface content.Acknowledgment: This work is co-funded by the European Union through the European Regional Development Fund, included in the COMPETE 2020 (Operational Program Competitiveness and Internationalization) through the ICT project (UID/GEO/04683/2013) with the reference POCI-01-0145-FEDER-007690, and in the INTERREG 2014-2020 Program, through the "Innovación abierta e inteligente en la EUROACE" Project, with the reference 0049_INNOACE_4_E. [ABSTRACT FROM AUTHOR]

The "Pas de l’Ours" landslide, located in the Queyras valley (Southeast France) reactivated inthe Spring 2017. The total moving mass is estimated at 17 million cubic meters, with a widthof 1 km and a length of 600 m, which makes it currently one of the largest active landslides inthe French Alps. In addition to the large deforming mass, numerous rockfalls and mudflowshave occurred and have severely damaged the road that is located at the foot of thelandslide. Ground-based instruments including Ground-based radar (GBSAR), GPS, andseismometers have been deployed on-site to monitor the landslide evolution. In addition, weare monitoring the landslide motion using SAR and optical data from the Sentinel-1 (S1) and-2 (S2) satellites. These datasets are processed separately: a) the Sentinel-1 (ascending anddescending) data are processed with the InSAR Scientific Computing Environment (ISCE), b)the Sentinel-2 are processed by Multiple-Pair Image Correlation (MPIC) and c) the GBSARdata are processed with IBIS Guardian, a persistent scatterer interferometry algorithm. Theanalysis of the landslide deformation pattern measured by the different techniques shows thatthe landslide is undergoing periods of rapid acceleration (> 0.5 m.d−1) and deceleration(m.yr−1). These complementary datasets are exploited to invert for the 3D surface displacement fora small section of the landslide where all three datasets overlap using a Weighted LeastSquare (WLS) inversion. The retrieved 3D displacements are compared to GNSS datasets atthree different locations. Our work highlights how complementary monitoring techniques canbe combined to retrieve the evolution of slope instabilities with various kinematicregimes, and offers a new perspective into the physical processes controlling landslidedynamics. [ABSTRACT FROM AUTHOR]

Early warning and satellite monitoring systems are nowadays widely used in flood risk management. Two examples operating at European scale are the European Flood Awareness System (EFAS), providing flood predictions for the major European rivers, and the Rapid Mapping (RM) component of the Copernicus Emergency Management Service (EMS), used for satellite-based flood mapping and monitoring. Here, we present a recently developed procedure to improve timeliness and precision of EMS satellite mapping using flood forecasts. Specifically, EFAS daily streamflow forecasts are used to derive event-based flood hazard maps, which are then combined with exposure and vulnerability information to evaluate expected impacts of the forecasted flood events. Areas with high potential impact are identified and used to pre-task and prioritize acquisition and processing of satellite images for monitoring flood events. The procedure is operational since 2017 and it has been applied in recent flood events, such as the 2018 floods in France and Spain. These applications demonstrated that the EFAS-RM integration can improve acquisition time and precision of satellite images, thus helping emergency response in case of severe floods. [ABSTRACT FROM AUTHOR]

In NE Italy, fast-moving landslides represent a significant threat both to the population and to the built environment. In the eastern portion of the Italian Alps, rockfalls are common and are often responsible for casualties or severe damage to infrastructure. To assess rockfall susceptibility, field surveys are crucial to accurately recognize the source areas of falling rock phenomena, to quantify the volumes of unstable blocks and identify the possible block trajectories.This work takes in detailed geological and geomorphological mapping, through the identification of dislodged blocks, fractures and fallen blocks on a much-frequented limestone cliff in the Trieste Karst used as a climbing spot. Moreover, a popular tourist path (the Napoleonica) lies at the foot of the climbing cliffs.The first identification of the most hazardous sections of the climbing sites were performed starting from field surveys and aerial photo analysis, morphometric and geomechanical characterization of the rock masses together with mapping activities.The field data were implemented with the analysis of UAV (Unmanned Aircraft Vehicle) images and videos; this innovative technique allows to reach and study hard-to-reach sites with high resolution and precision. The main aim of the study is to perform a semiautomatic morphometric and geomechanical characterization of two of the most hazardous sections. This analysis allowed to extract orientation and spacing of discontinuities of the studied walls and to recognize and measure unstable or pre-dislodged blocks by means of 3D orthophotos obtained from the UAV image processing. The main output of this work is a GIS-based map that will be the starting point for future 2D and 3D simulations in order to provide a rockfall susceptibility evaluation of the study area.Moreover, the methodology used could serve as model for the study of the other popular climbing sites, where the geological, geomorphological and climatic conditions are similar. The latters in fact are the predisposing factors to slope failures and consequently could cause casualties during climbing activities.The final outputs could form an important basis for the growth of investment in tourism policies aimed at increasing the quality of local attractions. [ABSTRACT FROM AUTHOR]

In the frame of the ERA-NET project SODEEP, we examine the impact of climate change on the permafrost degradation and related biophysical feedbacks over various spatial and temporal scales. Our aim is to assess the role of land-atmosphere interactions on the severity and frequency of extreme events over circum-Arctic land areas. Our focus regions are major bio-climatic zones of the Russian Arctic and sub-Arctic, which are not only extremely vulnerable habitats to the permafrost degradation due to climate change, but also exhibit a threat for the global climate due to frozen soil-bounded carbon. The data from long-term in situ observations and field studies as well as medium and high-resolution satellite-borne and aerial observations are combined with a state of the art hierarchy of climate models to identify missing or misrepresented permafrost-relevant key processes in numerical models. The rate of the permafrost temperature rise is different in different climatic zones. The maximum growth rate is observed in the southern tundra, the minimum in the southern forest-tundra and northern taiga, where the permafrost temperature is close to 0°C. The ground temperature rise reduce the bearing capacity. The layers of the "zero curtain" with a thickness of several meters occur at temperatures close to 0°C. Its presence inhibits the degradation of permafrost.Here, we present the project and first results. The latter comprise preliminary results from field campaign and development of a permafrost temperature dynamics and active layer thickness database, processing of optical satellite images (spatio-temporal changes related to lake extent and trend of several multi-spectral indices, i.e. NDVI, NDMI, Tasseled Cap indices based on Landsat and Sentinel-2 images selected from archives and acquired between 1987-2018), as well as the progress in understanding and implementation of permafrost related processes into the regional climate model REMO and simulations with Geophysical Institute Permafrost Laboratory (GIPL) model to reproduce historic permafrost distribution 1960-2009 on the pan-Arctic at 0.5° spatial resolution. Thermal and hydrologic properties of the ground material and snowpack are parameterized using vegetation cover, surface geology and extensive empirical observations. This model will be augmented later in the project with advances in techniques as geographic object-based image analysis for interpretation of remote sensing imagery through automated identification of landscape changes and used to update permafrost scheme in climate models. The approach using empirical regional trends detected from long records was applied for mapping of current and future (2050 and 2100) permafrost temperature. Trends are identified by taking into account the current changes in climate and permafrost characteristics.This work is funded in the frame of ERA-Net plus Russia. TSU is supported by MOSC RF # 14.587.21.0048 (RFMEFI58718X0048), AWI and HZG are supported by BMBF (Grant no. 01DJ18016A and 01DJ18016B), and WUT by a grant of the Romanian National Authority for Scientific Research and Innovation, CCDI-UEFISCDI, project number ERANET-RUS-PLUS-SODEEP, within PNCD III. [ABSTRACT FROM AUTHOR]

Global warming has a large impact on the timing of phenological phases. Due to climate change phenological phases occurred earlier during the last years.Photography has been used for several years now for the study of the phenological phases. Most of the photographs are analysed manually. There are however some few studies dealing with the automatic analysis of webcam images for the detection of the different phenological phases. One of the major problems are changes in the illuminance of the environment. Satisfying results were among others obtained using the greenness index for the detection of leaf growth. More difficulties are encountered trying to detect flowering because of white background objects which lead to errors in pixel detection. Using web cam images of apricot trees from Wachau, Austria we successfully tested the greenness index for the detection of green vegetation. A method for the detection of flowering was developed using data of one year and was tested using data of supplementary years. An analysis of the accuracy of the method is presented.In the end, the timing of the phenological phases is analysed as a function of meteorological conditions (temperature, precipitation) for the time period 2013-2018. [ABSTRACT FROM AUTHOR]

Satellite-based rainfall estimates are a valuable source of information. This is especially the case for regions with low gauge density, such as many regions in sub-Saharan Africa. Satellites do not measure rainfall directly, instead it is derived from other measured variables. This can lead to errors in the position of the rainfall events beside the errors in their intensity. Such errors are particularly critical for localized events such as the convective storms occurring during the monsoon season which produce most of the rainfall in sub-Saharan Africa. We propose a morphing approach to adjust both the intensity and the position of satellite estimates with gauge data. Morphing is a well-known image processing technique that transforms an image into another, using cross-dissolving and warping. Thus, both the intensity and the position are taken into account. Our approach uses an automatic registration, so only the two images are needed as input. There is no need to predefine the rain events, or any other features. The automatic morphing method is applied for a rainfall event during the monsoon season over the Volta Basin region. We gauge-adjust the satellite-based estimates IMERG-early (Integrated Multi-Satellite Retrievals for GPM), using the gauge network from TAHMO (Trans-African Hydro-Meteorological Observatory). The automatic morphing has encouraging results when the precipitation fields are similar enough. [ABSTRACT FROM AUTHOR]

The artificial catchment "Hühnerwasser" was built in a post-mining landscape, as a field experiment to observe and monitor early-development ecosystems at first catchment scale. As vegetation is a key driver of hydrological catchment behavior, spatial distribution and temporal dynamics of vegetation affects water redistribution from plot to catchment scale. In the context of early ecosystem development, quantifying changes in vegetation structures is an obvious indicator for state transitions. The first years of ecosystem development at the Hühnerwasser catchment showed rapidly increasing complexity of emerging structures associated with rising vegetation cover and increasing number of plant species. This work will focus on characterizing vegetation cover using aerial images aiming to describe spatial structures and how those evolve in time. The early stages are especially of interest. The structure is therefore characterized by the area of the catchment covered by vegetation, the number of vegetation patches, the mean and maximum patch size and a form factor (area of patch divided by its perimeter).Aerial images with a resolution at cm scale were taken once per year from 2007 to 2018. Binary maps are generated by setting thresholds for red, green and blue channels to differentiate between vegetation cover and bare soil. To evaluate the consistency of the binary images of each channel these images were stacked and compared. The performance of the method was tested by using a set of combinations of thresholds and a comparison with manual mapping of vegetation cover at an image subset was made.The blue channel seems to be very sensitive to detect vegetation and a better differentiation of vegetation and dark/wet soil can be achieved by setting the thresholds of the channels in a specific order. The structures derived by the classification into vegetated and bare soil are more important in the early years of ecosystem development. In those years (2007 to 2011) the most changes took place. As time advances vegetation became less patchy and other characteristics need to be implemented to describe the vegetation cover, taking into account different plant functional types. [ABSTRACT FROM AUTHOR]

Freeze-thaw weathering is one of the main contributors to disintegration of mineral buildingmaterials. Following the main theories on crystallization in porous media, unfrozen porewater is possibly drawn to sites of ice crystal growth, where crystallization pressure exertsforce on the pore walls. Research that concentrates on these processes often use relativelylarge experimental setups and samples, and focus on proxies, such as length change andtemperature of a sample. Therefore, mostly indirect evidence on the acting damagingmechanisms has been gathered over time. In situ monitoring of the water redistribution is afirst logical step towards a direct assessment of the acting damage mechanisms. X-raycomputed micro-tomography (μCT) has proven its value in earlier frost-related research andcan be used to locate different phases within the pore space of different mineralmaterials. A custom-made freezing cell was installed on HECTOR (UGCT) to performfreeze-thaw cycles on a water-saturated porous limestone core. After producing a firstμCT scan at a constant temperature of 4 ˚ C, the sample was cooled and kept at aconstant temperature of -8 ˚ C and a new full μCT scan was taken. This cycle wasrepeated four times. Through image processing and analysis, it was possible toobserve the volumes of water and ice throughout the freeze-thaw cycles. Moreover, byapplying differential imaging, we were able to filter out differences between water-icedistributions, enabling us to observe several evolutions. Despite some restrictions andconsiderations, these observations help to improve the understanding of pore-scale processes. [ABSTRACT FROM AUTHOR]

Quantification of sulfur dioxide (SO2) emission flux is a fundamental task in volcanology to have insights of the composition and the spatial evolution of volcanic plumes. The ground based InfraRed hyperspectral imager HyperCam, from Telops Company, was deployed during IMAGETNA campaign in 2015 and provided high spatial and spectral resolution images of Mt Etna plume. The spectral range of the hyperspectral imager is [7.7 - 11.8 µm] and the measured images contained 320 x 64 pixels with a spectral resolution of 2 cm-1. To process hyperspectral images in quasi real-time, a fast and reliable radiative transfer retrieval model is required. The LATMOS Atmospheric Retrieval Algorithm (LARA), used to retrieve the slant column densities of SO2, includes an accurate line-by-line radiative transfer model and an efficient minimization algorithm of the Levenberg-Marquardt type. But the calculation time remains too high to infer near real time (NRT) estimation of SO2 fluxes. As first, to reach NRT target, a classification methodology of the brightness temperature spectra was developed and then applied on each measured sequence to significantly decrease the processing time. One image previously took a week of calculation to be retrieved. The classification of the spectra allows the retrieval of a complete measurement sequence of the field campaign (~400 images) in only a couple of hours. The accuracy of the methodology was confirmed, by comparing the SO2 slant column density images obtained after classification with the one obtained by the accurate and time expensive pixel by pixel retrieval processing. The LARA model, the spectra classification methodology and a comparison of the results will be presented. [ABSTRACT FROM AUTHOR]

The secondary succession is currently an often observed process taking place on agricultural areas. Even though this phenomenon can have a highly positive impact on environment e.g. climate, carbon storage or biodiversity, it can also be perceived as a threat when present on Natura 2000 habitats. The process of succession taking place on the protected areas needs to be therefore monitored in order to understand and assess it's influence on ecosystems' functioning. This can be achieved and automated by the use of remote sensing data and novel image processing techniques. In our work we took advantage of the simultaneously acquired, high-resolution aerial remote sensing data - hyperspectral imagery and LiDAR data - to develop a complete methodology for monitoring the succession process. The algorithms implemented within the methodology included i.a. Structure from Motion and Dense Image Mapping for processing the archival images, segmentation and Voronoi Tessellation for succession spatial extent delineation, as well as Machine Learning Random Forest classifier, Recursive Feature Elimination and t-distributed Stochastic Neighbor Embedding algorithms used for succession species differentiation. The prepared methodology was tested on seven research areas located in Poland. In our presentation we would like to show a set of different maps being the products of the developed methodology: a map depicting the spatial extent of shrubs and trees forming the succession process, maps showing the internal and external threat to the habitats, a map showing the spatial extent of succession species present on an example research area and maps presenting the dynamics of the succession process. We will also explain how different data and algorithms can be used to develop these products and how different management units can use them to make informed decisions about Natura 2000 protected areas. The research is a part of the project HabitatARS (BIOSTRATEG2/297915/3/ NCBR/2016) The innovative approach supporting monitoring of non-forest Natura 2000 habitats, using remote sensing methods, co-financed by The National Centre for Research and Development. It's results are now being implemented and commercialized by the MGGP Aero Sp. z.o.o. company. [ABSTRACT FROM AUTHOR]