Search

Image segmentation algorithms are a type of image classifier that assigns a label to each individual pixel in an image. U-Nets, initially developed for the analysis of biomedical images and now widely used in a variety of fields, are an example of such algorithms. It has been shown that U-Nets are specially interesting when working with small training datasets and combined with data augmentation techniques.In this study, we used satellite images with labelled landslide masks from known events to train a U-Net to identify areas of potential landslide. These landslide masks are time-consuming to create, resulting in a small initial training set. Even when working with U-Nets, the success of machine learning and AI tools depends on the availability and quality of training data, as well as the algorithm settings during the training process. Tuning machine learning models to achieve the best performance possible from limited amounts of data is important to generate trustworthy results that can be used to advance the knowledge of landslide events around the world.Here, we show the differences in algorithm performance as we use different types of data augmentation and model parameters. We also explore and assess the effects on performance of options such as including different satellite bands, terrain information and alternative colour band representations.

SAR images can be used to measure changes in the surface of the Earth over time using Time Series Interferometric Synthetic Aperture Radar (TS-InSAR) techniques. TS-InSAR enables the detection and measurement of very small changes in surface deformation, often on the order of millimetres or less. This makes it a powerful tool for monitoring a wide range of natural and man-made phenomena, such as tectonic activity, subsidence, ground water extraction, and the behaviour of engineered structures like buildings and bridges. While TS-InSAR provides deformation measurements, further analysis must be taken to understand the underlying cause of the deformation. In this study, a novel framework has been developed to extract the vast amount of information embedded within the large number of ground deformation Measurement Points (MPs) derived from the Small BAseline Subset (SBAS; Berardino et al., 2002) TS-InSAR technique. The proposed automatic data-mining approach begins with clusterization the TS-InSAR MPs by applying a nonlinear dimensionality-reduction technique, Uniform Manifold Approximation and Projection (UMAP; McInnes et al., 2018), prior to performing clustering with Hierarchical Density based Spatial Clustering of Applications with Noise (HDBSCAN; Campello et al. 2013) in order to group together MPs exhibiting similar deformation behaviour on a large scale. Next, every extracted cluster time series is further investigated by applying a piecewise linear function as a method to detect and quantify accelerations and decelerations of deforming areas.A test of the method has been conducted over the Bandung Basin (Indonesia) using Sentinel-1 data from October 2015 to December 2020. Application of the method provides an objective way to identify changes in displacement rates over time and provides a wealth of information on the dynamics of surface displacement over a large area. The displacement rates, their spatial variation, and the timing and location of accelerations and decelerations can be used to investigate the physical behaviour of the deforming ground by linking the timing and location of changes in displacement rates to causal and triggering factors.ReferencesBerardino, P., Fornaro, G., Lanari, R., Sansosti E. A new algorithm for surface deformation monitoring based on small baseline differential SAR interferograms. IEEE Transactions on Geoscience and Remote Sensing, 40 (11) (2002), pp. 2375-2383.Campello, R.J., Moulavi, D., Sander, J. Density-based Clustering Based on Hierarchical Density Estimates. In Advances in Knowledge Discovery and Data Mining, Proceedings of the Pacific-Asia Conference on Knowledge Discovery and Data Mining; Pei, J., Tseng, V.S., Cao, L., Motoda, H., Xu,G., Eds.; Springer: Berlin, Germany, 2013; pp. 160–172 McInnes, L. and Healy, J. UMAP: uniform manifold approximation and projection for dimensionreduction. Preprint at https://arxiv.org/abs/1802.03426 (2018).

To extract the most information from a national InSAR dataset it is imperative to understand the mechanisms leading to motion and how these manifest in an InSAR dataset such as the EGMS. The British Geological Survey (BGS) have been at the forefront of UK InSAR ground motion interpretation for the past 22 years; in projects such as ESA’s Terrafirma, thematic FP7 projects e.g. PanGeo and SubCoast, and provide expert advice to the UK government surrounding potential fracking sites and CCS sites along with a sustained research programme. BGS also produce national hazard susceptibility mapping, known as GeoSure, which is routinely used by insurance companies to assess ground motion hazards. The study of many epochs of InSAR data for many study sites provide the opportunity to examine how patterns of subsidence have evolved with time and how this relates to the processes taking place. This presentation will illustrate the evolution of ground motions within the UK over the last 30 years, case studies will highlight the lessons learnt especially with respect to how the change in geological process manifests in the InSAR signal. For example, a striking change in ground motion patterns over a time period occurred in the Newcastle and Durham Coalfield (Gee et al., 2017) where a dramatic change in the pattern of motion was observed between subsidence in the 1990’s ERS data and uplift in the 2000’s ENVISAT data. This was found to relate to change in minewater pumping. Therefore, to create a national interpreted ground motion product it is important to not only understand the mechanism of motion but also understand how that process changes over time. Recent research has focused on the prediction of which UK hazards will be visible and measurable by InSAR (Novellino et al., 2023), and on the prediction of likely natural motion rates for the natural geological subsidence processes (Jones et al., 2013). BGS have developed automatic AI and ML tools which examine not only the InSAR average velocity but also the time series to group areas of similar motion characteristics and to then detect when changes occur (Festa et al., 2023, Hussain et al., 2021.). The application of such tools to the EGMS time series and integration of results with BGS GeoSure national hazard susceptibility datasets provide a pathway to the ongoing interpretation of national GB wide InSAR datasets. The above experience puts the BGS in a unique position for the exploitation of the new Copernicus European Ground Motion Service which represents the first freely available national UK InSAR dataset. Over the coming years it is BGS’ ambition is to apply our experience of UK hazard susceptibility, how these hazards manifest in InSAR data, our fledgling automated interpretation tools and visibility mapping to produce the first dynamic country-wide interpreted ground motion information layer; a value added product which not only tells the user what the motion is but also identifies the likely reason for the motion along with forecasts of how such motion might evolve in the coming years.

Landslides in remote or uninhabited regions can be undocumented, leaving gaps in landslide inventories which are a key input for hazard and risk assessments. This can lead to landslide events being missing from research studies, and contribute to a bias in the events used for training of machine learning models.In this work we use satellite images, terrain information, and labelled examples of landslides to train a convolutional neural network (U-Net), for the purpose of adding previously undocumented and new landslides to inventories. This model segments the input images and highlights the pixels it labels as landslides.Our work focusses on landslides with a range of types and triggers, so that the model is exposed to a variety of training data. We describe the key properties of the landslides in the training set, and discuss the implications for future uses of the trained model.

Many of the UNESCO World Heritage Sites face geological threats which could have negative effects on the value, integrity and accessibility of their heritage assets. A relevant example is the Derwent Valley Mills UNESCO World Heritage site, one of the key sites of Britain's industrial revolution of the 18th century and located along the Derwent River Valley. Individual susceptibility scenarios of natural hazards in the area like collapsible deposits, compressible ground, debris flow, landslide, running sands, shrink-swell, soluble rock and flooding (both riverine and groundwater) are available, but a comprehensive product able to support disaster mitigation measurement and land planning still does not exist. On this basis, a multi-hazard susceptibility analysis was completed with the added benefit of reducing the complexity and providing a methodological framework for multi-hazard estimation. The analysis was completed in a GIS environment through an Analytical Hierarchy Process (AHP) multicriteria decision-making process. Since the AHP method is affected by a user selection bias, a quantitative Relative significance index was derived to rank the AHP factors during the susceptibility estimation. This index suggests that flooding is the principal natural hazard for the Derwent Valley Mills UNESCO World Heritage site. The multi-hazard susceptibility map also indicates that most of the areas where the mills are located are subject to significant susceptibility to natural hazards.

Delineating hydrothermal alteration and supergene caps is fundamental for mineral exploration of sulfide ores. The aim of this study is to apply a multi-scale workflow based on hyperspectral remote and proximal sensing data in order to delineate hydrothermal dolomitization and supergene alteration associated with the Mississippi Valley-Type Zn-Pb(-Ag) deposit of Jabali (Western Yemen). The area was investigated through hyperspectral images derived from the new launched Italian Space Agency’s PRISMA satellite, which has a higher spectral resolution compared to multispectral sensors and covers the mineral-diagnostic wavelength regions (such as the 2100 nm to 2300 nm range) with a Signal to Noise Ratio (SNR) ≥ 100. Spectral mineral maps were produced through the band ratios method using specific feature extraction indices applied to the hyperspectral satellite data. The results were validated by using Visible Near InfraRed (VNIR) to Short Wave InfraRed (SWIR) reflectance spectra, mineralogical (XRPD) and geochemical (ICP-ES/MS) analyses on rock samples collected in the Jabali area. The dolomites footprint was mapped using a PRISMA Level 2C image, by enhancing the spectral differences between limestones and dolomites in the SWIR-2 region (major features centered at 2340 nm and 2320 nm, respectively). Gossans were detected due to the Fe3+ absorption band in the VNIR region at 900 nm. The Zn-Pb mineralized area, extended for approximately 25 km2, was thus identified by recognizing gossan occurrences in dolomites. The study demonstrates that the PRISMA satellite is effective in identifying Zn-Pb mineralized outcrops in sedimentary basins.

Analysis of shoreline change is fundamental to a broad range of investigations undertaken by coastal scientists, coastal engineers, and coastal managers. Multispectral Satellite Imagery (MSI) provides high resolution datasets that allow coastlines to be monitored more frequently and on a global scale. The Landsat and Sentinel-2 MSI datasets are free for public use, which has increased the frequency of studies focusing on coastal change using satellite imagery. However, despite access to global and free satellite imagery, a method has yet to be developed to monitor different shoreline types and indicators globally, as not all shorelines are sandy beaches, and the waterline cannot be representative of all shoreline changes. The review paper introduces different techniques used currently to extract shoreline features, including water indexing, Machine Learning (ML) and segmentation methods. We presented here a comprehensive review of range of the methods available for shoreline extraction from MSI and discuss why some shoreline features have been identified using multispectral satellite imagery and others not. This approach helps to signal where the gaps are on the current methods for shoreline extraction and provides a roadmap of the key challenges that prevents MSI to be used for understanding shoreline changes at a global scale.

The lateral collapse of Anak Krakatau volcano, Indonesia, in December 2018 highlighted the potentially devastating impacts of volcanic edifice instability. Nonetheless, the trigger for the Anak Krakatau collapse remains obscure. The volcano had been erupting for the previous six months, and although failure was followed by intense explosive activity, it is the period immediately prior to collapse that is potentially key in providing identifiable, pre-collapse warning signals. Here, we integrate physical, microtextural and geochemical characterisation of tephra deposits spanning the collapse period. We demonstrate that the first post-collapse eruptive phase (erupting juvenile clasts with a low microlite areal number density and relatively large microlites, reflecting a crystal-growth dominated regime) is best explained by instantaneous unloading of a relatively stagnant upper conduit. This was followed by the second post-collapse phase, on a timescale of hours, which tapped successively hotter and deeper magma batches, reflected in increasing plagioclase anorthite content and more mafic glass compositions, alongside higher calculated ascent velocities and decompression rates. The characteristics of the post-collapse products imply downward propagating destabilisation of the magma storage system as a response to collapse, rather than pre- collapse magma ascent triggering failure. Importantly, this suggests that the collapse was a consequence of longer-term processes linked to edifice growth and instability, and that no indicative changes in the magmatic system could have signalled the potential for incipient failure. Therefore, monitoring efforts may need to focus on integrating short- and long-term edifice growth and deformation patterns to identify increased susceptibility to lateral collapse. The post-collapse eruptive pattern also suggests a magma pressurisation regime that is highly sensitive to surface-driven perturbations, which led to elevated magma fluxes after the collapse and rapid edifice regrowth. Not only does rapid regrowth potentially obscure evidence of past collapses, but it also emphasises the finely balanced relationship between edifice loading and crustal magma storage.

The lateral collapse of Anak Krakatau volcano, Indonesia, in December 2018 highlighted the potentially devastating impacts of volcanic edifice instability. Nonetheless, the trigger for the Anak Krakatau collapse remains obscure. The volcano had been erupting for the previous six months, and although failure was followed by intense explosive activity, it is the period immediately prior to collapse that is potentially key in providing identifiable, pre-collapse warning signals. Here, we integrate physical, microtextural and geochemical characterisation of tephra deposits spanning the collapse period. We demonstrate that the first post-collapse eruptive phase (erupting juvenile clasts with a low microlite areal number density and relatively large microlites, reflecting a crystal-growth dominated regime) is best explained by instantaneous unloading of a relatively stagnant upper conduit. This was followed by the second post-collapse phase, on a timescale of hours, which tapped successively hotter and deeper magma batches, reflected in increasing plagioclase anorthite content and more mafic glass compositions, alongside higher calculated ascent velocities and decompression rates. The characteristics of the post-collapse products imply downward propagating destabilisation of the magma storage system as a response to collapse, rather than pre-collapse magma ascent triggering failure. Importantly, this suggests that the collapse was a consequence of longer-term processes linked to edifice growth and instability, and that no indicative changes in the magmatic system could have signalled the potential for incipient failure. Therefore, monitoring efforts may need to focus on integrating short- and long-term edifice growth and deformation patterns to identify increased susceptibility to lateral collapse. The post-collapse eruptive pattern also suggests a magma pressurisation regime that is highly sensitive to surface-driven perturbations, which led to elevated magma fluxes after the collapse and rapid edifice regrowth. Not only does rapid regrowth potentially obscure evidence of past collapses, but it also emphasises the finely balanced relationship between edifice loading and crustal magma storage.

In this paper, we aim to define a procedure of flood hazard assessment applicable to large river basins in which flood events can be induced/sustained by the full basin area or by fractions of the total area as functions of the extent of the triggering precipitation event. The proposed procedure is based on a combined approach accounting for (1) the reconstruction of intensity–duration–frequency curves expressing the magnitude in terms of intensity for multiple return periods; (2) the application of the soil conservation service method for runoff estimation from a selected rainfall scenario considering some characteristics of the basin (i.e. soil type, land use/treatment, surface condition, and antecedent moisture conditions); (3) 2D hydrodynamic modelling conducted by the HEC-RAS model using runoff hydrographs as hydrological input data; (4) the reconstruction of flood hazard maps by overlaying multiple inundation maps depicting flood extent for different return periods. To account for the variability in the extent of the triggering precipitation event and the resulting input hydrograph, multiple contributing areas are considered. The procedure is tested at the archaeological site of Sybaris in southern Italy, which is periodically involved in flood events of variable magnitude. The obtained results highlight that the variable extent of the floodable area is strongly conditioned by the extent of the contributing area and return period, as expected. The archaeological site is always involved in the simulated flooding process, except for the smallest contributing area for which only a 300-year event involves this part of the site. Our findings may be useful for developing and supporting flood risk management plans in the area. The developed procedure might be easily exported and tested in other fluvial contexts in which evaluations of multiple flood hazard scenarios, due to the basin geometry and extent, are needed.

Coastal dunes play an important role in coastal erosion risk management, where they act as a dynamic natural sea defence line. Formby coast is part of the Sefton coast in the Northwest of England and is one of the largest and most rapidly evolving sand dune systems in the UK. Such dune systems require continuous comprehensive monitoring activity to understand their dynamics. In this research, we investigate the use of airborne LiDAR digital terrain model DTMs for monitoring the dynamics of the sand dunes at Formby between 1999 and 2020. We found that the rate of elevation change for the beach and the dune areas ranges from −0.78 to 0.02 m/year and −0.92 to 0.73 m/year, respectively. The beach and the frontal dunes have had significant sand erosion, while the inner dunes gained sand during the measurement period. Vegetated areas remained unchanged due to the impact of vegetation in stabilizing the movement of the dunes. Formby beach had a volume loss of about 907,000 m3 in the last 21 years, while the dunes had a volume increase of about 1,049,000 m3 over the same period. The total volume of the entire dune system, consisting of both the beach and dune areas, remained unchanged, which indicates that the growth of the inland dunes is fed by sand from the beach. All the volumetric changes occurred due to sand redistribution within the system, with erosion along the beach, and deposition and erosion in the dune areas.

The Campania region has been recurrently hit by severe landslides in volcanoclastic deposits. The city of Naples, and in particular the Camaldoli and Agnano hills, also suffered several landslide crises in weathered volcanoclastic rocks as a consequence of intense rainfalls or wildfires. This work provides an updated landslide database for the suburbs of Naples. The obtained database consists of about 1322 landslides considering available information dated back to 1816, identified thanks to the historical newspaper's examination; 62 landslides were recorded between 2019 and 2020. Furthermore, the achieved database is the result of the standardization of available information from different sources, organized in a completely different way. The newly identified phenomena have been recognized thanks to a combination of optical satellite imagery with Google Earth, Sentinel-1 radar satellite imagery and field investigation. The implemented methodology is based on change detection analysis of satellite imagery by using polarimetric features. Subsequently, output derived from the segmentation procedure of satellite images have been compared with field trip observations. The main purpose of this procedure is to emphasize areas where land cover changes, potentially related to slope failures, occur in the Phegraean Fields, facilitating the following possible phases of mapping and/or field survey. Eventually, this type of information is expected to help decision-makers with land planning and risk assessment.

The Dec. 22, 2018 lateral collapse of the Anak Krakatau (AK) volcano in the Sunda Straits of Indonesia discharged volcaniclastic material into the 250 m deep caldera southwest of the volcano and generated a large tsunami, causing runups of up to 85 m in the near-field, and 13.5 m in the far-field, on the nearby coasts of Sumatra and Java. The tsunami caused 437 fatalities, the greatest number from a volcanically-induced tsunami since the catastrophic explosive caldera-forming eruption of Krakatau in 1883 and the sector collapse of Ritter Island in 1888. For the first time in over 100 years, the 2018 AK event provides an opportunity to study a major volcanically-generated tsunami that caused widespread loss of life and significant damage. Here, we present numerical simulations of the collapse and tsunami generation, propagation, and coastal impact, with state-of the-art numerical models, using both a new parametrization of the collapse and a near-field bathymetric dataset based on our 2019 field surveys and satellite images. These subaerial and submarine data sets are used to constrain the geometry and magnitude of the landslide mechanism, which show that the primary landslide scar bisected the AK edifice, cutting behind the central vent and removing 50% of its subaerial volume. The primary landslide volume is estimated to range from 0.175–0.313 km3, based on uncertainties in the shape of the submerged part of the failure plane. This is supported by an independent estimate of the primary landslide deposit volume of 0.214 ± 0.036 km3. Given uncertainties in the failure volume, we define a range of potential failure surfaces that span these values in 4 collapse scenarios of volume ranging from 0.175 to 0.313 km3. These AK collapses are modeled, assuming either a granular or viscous fluid rheology, together with their corresponding tsunami generation and propagation. Observations of a single tsunami, with no subsequent waves, are consistent with our interpretation of landslide failure in a rapid, single phase of movement rather than a more piecemeal process, generating a tsunami which reached nearby coastlines within ~30 min. For both modeled rheologies, the 0.224 km3 collapse (second and preferred scenario) most successfully reproduces the near- and far-field tsunami flow depth and runup observed in all post-event field survey results, tide gauge records, and eyewitness reports to date, suggesting our estimated landslide volume range is appropriate. This event highlights the significant hazard posed by relatively small-scale lateral volcanic collapses, which can occur en-masse, without any precursory signals, and are an efficient and unpredictable tsunami source. Our successful simulations demonstrate that current numerical models can accurately forecast tsunami hazards from these events. In cases such as Anak Krakatau's, the absence of precursory warning signals, together with the short travel time following tsunami initiation present a major challenge for mitigating tsunami coastal impact, stressing the need to develop and install early warning systems for such events.

Subsurface geonergy can induce ground motion and seismicity, however a scarcity of observations usually obscures the mechanisms underpinning such behaviour. Here, we analyse Interferometric Synthetic Aperture Radar (InSAR) data from ERS, ENVISAT and Sentinel-1 satellites for the period 1995–2017 and interpret ground deformation in the area of the planned Cheshire UK GeoEnergy Observatory ahead of facility contruction. Ground motion is dominated by the compaction of tidal flat deposits overlying two paleo-valleys, trending NNW–SSE. The western paleo-valley experienced faster subsidence rates in the period 1995–2007, whereas the eastern paleo-valley subsided faster in the period 2016–2017. The research highlights how baseline assessment can help differentiate natural variation from any anthropogenic effects associated with the growth of new subsurface technologies.

Multi-temporal Interferometric Synthetic Aperture Radar (MTInSAR) is a solid and reliable technique used to measure ground motion in many different environments. Today, the scientific community and a wide variety of users and stakeholders consider MTInSAR a precise tool for ground motion-related applications. The standard product of a MTInSAR analysis is a deformation map containing a high number of point-like measurement points (MP) which carry information on ground motion. The density of MPs is uneven, and they cannot be extracted continuously at large scale due to geometrical distortions and unfavourable landcover. It is a good practice to assess the feasibility of the interferometric analysis ahead of data processing. This technical note proposes a ready-to-use set of tools aimed at updating existing methods for modelling the effects of local topography and land cover on MTInSAR approaches. The goal of the tools is to provide InSAR experts and non-experts with a fast and automatic way to derive visibility maps, useful for pre-processing screening of a target area, and to forecast the expected density of MP over a specified area. Moreover, the visibility maps are a valid support for users to better understand the available standard and advanced interferometric results. Two workflows are proposed: the first generates the so-called Rindex map (Ri_m) to estimate the influence of topography on MP detection, the second is used to derive a land cover-calibrated Ri_m seen as a probabilistic model for MP detection (MPD_m). The proposed set of tools was applied in the context of the Alpine arc, whose climatic, morphological, and land cover characteristics represent a challenging environment for any interferometric approach.

With the advances of ESA’s Sentinel-1 InSAR (Interferometric Synthetic Aperture Radar) mission there are freely available remote sensing ground deformation observations all over the globe that allows continuous monitoring of natural hazards and structural instabilities. The Digital Environment initiative in the UK aims to include these remote sensing data in the effort of forecasting and mitigating hazards across the UK.However, analyses of low coherence areas (e.g. forested and vegetated areas) with conventional InSAR methodologies are difficult to perform due to the limiting factor of temporal and geometric decorrelation. Even the application of the permanent scatterer (PS) technique may not be successful when there is a low density of stable radar targets. Using artificial reflectors with high radar cross section (RCS) can be a way of overcome this limitation and achieve measurements with a good signal-to-clutter ratio (SCR).In order to be able to include Sentinel-1 data in the UK’s Digital Environment it is important to understand the advantages and limitations of these observations and interpret them appropriately. The Hollin Hill landslide observatory in North Yorkshire is used by the British Geological Survey in their efforts to understand landslide processes, and to trial new technologies and methodologies for slope stability characterisation and monitoring.We present InSAR results of the Hollin Hill landslide where a variety of ground-based geophysical measurements (e.g. GPS, Electric resistivity tomography, meteorological observations) are available for comparison with InSAR data. We use Sentinel-1 InSAR data acquired between Oct 2015 and Jan 2021 to study the behaviour of this landslide. We find that the Line of Sight component of the down-slope movement is 2.7 mm/yr in the descending track, and 7.5-7.7 mm/yr in the ascending track. The InSAR measurements also highlight the seasonal behaviour of this landslide.In July 2019 six corner reflectors were installed to improve the coherence of the InSAR measurements, especially in the ascending acquisition mode. We present comparison with ground-based measurements such as the movement recorded by the GPS measurements of the pegs of the ERT survey or the moisture recorded by the various instruments at the site, and show the improvement introduced by the corner reflectors.In addition we present results of an experiment that explores the use of smaller corner reflectors for potential urban applications of infrastructure monitoring. A single corner reflector needs to be at least ~67cm wide and tall to be seen by the Sentinel-1 satellites. We show that by placing 4 reflectors with 33cm dimensions in the same pixel coherent signal can be acquired. It is feasible to install small reflectors on bridges, tall buildings, or incorporate “corner-like” features in newly built structures,but care needs to be taken on the precise spacing of the reflectors to avoid destructive interference. Continuous monitoring of infrastructure with remote sensing and machine learning can alert to potential failures where further investigation is needed.

In this work, we provide an innovative route for analysing urban expansion and population growth and their link to the consumption of construction materials by combining satellite data with material consumption analysis within the Hanoi Province (Vietnam). Urban expansion is investigated with the use of landcover maps for the period 1975–2020 derived from satellite. During this period, artificial surfaces and agricultural areas have increased by 11.6% and 15.5%, respectively, while forests have decreased by 26.7%. We have used publicly available datasets to calculate and forecast the construction materials consumption and measure its statistical correlation with urban expansion between 2007 and 2018. Our results show that official figures for sand consumption are currently underestimated, and that by 2030, steel and sand and gravel consumption will increase even further by three and two times, respectively. Our analysis uses a new method to assess urban development and associated impacts by combining socio-economic and Earth Observation datasets. The analysis can provide evidence, underpin decision-making by authorities, policymakers, urban planners and sustainability experts, as well as support the development of informed strategies for resource consumption. It can also provide important information for identifying areas of land conservation and ecological greenways during urban planning.

This paper describes a novel methodology where Machine Learning Algorithms (MLAs) have been integrated to assess the landslide risk for slow moving mass movements, processes whose intermittent activity makes challenging any risk analysis worldwide. MLAs has been trained on datasets including Interferometric Synthetic Aperture Radar (InSAR) and additional remote sensing datasets such as aerial stereo photographs and LiDAR and tested in the Termini-Nerano landslides system (southern Apennines, Italy). The availability of such a wealth of materials allows also an unprecedented spatio-temporal reconstruction of the volume and the kinematic of the landslides system through which we could generate and validate the hazard map. Our analysis identifies fifteen slow-moving phenomena, traceable since 1955, whose total area amounts to 4.1 × 105 m2 and volume to ~1.4 × 106 m3. InSAR results prove that seven out of the fifteen slow-moving landslides are currently active and characterized by seasonal velocity patterns. These new insights on the dynamic of the landslides system have been selected as the main independent variables to train three MLAs (Artificial Neural Network, Generalized Boosting Model and Maximum Entropy) and derive the landslide hazard for the area. Finally, official population and buildings census data have been used to assess the landslide risk whose highest values are located in the crown area, south of Termini village, and nearby Nerano. This new methodology provides a different landslide risk scenario compared to the existing official documents for the study area and overall new insights on how to develop landslide risk management strategies worldwide based on a better understanding of slope processes thanks to the latest satellite technologies available.

Traditional applications of Interferometric Synthetic Aperture Radar (InSAR) data involved inverting an interferogram stack to determine the average displacement velocity. While this approach has useful applications in continuously deforming regions, new tools are needed for automatically and regularly identifying changes in the time series. Thanks to regular acquisitions across most of the world by the ESA Sentinel-1 satellites constellation, we are now in a position to explore opportunities for near-real time deformation monitoring. In this paper we present a statistical approach for detecting offsets and gradient changes in InSAR time series. Our key assumption is that 5 years of Sentinel-1 data is sufficient to calculate the population standard deviation of the detection variables. Our offset detector identifies statistically significant peaks in the first, second and third difference series. The gradient change detector identifies statistically significant movements in the second derivative series. We exploit the high spatial resolution of Sentinel-1 data and the spatial continuity of geophysical deformation signals to filter out false positive detections that arise due to signal noise. When combined with near-real time processing of InSAR data these detectors, particularly the gradient change, could be used to detect incipient ground deformation associated with geohazards such as landslides or volcanic eruptions.

The Campania region has been recurrently hit by severe landslides in volcanoclastic deposits. The city of Naples, and in particular the Camaldoli and Agnano hills (Phlegraean Fields), also suffered several landslide crises in weathered volcanoclastic rocks as a consequence of intense rainfalls or wildfires. To identify slope failures phenomena occurred in the winter season 2019–2020 an innovative procedure has been proposed. The purpose of this procedure is to highlight areas where major land cover changes occurred within our area of study, which can be potentially related to mass movements. The amplitude of spaceborne SAR images has been exploited for the change detection analysis and the output derived from the segmentation procedure has been compared with field observations. The amplitude-based method has been already applied in the detection of landslides, but never on the event with limited extensions, such as for this application. The achieved outcomes allowed the mapping of 62 new landslides that have been used to update the current landslide inventory database. This type of information is expected to help decision-makers with land planning and risk assessment.

Advances in differential interferometric synthetic aperture radar (DInSAR) processing algorithms, such as the Intermittent Small Baseline Subset (ISBAS), and increased data availability from SAR systems, such as Sentinel-1, provide the opportunity to increase the spatial and temporal density of ground deformation measurements. Such measurements, when combined with modelling, have the potential to make a significant cost-effective contribution to the progressive abandonment strategy of recently closed coalfields. Applications of DInSAR over coalfields have observed heave in coal measures rocks and temporal correlations between the rise of mine water and deformation time-series. The cessation of systematic dewatering can have a variety of detrimental impacts and knowledge of the time-scales (i.e. the rate of rebound) and structure of the mine system are crucial to the remediation strategy. Although mine plans and borehole measurements provide vital information in this regard, mine plans are often incomplete or inaccurate, whereas monitoring boreholes are spatially sparse. Consequently, groundwater can flow in unanticipated directions via goaf, mine shafts and roadways, making it difficult to determine where the impacts of rebound are likely to occur. In this study, ground deformation data obtained using ISBAS DInSAR on ENVISAT (2002–2009) and Sentinel-1 (2015–2019) data are used to develop a simple method to model groundwater rebound in abandoned coalfields. A forward analytical model based upon the principle of effective stress and mine water ponds is first implemented to estimate surface heave in response to changes in groundwater levels measured in monitoring boreholes. The forward model is then calibrated and validated using the ground deformation data. The DInSAR data were subsequently inverted to map the change in groundwater levels in greater detail across the coalfield and forecast surface discharges in order to support mitigation strategies.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. Sand movement is one of the main environmental hazards in Northern Sudan that threaten livelihood and rural communities. This paper investigates for the first time the use of the Synthetic Aperture Radar (SAR) offset tracking technique for detecting sand movement in Northern Sudan, and distinguishes the impact of the movement influencing factors: wind speed/direction, vegetation and topography. High-resolution images from the Sentinel-1 satellite were used for the generation of displacement maps. Three different dune fields with different characteristics were investigated for a study period between 4 June and 14 October 2017 (133 days). Dune field 1 is vegetated and near a built-up area, dune field 2 is in an open environment with sand dunes overlaying rocky substrate, and dune field 3 is located near mountains. The cumulative east displacement over the study period was 1.8 m, −1.1 m and 4.8 m for the three dune fields, respectively, while the cumulative north displacement was 0.7 m, 2.9 m and 4.2 m. Large movement is detected in the non-vegetated dune fields, with an average dune velocity of 0.18 m/d, while the vegetated dune field had a velocity of 0.09 m/d, which emphasizes the fact that vegetation is an effective stabiliser of dune movement. The pixel offset results showed a positive correlation between the wind speed/direction and the dune movement. In addition to vegetation, topography also played a major role in diverting the direction of the blown sand mainly near the edges to the mountains and the vegetation barriers. This technique showed high competency in monitoring the movement of sand dunes, in addition to identifying areas exposed to large sand drifting as a risk mapping technique.

According to the Emergency Events Database (https://public [...]

A new method, combining empirical modeling with time series Interferometric Synthetic Aperture Radar (InSAR) data, is proposed to provide an assessment of potential landslide volume and area. The method was developed to evaluate potential landslides in the Heitai river terrace of the Yellow River in central Gansu Province, China. The elevated terrace has a substantial loess cover and along the terrace edges many landslides have been triggered by gradually rising groundwater levels following continuous irrigation since 1968. These landslides can have significant impact on communities, affecting lives and livelihoods. Developing effective landslide risk management requires better understanding of potential landslide magnitude. Fifty mapped landslides were used to construct an empirical power-law relationship linking landslide area (AL) to volume (VL) (VL = 0.333 × AL1.399). InSAR-derived ground displacement ranges from −64 mm/y to 24 mm/y along line of sight (LOS). Further interpretation of patterns based on remote sensing (InSAR & optical image) and field survey enabled the identification of an additional 54 potential landslides (1.9 × 102 m2 ≤ AL ≤ 8.1 × 104 m2). In turn this enabled construction of a map that shows the magnitude of potential landslide activity. This research provides significant further scientific insights to inform landslide hazard and risk management, in a context of ongoing landscape evolution. It also provides further evidence that this methodology can be used to quantify the magnitude of potential landslides and thus contribute essential information towards landslide risk management.

We use satellite imagery to investigate the shoreline changes associated with volcanic activity in 2018&ndash, 2019 at Anak Krakatau, Indonesia, spanning a major lateral collapse and period of regrowth through explosive activity. The shoreline changes have been analyzed and validated through the adaptation of an existing methodology based on Sentinel-2 multispectral imagery and developed on Google Earth Engine. This work tests the results of this method in a highly dynamic volcanic environment and validates them with manually digitized shorelines. The analysis shows that the size of the Anak Krakatau Island increased from 2.84 km2 to 3.19 km2 during 15 May 2018&ndash, 1 November 2019 despite the loss of area in the 22 December 2018 lateral collapse. The lateral collapse reduced the island area to ~1.5 km2 but this was followed by a rapid increase in area in the first two months of 2019, reaching up to 3.27 km2. This was followed by a period of little change as volcanic activity declined and then by a net decrease from May 2019 to 1 November 2019 that resulted from erosion on the SW side of the island. This history of post-collapse eruptive regrowth and coastal erosion derived from the shoreline changes illuminates the potential for satellite-based automated shoreline mapping to provide databases for monitoring remote island volcanoes.

This study presents a new method, called the Radial Interpolation Method, to interpolate data characterized by an approximately radial pattern around a relatively constrained central zone, such as the ground deformation patterns shown in many active volcanic areas. The method enables the fast production of short-term deformation maps on the base of spatially sparse ground deformation measurements and can provide uncertainty quantification on the interpolated values, fundamental for hazard assessment purposes and deformation source reconstruction. The presented approach is not dependent on a priori assumptions about the geometry, location and physical properties of the source, except for the requirement of a locally radial pattern, i.e., allowing multiple centers of symmetry. We test the new method on a synthetic point source example, and then, we apply the method to selected time intervals of real geodetic data collected at the Campi Flegrei caldera during the last 39 years, including examples of leveling, Geodetic Precise Traversing measurements and Global Positioning System. The maps of horizontal displacement, calculated inland, show maximum values lying along a semicircular annular region with a radius of about 2–3 km in size. This semi-annular area is marked by mesoscale structures such as faults, sand dikes and fractures. The maps of vertical displacement describe a linear relation between the maximum vertical uplift measured and the volume variation. The multiplicative factor in the linear relation is about 0.3 × 106 m3/cm if we estimate the proportion of the ΔVthat is captured by the GPS network onland and we use this to estimate the full ΔV. In this case, the 95% confidence interval on K because of linear regression is ± 5%. Finally, we briefly discuss how the new method could be used for the production of short-term vent opening maps on the base of real-time geodetic measurements of the horizontal and vertical displacements.

The advent of new satellite and data processing techniques have meant that routine, operational and reliable surveys of land motion on a regional and national scale are now possible. In this paper, we apply a novel satellite remote sensing technique, the Intermittent Small Baseline Subset method, to data from a new satellite mission, Sentinel-1, and demonstrate that a wide area map of ground deformation can be generated that supports the regulation of a range of energy-related activities. The area for the demonstration is mainland Scotland (∼75,000 km2) and the land motion map required the processing of some 627 images acquired from March 2015 to April 2017. The results show that land motion is encountered almost everywhere across Scotland, dominated by subsidence over peatland areas. However, many other phenomena are also encountered including landslides and deformation associated with mining and civil engineering activities. Considering specifically Petroleum Exploration and Development Licence areas offered under the 14th Onshore Licensing Round in the UK, examples of the types of land motion are shown, including an example related to soil restoration by a wind farm. It is demonstrated that, in Scotland at least, almost all licence areas contain deformation of one form or another and, furthermore, the causes of that subsidence are dynamic and likely to be changing from year-to-year. Therefore, maps like this are likely to be of enormous use in a regulatory framework to scope out pre-existing problems in a licence area and to ensure that the correct monitoring framework is put in place once activities begin. They can also provide evidence of good practice and give assurance against litigation by third parties.

A large scale study of landslide processes was undertaken by coupling conventional geomorphological field surveys with aerial photographs along with an advanced Interferometric Synthetic Aperture Radar (InSAR) analysis of ground instability in north-western Sicily. COSMO-SkyMed satellite images for the period between 2008 and 2011 were processed using the Intermittent Small BAseline Subset (ISBAS) technique, recently developed at the Department of Civil Engineering of the University of Nottingham. The use of ISBAS allowed the derivation of ground surface displacements across non-urbanized areas, thus overcoming one of the main limitations of conventional interferometric techniques. ISBAS provides ground motion information not only for urban but also for rural, woodland, grassland and agricultural terrains, which cover > 60% of north-western Sicily, thereby improving by 40 times in some cases, the slope instability investigation capabilities of InSAR methods. ISBAS ground motion data enabled the updating of the landslide inventory for the areas of Piana degli Albanesi and Marineo (over 130 km2), which encompass a number of active, dormant and inactive landslides according to the pre-existing landslide inventory maps produced through aerial photo-interpretation and local field checks. An average of ∼ 7000 ISBAS pixels km− 2 allowed the detection of small displacements in regions difficult to access. In particular, 226 landslides – mainly slides, flows and creep and four badlands were identified, comprising a total area of 25.3 km2. When compared to the previous landslide inventory maps, 84 phenomena were confirmed, 67 new events were detected and 79 previously mapped events were re-assessed, modifying their typology, boundary and/or state of activity. Because the InSAR method used here is designed to measure slow rates of velocity and therefore may not detect fast-moving, events such as falls and topples, the results for Piana degli Albanesi and Marineo demonstrate the validity of this method to support land management, underlying the time and cost benefits of a combined approach using traditional monitoring procedures and satellite InSAR methods especially if slow-moving slope movements prevail.

Differential Interferometric Synthetic Aperture Radar (DInSAR) techniques have repeatedly proved to be an effective tool for built environments monitoring in areas affected by geological hazards. This paper describes how the Coherent Pixel Technique (CPT) approach has been successfully applied to assess the response of an unstable slope to the different phases of remedial works following a landslide event. The CPT technique was performed on 59 COSMO-SkyMed images obtained between May 2011 and August 2016 and centred on the Quercianella settlement (a small hamlet of Livorno municipality, Tuscany, Italy), where the reactivation of a dormant shallow slide had occurred in March 2011 and, hereafter, a geotechnical intervention, designed with the aim of mitigating the risks, has been conducted from August 2013, lasting thirteen months. The time series of CPT results show a deformation pattern with sudden accelerations (up to 21 mm in few months) corresponding to the beginning of the interventions, during which the area has been excavated to install a drainage well, followed by mild decelerations resulting from the stabilization of the area after the conclusion of the works. In particular, the integration of ground-based subsurface monitoring (inclinometers and piezometers) and DInSAR superficial data has provided consistent results for landslide characterization and helped defining the state of activity and the areal distribution of the sliding surface. Moreover, the performance of remedial works in the landslide-affected area has been observed, showing stabilization in the upper part of the hamlet and the ongoing movement in the lower part. The combined monitoring system also led the geotechnical company in charge of remedial works to design further stabilization works in order to preserve buildings and roads in the moving area. Therefore, the integration of remote sensing techniques and in situ instruments represents a timely and cost-efficient solution for intervention works monitoring, opening new perspectives on designing engineering solutions for the stabilization of unstable slopes.