NH6.7

In the last decades, satellite monitoring techniques allowed to discover non-catastrophic slope movements triggered by earthquake shaking and involving deep blind sliding surfaces of old paleo-landslides. Understanding the triggering and attenuation mechanisms of such mass movements is crucial to assess their hazard. On December 2018, the Etna volcano (southern Italy) began a very intense eruption, accompanied by a seismic swarm with magnitudes up to 4.9. Synthetic Aperture Radar data from Sentinel-1 and ALOS-2 identified some local displacements over a hilly area to the southwest of the Etna volcano, near Paternò village. We evaluated the contribution of seismically-induced surface instabilities to the observed ground displacement by employing a multidisciplinary analysis comprising geological, geotechnical and geomorphological data, together with analytical and dynamic modelling. The results of our study allowed us to identify the geometry and kinematics of a previously unknown paleo-landslide. A pseudostatic, limit-equilibrium back-analysis of the landslide mass highlighted that the displacements detected by InSAR data were caused by the undrained seismic instability of the landslide mass, which was dormant before the volcanic eruption, under the light-to-moderate seismic shacking of the December 26, Mw 4.9 earthquake. Such a new observation allowed to identify the geometry and kinematics of a previously unknown landslide mass and confirms that earthquakes have a cumulative effect on landslides that doesn't necessarily manifest as a failure but could evolve in a catastrophic collapse after several earthquakes. Such an aspect must be adequately investigated to identify unknown quiescent landslide bodies and to prevent the effects of their potential collapse during an earthquake.

How to cite: Albano, M., Saroli, M., Atzori, S., Moro, M., Tolomei, C., Bignami, C., and Stramondo, S.: Analysis of a large, seismically-induced mass movement after the December 2018, Etna volcano (southern Italy) seismic swarm, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-942, https://doi.org/10.5194/egusphere-egu21-942, 2021.

Abstract: Riverbank collapses frequently occur in the lower reaches of the Yellow River, China, which cause environmental changes around the riverbanks and result in a great loss of farmland. An analysis was carried out to understand the riverbanks of the Jiyang Reach via Google Earth. The results show that the three representative segments in the Jiyang Reach, the maximum annual-lateral displacement and average retreat area were 26.0 m/a and 1083.8 m² during the period 3/31/2016–5/10/2018, respectively. Several factors such as the soil properties, upstream river-control works, bridge piers, and channel bends may change the river flow direction and the scour intensity, thereby increasing the probability of downstream riverbank collapse, which are all causes of riverbanks collapse in the lower Yellow River. Field investigation and research data show that the lower reaches of the Yellow have serious bank-collapse disasters and their riverbanks are still in an unstable state.

keywords: Riverbank collapse; Yellow River; Channel evolution; Riverbank protection

How to cite: Gao, L., Xu, X., Zhao, Y., and Tarolli, P.: Assessment of avalanche hazards using remote sensing in the lower Yellow River, China, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-1000, https://doi.org/10.5194/egusphere-egu21-1000, 2021.

Abstract: To control the river regime in the wandering river channels is an important work of ecological protection and high-quality development in the Yellow River Basin. Using MIKE21, this study compared and analyzed the control effects of the spur dike group on the river regime under different oriented angles, layout methods, and dam types. The results show that: (1) A optimal oriented angle existed that can efficiently control the river regime. Among the dikes with three oriented angles designed in this study, the spur dam of 45° has the strongest effect blocking the flow, and the corresponding uniformity coefficient of the flow velocity CV reached the lowest value, 0.44, at this time. Under this condition, the flow-velocity distribution was more stable than that of other angles, dynamic pressure on the bank foundation was relatively small, and thus the groins could play a relatively effective influence on the protection of the river bend. (2) The effect on the river regime of a spur-dike group was more than the total amount of all single spur dikes. If only a single spur dike were arranged, the spur dike would keep the high-speed flow away from the concave bank and protect the riparian line with a length of about 80 m. In contrast, if the spur dikes worked as a group, a single spur dike would protect the riparian line with a length of about 100 m. (3) The diversion effect of the permeable groin in the lower Yellow River is the same as that of the solid groin with the same layout. Both the flow reduction rates of the permeable and solid groins are all close to 80%. It is concluded that the impermeable groins can be widely used in the lower Yellow River for it is able to achieve the expected control effect and relatively safe operation condition in virtue of permeability.

Keywords: wandering river channel; permeable groins; flow characteristics; MIKE21

How to cite: Peng, X., Xu, X., and Gao, L.: Effects of permeable groins on river regime in the lower Yellow River, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-1463, https://doi.org/10.5194/egusphere-egu21-1463, 2021.

InSAR images allow to detect the coseismic deformation, delimiting the epicentral area where the larger displacement has been concentrated. The main observations are: 1) the most deformed area in the ideal case is elliptical (for dip-slip faults) or quadrilobated (for strike-slip faults) and coincides with the surface projection of the volume coseismically mobilized in the hanging wall of thrusts and normal faults, or the crustal walls adjacent to strike-slip faults; 2) the dimension of the deformed area detected by InSAR scales with magnitude of earthquake and for M≥6 is always larger than 100 km, increasing to more than 550 km² for M≈6.5; 3) the seismic epicenter rarely coincide with the area of larger vertical shaking (either downward or upward); 4) the higher macroseismic intensity corresponds to the area of larger vertical displacement, apart from local site amplification effects; 5) outside this area, the vertical displacement is drastically lower, determining the strong attenuation of seismic waves and the decrease of the peak ground acceleration in the surrounding far field area, apart from local site amplifications; 6) the segment of the activated fault constrains the area where the vertical oscillations have been larger, allowing the contemporaneous maximum freedom degree of the crustal volume affected by horizontal maximum shaking, i.e., the near field or epicentral area; 7) therefore, the epicentral area and volume are active, i.e., they coseismically move and are contemporaneously crossed by seismic waves (active volume), whereas the surrounding far field area is mainly fixed and passively crossed by seismic waves (passive volume). Therefore, here we show how the InSAR images of areas affected by earthquakes represent the fingerprint of the epicentral area where the largest shaking has taken place during an earthquake. Seismic hazard assessments should rely on those data.

How to cite: Petricca, P., Bignami, C., and Doglioni, C.: The epicentral fingerprint of earthquakes, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-2938, https://doi.org/10.5194/egusphere-egu21-2938, 2021.

Coastal communities, land covers and intertidal habitats are vulnerable receptors of erosion, flooding or both in combination. This vulnerability is likely to increase with sea level rise and greater storminess over future decadal-scale time periods. The accurate, rapid and wide-scale determination of shoreline position, and its migration, is therefore imperative for future coastal risk adaptation and management. Developments in the spectral and temporal resolution and availability of multispectral satellite imagery opens new opportunities to rapidly and repeatedly monitor change in shoreline position to inform coastal risk management decisions. This presentation discusses the development and application of an automated tool, VEdge_Detector, to extract the coastal vegetation line from high spatial resolution (Planet's 3 – 5 m) remote sensing imagery, training a very deep convolutional neural network (Holistically-Nested Edge Detection) to predict sequential vegetation line locations on annual/decadal timescales. The VEdge_Detector outputs were compared with vegetation lines derived from ground-referenced positional measurements and manually digitised aerial photographs, revealing a mean distance error of <6 m (two image pixels) and > 84% producer accuracy at six out of the seven sites. Extracting vegetation lines from Planet imagery of the rapidly retreating cliffed coastline at Covehithe, Suffolk, UK identified a mean landward retreat rate >3 m a^-1 (2010 - 2020). Plausible vegetation lines were successfully retrieved from images of other global locations, which were not used to train the neural network; although significant areas of exposed rocky coastline proved to be less well recovered by VEdge_Detector. The method therefore promises the possibility of generalising to estimate retreat of sandy coastlines in otherwise data-poor areas, which lack ground-referenced measurements. Vegetation line outputs derived from VEdge_Detector are produced rapidly and efficiently compared to more traditional non-automated methods. These outputs also have the potential to inform upon a range of future coastal risk management decisions, including hazard and risk mapping considering future shoreline change.

How to cite: Rogers, M., Spencer, T., Bithell, M., and Brooks, S.: VEdge_Detector: Automated coastal vegetation edge detection using a convolutional neural network, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-3206, https://doi.org/10.5194/egusphere-egu21-3206, 2021.

The Tsaoling Landslide is one of the largest landslides in Taiwan caused by the Chi-Chi Earthquake in 1999. More than 130 million cubic meters of rocks and debris blocked the Chingshui Stream channel and formed a landslide dammed lake. In July 2004, Typhoon Mindulle completely filled the dam by the debris of the landslides initially situated on the higher upstream regions. Since then, the river channel in the region of the filled dam lake and the seismogenic Tsaoling landslide accumulation began to cut down by fluvial erosion and transportation, eventually forming multiple river terraces and deep valley. In 2009, extreme heavy rain fall hit the area again by the typhoon Morakot, causing deformation of the eastern flank of the landslide area and major river channel migration. However, relative environmental changes and geomorphical evolution in Tsaoling landslide area have received less attention. In recent years, the remote sensing technology improves rapidly, providing a wide range of image, essential and precise geoinformation. The Small unmanned aircraft system (sUAS) has been widely used in landslide monitoring and geomorphic change detection. On the basis of self-made drones, we have established a multi-temporal high-resolution DTMs, so as to access and to monitoring the post-landslide activities and topographic changes the Tsaoling area regularly and continuously. The result shows that, especially during the monsoon (spring rainy season) in June 2017, the small cliff of minor scarp on the main sliding surface has an important cliff line retreat. The maximum retreat distance exceeds 150 meters, and the volume of the landslide situated on the original sliding surface exceeds 1.5 million cubic meters. Over the next few years, the data set indicated that the topography of the area change continued. In this study, on the one hand, we are actively exploring new algorithms to minimize the relative error of the terrain in each period to accurately calculate the morphological changes in each period. On the other hand, the geomorphic changes indicate the landslide activity, and the characteristics of the river processes in the Tsaoling landslide area. Since 2016, through 8 multi-temporal UAS missions in Tsaoling area, the results indicate that the area is continues to deform and remain active. As a result, it is still worthwhile to monitor continuously.

How to cite: Chang, K.-J., Chang, H.-H., Hsieh, Y.-C., and Huang, M.-J.: Improve Drone Survey Hazard Mapping technology to decipher landslide activity and geomorphological evolution, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-5799, https://doi.org/10.5194/egusphere-egu21-5799, 2021.

Coastal flooding are natural processes that are both i) essential (providing nutrients to the coastal vegetation, habitats) and ii) hazardous (negatively impact human activities, livelihood, assets, livestock and so on). Climate changes have induced higher frequency of floods, rising sea levels, high amplitude tides and other climatic extremes at regional to global scales. The increasing intensity, duration of floods is proportionately increasing the risks associated with coastal human habitations. The regional risks are defined based on the physical, demographic, socio-economic vulnerability of the habitants. Sea level rise would further enhance the coastal inundations permanently breaching these productive, densely populated regions. This necessitates the need for spatially assessing the relative hazard, vulnerability and risks at regional scales to reduce/mitigate risks.

Indian subcontinent supports the second largest global population, with numerous megacities, towns and villages along the coast and mainland. This study's main objective is to quantify the risk associated with inundations caused by rising sea levels, tidal surge at the regional level. As a case study, Sagar Island located in the verge of Sundarbans, south of West Bengal is considered. Flood risk assessment in the island has been carried out using Multi-Criteria Decision Analysis (MCDA) framework based on 23 spatial parameters.

Results indicate, within a century (1922 – 2020), the island has lost most of its natural vegetation (mangroves - Sundarbans) (47% to 3%), with increasing cultivated (agriculture, horticulture) spaces (77.4 %) and built-up environs (8.2%). Sea level rise varies from 4.4 mm/year (South) to 5.25 mm/year (North) and in the last century has breached over 2824 hectares of mainland. The study's findings reveal 19.8% of horticulture and 33.3% of agriculture assets are highly exposed to natural hazards. 1.34% population are at relatively very high-risk levels, 17.81% at high-risk levels. The study's findings reveal the variable importance of socio-economic, demographic, topographic and proximity to public service, in defining the flood vulnerability and risk towards the habitants. The approach and findings of paves the way for planning authorities to prioritise risk mitigation strategies that are region-specific to reduce the impact of inundation due to natural hazards

Keywords: Sea level rise, Flood risk, MCDA, Vulnerability, flood hazard

How to cite: Shivamurthy, V. and Aithal, B.: Multi-Criteria Decision Analysis of Coastal Inundation at Regional scale, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-6963, https://doi.org/10.5194/egusphere-egu21-6963, 2021.

Remote sensing for natural hazard assessment and applications offers data on even vast areas, often difficult and dangerous to access. Today, satellite data providers such as PlanetLabs Inc. and the European Copernicus program provide a sub-weekly acquisition frequency of high resolution multispectral imagery. The availability of this high temporal data density suggests that the detection of short-term changes is possible; however, limitations of this data regarding qualitative, spatiotemporal reliability for the early warning of gravitational mass movements have not been analysed and extensively tested.

This study analyses the effective detection and monitoring potential of PlanetScope Ortho Tiles (3.125 m, daily revisit rate) and Sentinel-2 (10 m, 5-day revisit) satellite imagery between 2018 and 2020. These results are compared to high accuracy UAS orthoimages (0.16 m, 5 acquisitions from 2018-2020). The analysis is conducted based on digital image correlation (DIC) using COSI-Corr (Caltech), a well-established software and the newly developed IRIS (NHAZCA). The mass wasting processes in a steep, glacially-eroded, high alpine cirque, Sattelkar (2’130-2’730 m asl), Austria, are investigated. It is surrounded by a headwall of granitic gneiss with a cirque infill characterised by massive volumes of glacial and periglacial debris including rockfall deposits. Since 2003 the dynamics of these processes have been increased, and between 2012-2015 rates up to 30 m/a were observed.

Similar results are returned by the two software tools regarding hot-spot detection and signal-to-noise ratio; nonetheless IRIS results in an overall better detection, including a more delimitable ground motion area, with its iterative reference and secondary image combination. This analysis is supported by field investigations as well as clearly demarcated DIC-results from UAS imagery. Here, COSI-Corr shows limitations in the form of decorrelation and ambiguous velocity vectors due to high ground motion and surface changes for very high resolution of this input data. In contrast, IRIS performs better returning more coherent displacement rates. The results of both DIC tools for satellite images are affected by spatial resolution, data quality and imprecise image co-registration.

Knowledge of data potential and applicability is of high importance for a reliable and precise detection of gravitational mass movements. UAS data provides trustworthy, relative ground motion rates for moderate velocities and thus the possibility to draw conclusions regarding landslide processes. In contrast satellite data returns results which cannot always be clearly delimited due to spatial resolution, precision, and accuracy. Nevertheless, iterative calculations by IRIS improve the validity of the results.

How to cite: Hermle, D., Gaeta, M., Keuschnig, M., Mazzanti, P., and Krautblatter, M.: Multi-temporal analysis of optical remote sensing for time-series displacement of gravitational mass movements, Sattelkar, Obersulzbach Valley, Austria, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-8011, https://doi.org/10.5194/egusphere-egu21-8011, 2021.

Bushfire is one of the dangerous natural manmade hazards. It can cause great damges to the air quality, human health, environment and bio-diversity. In addition, forest fires may be a potential and signigicant source of polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans. In early 2020, Australia experienced serious bushfires with over an area of estimated 18.6 million hectares burned, over 5,900 buidlings (including 2, 779 homes) destroyed, and at least 34 people (including three fire fighters) and billion animals and some endangered species killed. Subsequently, air quality was degraded to hazardous levels. It was estimated that about 360 million tonnes of CO₂ was emitted as of 2 Jan. 2020 by NASA. Remote sensing data has been instrumental for the environmental monitoring in particular the bushfire. Many methods and algorithms have been proposed to detect the burned areas in the forest. However, it is challenging or even infeasible to routinely apply them by non-experts due to a chain of sophisticated schemes during their implementation. Here, we present a simple and effective method for mapping a burned area. The performances of different optical sensors and indices are conducted. Sentinel-2 MSI and Landsat 8 data are ultilized for the comparison of burned forest by analyzing different indices (including NDVI, NDBR and newly development index Nomarlized Difference Laten Heat Index (NDLI)). The forest damages are estimated over the Katoombar, Austrialia and the burning severity map is generated and classified into eight levels (none, high regrowth, lowregrowth, unburned, low severity, moderate low severity, moderate high severity, and high severity). The comparision in results from Sentinel-2 MSI data and Landsat image is performed and presented.

How to cite: Nguyen, K.-A., Liou, Y.-A., and Ho, L.-T.: A view of recent forest fire in Australia by satellite derived indices, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-8562, https://doi.org/10.5194/egusphere-egu21-8562, 2021.

It is well known that groundwater overexploitation can generate land subsidence due to the compaction of compressible aquitards. Mexico City's soils are an important example of highly compressible lake sediments in compaction due to groundwater extraction that have significantly damaged the urban and commercial building structures. Previous studies indicate that there is annual subsidence of 15 to 25 cm in the Mexico City International Airport, 10 cm in downtown, and between 10 to 15 cm in the Southeast Mexico City area. Soil fracturing is an indicator of differential subsidence that has damaged buildings and infrastructure, including hydraulic pipes, sidewalks, and pavements. For this reason, it is necessary to carry out specific studies related to topographic deformation. This talk presents a characterization of the terrain changes over time and a zoning map for Mexico City subsidence susceptibility. To this end, free access elevation models generated from 2000 to 2018 by different sensors and methodologies were compared. The resulting model is validated by mapping information from active GPS stations, whose data is also freely available. Besides, a spatial comparison of land subsidence areas and sites previously identified as flooding and aquifer overexploitation areas is presented. The results will serve as a basis for future monitoring to be carried out in the area with high-resolution tools.

How to cite: Vidal, G., Nieto Butrón, J., Hernández Hernández, M. A., Herrera Zamarrón, G., Cabral Cano, E., Yépez Rincón, F. D., and Ramírez Serrato, N. L.: Morphometric analysis of the terrain over time to characterize subsidence. Case study: Mexico City, Mexico., EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-10282, https://doi.org/10.5194/egusphere-egu21-10282, 2021.

Wildfires are occurring throughout the world, causing immense damage to forest resources, flora, and fauna. The Fire Danger indices are used as a tool for the decision-makers to issue warnings to the public, based on the level of fire danger classes for implementing mitigation measures to control wildfires. In this study, a Wildfire danger index (WDI) is developed from static and dynamic factors, which are derived from satellite datasets. The Static fire danger Index (SFDI) is generated using MODIS Land cover type (MCD12Q1), Shuttle Radar Topography Mission (SRTM) DEM, and Open Street Map datasets. The Random Forest algorithm is used to generate SFDI from the parameters LULC map, slope map, aspect map, and elevation maps based on the historical MODIS active thermal anomaly product (MCD14). Dynamic Fire Danger Index (DFDI) is developed from the MODIS Terra datasets such as Land Surface Temperature (MOD11A2) and surface reflectance (MOD09A1) datasets.  The DFDI is developed from four parameters viz. Land surface temperature, Visible Atmospherically Resistant Index (VARI) and Normalized Multiband Drought Index (NMDI), and Normalized Difference Infrared Index – B6 (NDIIB6). Finally, the wildfire danger index is calculated by integrating SFDI and DFDI and found that the accuracy is more than 80% during the 2018-19 fire season. Therefore, the WDI can be useful for disseminating daily fire danger maps on near real-time basis using the MODIS TERRA Near Real-Time datasets so that the fire officials to take necessary actions to control the spread of wildfires.

How to cite: Babu, S., Visser, V., Moncrieff, G., Slingsby, J., and Altwegg, R.: Developing a Wildfire danger index based on the satellite-derived parameters, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-10320, https://doi.org/10.5194/egusphere-egu21-10320, 2021.

The Planpincieux Glacier lies in the Italian side of the Grandes Jorasses massif (Mont Blanc area), toward the Ferret Valley, in the Courmayeur municipality. This is a highly touristic area, visited every year by tens of thousands of people.

In summers 2019 and 2020, large portions of the Montitaz Lobe of the glacier (estimated volumes of 250000 m³ and 500000 m³ respectively) became unstable and menaced the Planpincieux village. According to runout simulations, such volumes could have reached and damaged a small bridge, buildings or the main valley road, depending on the volume involved in the collapse. Therefore, robust volume estimation was required for the realisation of effective safety plans.

To this aim, a helicopter-borne ground-penetrating-radar (GPR) survey was conducted in July 2020 with the novel dual polarization AIRETH system. Such a survey provided the ice thickness (20-60 ±10 m) of the unstable portion  and the bedrock topography along transects.

Besides, multiple helicopter and drone photogrammetric surveys were acquired since 2017, which provided the digital terrain model (DTM) and the orthophotos of the glacier using structure from motion (SfM) technique.

Merging GPR and SfM allowed at reconstructing the evolution of the glacier shrinkage in the period where DTMs were available. Moreover, it was possible to assess the correspondence of several bedrock discontinuities with large recurrent fractures.

Even though it is commonly acknowledged that the bedrock topography influences the glacier morphology, their correspondence has been rarely demonstrated in an Alpine glacier.

Since the fractures provoked by the bedrock discontinuities might destabilise the underlying glacier portion, the knowledge of the actual position of such fractures can help in the quantitative evaluation of the glacier instability. This can have a strong impact in the potential glacier-related risk assessment and management.

How to cite: Dematteis, N., Trolio, F., and Giordan, D.: Assessing the influence of bedrock discontinuities on glacier fractures using ground-penetrating radar and structure from motion, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-10685, https://doi.org/10.5194/egusphere-egu21-10685, 2021.

Wildfires have been an integral part of the Mediterranean ecosystem. Moreover, the Intergovernmental Panel on Climate Change (IPCC) Fifth Assessment Report emphasizes that the Mediterranean basin is expected to be drier by the end of the 21st century, while future warming will possibly be higher than the global mean. Therefore, outbreaks of wildfires are expected to increase. One of the most important factors for wildfire behavior apart from the meteorological conditions, is fuel types. In this study, a detailed fuel type mapping in a case study area was addressed. To accomplish this goal, an object-based image analysis (OBIA) approach was implemented using the open-source Orfeo toolbox. The freely available Sentinel-2A satellite images were processed in combination with auxiliary European and National scale GIS data. The classification results demonstrate a high-quality Land Cover map with 84% of overall accuracy. The classified land cover polygons were associated with high-resolution tree cover density data derived from Copernicus Land Monitoring Service. This coupling led to the synthesis of the fuel type map. To this end, this approach can fulfill the efficient mapping of fuel types for operational purposes. This research has been co‐financed by the European Regional Development Fund of the European Union and Greek national funds through the Operational Program Competitiveness, Entrepreneurship and Innovation, under the call RESEARCH –CREATE –INNOVATE  (project code:T2EDK-01967)

How to cite: Karystinakis, K., Alexandridis, V., Stefanidis, S., and Kalantzi, G.: Fuel type mapping in a typical Mediterranean ecosystem using object-based image analysis of Sentinel 2 imagery and auxiliary GIS data, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-11355, https://doi.org/10.5194/egusphere-egu21-11355, 2021.

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.

How to cite: Novellino, A., Hussain, E., Jordan, C., and Bateson, L.: Offline-Online Change Detection for Sentinel-1 InSAR Time Series, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-11673, https://doi.org/10.5194/egusphere-egu21-11673, 2021.

Landslide identification is the fundamental step to reduce the potential damaging effects of landslide activities. A variety of techniques and approaches has been developed to detect landslides. Conventional landslide identification is a complex and laborious task due to a large amount of the field work and materials that have to be investigated. Additionally, the conventional geomorphological mapping mainly provides a subjective representation of landscape complexities at different scales. Sometimes, in certain conditions, such as densely-vegetated terrain, conventional landslide mapping is ineffective or even impossible.

Therefore, innovative methods that allow for the reduction of subjectivism, time, and effort have increasingly become the subject of interest in landslide research. These methods mainly focus on semi-automated or automatic landslide mapping and include analysis of remote sensing data, such as optical images, Digital Elevation Models (DEMs) derived by Light Detection and Ranging etc. Among them, the pixel-based approach (PBA) and the object-based image analysis (OBIA) methods can be distinguished, for which supervised classification methods are usually utilized.

The accuracy of supervised classification methods strongly corresponds to the training samples - its quality and amount. Supervised classification methods require the collection of training as well as testing data to generate and assess the accuracy of the classification results. It is a challenging task, especially in forested areas, to capture ground truths of the good quality to train the classifier and to identify landslides. Considering this, we decided to investigate the following research question: What is the appropriate training–testing dataset split ratio in supervised classification to detect landslides in a testing area based on DEMs? Since PBA and OBIA approaches are nowadays widely utilized, we investigated this issue for both methods. The Random Forest classifier was implemented for both methods. The experiments were performed in Poland in the Outer Carpathians.

Accuracy measures calculated for the region growing validation indicated that the training area should be similarly large to the testing area in DEM-based automatic landslide detection. Additionally, we found that the OBIA approach performs slightly better than PBA when the quantity of training samples is lower. Besides this, we also attempted to increase the detection performance and to generate final landslide inventory. For this purpose, the intersection of the OBIA and PBA results together with median filtering and the removal of small elongated objects were carried out. We achieved the Overall Accuracy of 80% and F1 Score of 0.50.

How to cite: Pawluszek-Filipiak, K. and Borkowski, A.: Automatic landslide detection using the Random Forest classification - the importance of the train-test split ratio, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-12046, https://doi.org/10.5194/egusphere-egu21-12046, 2021.

The MOMPA project (MOnitorización de Movimientos del terreno y Protocolo de Actuación - MOnitoring of ground Movements and Action Protocol) has been 65% co-financed by the European Regional Development Fund through the Interreg V-A Spain-France-Andorra programme (POCTEFA 2014-2020). POCTEFA aims to reinforce the economic and social integration of the French–Spanish–Andorran border. The study area of the project is in the Eastern Pyrenees, covering the whole Principality of Andorra, the Spanish areas of Alt Urgell and Cerdanya (Catalonia) and the French areas of Cerdanya-Capcir and Conflent (Occitanie). The aim of the Project is to provide a useful technical-operational tool for risk prevention and management, at a cross-border level, based on satellite DInSAR technique monitoring of ground movements. The tool includes two main elements: the assessment of the risk associated with active phenomena that affect structures and infrastructures; and the integration of the technique in an action protocol for Civil Protections. The results will be transferred to Civil Protections (associated partners of the project) and other organizations, such as local and regional Public Authorities.

The study area presents one main critical issue: it is not an easy area for what concerns the radar response. This means that the obtainable results in terms of displacement map (velocity map and time series of deformation), which is the main input of the project, can be strongly limited. A second issue is the variability of the available data (e.g. landslide inventory, geology, DEM) between Andorra, Spain, and France. In General, landslides inventories are not complete or exhaustive and do not cover areas far from human structures.

The project will face the risk assessment starting from the interregional scale displacement map (covering around 15,000 km²) and the extracted Active Deformation Areas (ADA), as inputs to then select movements with potential risk where focus the analysis at a local scale, based on traditional method (basically photointerpretation and field work). Both the medium-resolution, free data, acquired by Sentinel-1 and the high-resolution data acquired by COSMO-SkyMed will be used, the results will be compared and evaluated.

Moreover, the project focuses his attention on the specific case of “la Portalada” (in Andorra). This is a huge landslide that occurred on August 2019. Today, there is a slow movement up slope that could affect a main road located in the bottom of the valley . Because of the high interest for the local authorities to monitor and characterize the current movement of the slope located upper to the landslide scar eight passive and one active corner reflectors have been installed along the steep forested slope. The data obtained will be integrated in the prevention risk protocol.

The project started the 1^st of December 2019 and will finish in May 2022.  The aim of this work is to present the project and the first results achieved through satellite interferometry.

How to cite: Barra, A., Marturià, J., Copons, R., Gasc, M., Fabregat, I., Buxó, P., Dufour, N., Pigeot, L., Colell, X., Trapero, L., and Crosetto, M.: MOMPA Project: interregional DInSAR monitoring and action protocol in the Eastern Pyrenees, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-12161, https://doi.org/10.5194/egusphere-egu21-12161, 2021.

Over the last few years, wildfires have become more severe and destructive, having extreme consequences on local and global ecosystems. Fire detection and accurate monitoring of risk areas is becoming increasingly important. Satellite remote sensing offers unique opportunities for mapping, monitoring, and analysing the evolution of wildfires, providing helpful contributions to counteract dangerous situations.

Among the different remote sensing technologies, hyper-spectral (HS) imagery presents nonpareil features in support to fire detection. In this study, HS images from the Italian satellite PRISMA (PRecursore IperSpettrale della Missione Applicativa) will be used. The PRISMA satellite, launched on 22 March 2019, holds a hyperspectral and panchromatic  payload which is able to acquire images with a worldwide coverage. The hyperspectral camera works in the spectral range of 0.4–2.5 µm, with 66 and 173 channels in the VNIR (Visible and Near InfraRed) and SWIR (Short-Wave InfraRed) regions, respectively. The average spectral resolution is less than 10 nm on the entire range with an accuracy of ±0.1 nm, while the ground sampling distance of PRISMA images is about 5 m and 30 m for panchromatic and hyperspectral camera, respectively.

This work will investigate how PRISMA HS images can be used to support fire detection and related crisis management. To this aim, deep learning methodologies will be investigated, as 1D convolutional neural networks to perform spectral analysis of the data or 3D convolutional neural networks to perform spatial and spectral analyses at the same time. Semantic segmentation of input HS data will be discussed, where an output image with metadata will be associated to each pixels of the input image. The overall goal of this work is to highlight how PRISMA hyperspectral data can contribute to remote sensing and Earth-observation data analysis with regard to natural hazard and risk studies focusing specially on wildfires, also considering the benefits with respect to standard multi-spectral imagery or previous hyperspectral sensors such as Hyperion.

The contributions of this work to the state of the art are the following:

• Demonstrating the advantages of using PRISMA HS data over using multi-spectral data.
• Discussing the potentialities of deep learning methodologies based on 1D and 3D convolutional neural networks to catch spectral (and spatial for the 3D case) dependencies, which is crucial when dealing with HS images.
• Discussing the possibility and benefit to integrate HS-based approach in future monitoring systems in case of wildfire alerts and disasters.
• Discussing the opportunity to design and develop future missions for HS remote sensing specifically dedicated for fire detection with on-board analysis.

To conclude, this work will raise awareness in the potentialities of using PRISMA HS data for disasters monitoring with specialized focus on wildfires.

How to cite: Spiller, D., Ansalone, L., Longépé, N., Wheeler, J., and Mathieu, P. P.: Wildfire detection and monitoring by using PRISMA hyperspectral data and convolutional neural networks, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-12330, https://doi.org/10.5194/egusphere-egu21-12330, 2021.

High Mountain Asia (HMA) has the most complex terrain with active hydrologic and geomorphologic processes. Climate change has expedited glacial melt and altered monsoon rain intensity. This has increased flood vulnerability across the region. There have been a few initiatives to measure the vulnerability locally. However, to identify hotspots of flood risk across the region, investigation of the entire HMA region is necessary. Unfortunately, in ungauged basins, the use of traditional floodplain mapping techniques is prevented by the lack of the extensive data required. The present work aims to provide a remote sensing-based flood-risk assessment model that maps and quantifies susceptibility in flood-prone areas. We developed a procedure for floodplain delineation based on high-resolution terrain data and a geomorphic classifier, coupled with satellite-derived extreme rainfall quantiles, and records of past flood events. For this work, we used the unique 8-meter Digital Elevation Models (DEMs) for HMA that are available at the NASA National Snow and Ice Data Center Distributed Active Archive Center (NSIDC DAAC). The geomorphic classifier is based on the Hydraulic Scaling Function automatically derived from the DEM, which is used to normalize topography according to the ratio between the local elevations along the drainage network and the riverbanks.  We assess the flood risk hot spots for a specific year based on the spatial distribution of flood losses, drainage density, flood-prone areas, and rainfall. This local flood-risk assessment framework, gradually applied across the entire HMA domain, will increase the awareness of flood risk, towards improved measures for flood risk reduction.

How to cite: Khanam, M., Sofia, G., Nikolopoulos, E. I., and Anagnostou, E. N.: Identification of hotspots of flood risk in High Mountain Asia region based on geomorphology and climate data, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-13783, https://doi.org/10.5194/egusphere-egu21-13783, 2021.

Coal fires, land subsidence, roof collapse, and other life-threatening risks are a predictable phenomenon for the mineworkers and the neighbourhood population in coalfields. Jharia Coalfields in India are suffered heavily from land subsidence and coal fires for over a century. In addition to the loss of precious coal reserves, this has led to severe damage to the environment, livelihood, transportation, and precious lives.

Such incidents highlight the dire need for a well-defined methodology for risk analysis for the coalfield. In this study, we regenerated a Land Use Land Cover map prepared using Indian Remote Sensing satellite imagery and ground survey. Persistent Scatterer Interferometry analysis using Sentinel -1 images was carried out to study the land subsidence phenomenon between Nov 2018 and Apr 2019. For the same study period, coal fire zones were identified with Landsat – 8 thermal band imagery. Integration of coal fire maps, subsidence velocity maps, and land use maps was further implemented in a geographical information background environment to extract the high-risk zones. These high-risk areas include residential areas, railways, and mining sites, requiring immediate attention.

The results show that the coal mines are affected by subsidence of up to 20 cm/yr and a temperature anomaly of nearly 20^oC is noticed. A high-risk zone of almost 18 sq. km. was demarcated with Kusunda, Gaslitand, and West Mudidih collieries being the most critically affected zones in the Coal mines. The study demonstrates the potential to combine data from multiple satellite sensors to build a safer ecosystem around the coal mines.  

How to cite: Karanam, V., Garg, S., Motagh, M., and Jain, K.: The Risk of Coal Fires And Land Subsidence in Jharia Coalfields, India, Analysed Using Remote Sensing Techniques, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-14419, https://doi.org/10.5194/egusphere-egu21-14419, 2021.

Regional-scale landslide deformation can be measured using satellite-based synthetic aperture radar interferometry (InSAR). Our study focuses on the quantification of displacements of slow-moving landslides that impact a hydropower dam and reservoir in the tropical Ecuadorian Andes. We constructed ground surface deformation time series using data from the Copernicus Sentinel-1 A/B satellites between 2016 and 2020. We developed a new approach to automatically detect the onset of accelerations and/or decelerations within each active landslide. Our approach approximates the movement of a pixel as a piecewise linear function. Multiple linear segments are fitted to the cumulative deformation time series of each pixel. Each linear segment represents a constant movement. The point where one linear segment is connected to another linear segment represents the time when the pixel’s rate of movement has changed from one value to another value and is referred to as a breakpoint. As such, the breakpoints represent moments of acceleration or deceleration. Three criteria are used to determine the number of breakpoints: the timing and uncertainty of the breakpoints, the confidence intervals of the fitted segments’ slopes, and the Akaike Information Criterion (AIC). The suitable number of breakpoints for each pixel (i.e., the number of accelerations or decelerations) is determined by finding the largest number of breakpoints that complies with the three listed criteria. The application of this approach to landslides results in a wealth of information on the surface displacement of a slope and an objective way to identify changes in displacement rates. The displacement rates, their spatial variation, and the timing of acceleration and deceleration can further be used to study the physical behavior of a slow-moving slope or for (regional) hazard assessment linking the onset of change in displacement rate to causal and triggering factors.

How to cite: Urgilez Vinueza, A., Handwerger, A., Bakker, M., and Bogaard, T.: A new methodology to detect changes in displacement rates of slow-moving landslides using InSAR time series, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-14447, https://doi.org/10.5194/egusphere-egu21-14447, 2021.

Gathering systematic information on the effects of extreme weather events (e.g., flooded areas, shallow landslide and debris flow activations, windthrows) is a fundamental prerequisite for local authorities to put into practice management strategies and establishing early-intervention priorities. The collection of these data is a resource-demanding task requiring huge personnel effort and financial means. Furthermore, events occurring in remote areas with a low chance of intersecting human infrastructure, are rarely detected and mapped accurately, thus leading to incorrect assumptions in relation to both extreme events spatial distribution and especially to the real occurrence probability. The present work aims at tackling some of the above-mentioned issues by providing a framework for obtaining the automatic identification of severe weather events that may have caused important erosional processes or vegetation damage, combined with a quick and preliminary change detection mapping over the identified areas.
The proposed approach leverages the free availability of both high-resolution global scale radar rainfall products and Sentinel-2 multispectral images to identify the areas to be analyzed and to carry out change detection algorithms, respectively. Radar rainfall data are analyzed and areas where high intensity rainfall and/or very important cumulative precipitation has occurred are used as a mask for restricting the subsequent analysis, which, in turn, is based on a multispectral change detection algorithm.
The testing phase of the proposed methodology provided encouraging results:  applications to selected mountain catchments hit by the VAIA storm in northeastern Italy (October 2018) were capable of identifying flooded areas, debris-flow and shallow landslide activations and windthrows with good accuracy and with the ability to distinguish between erosional processes and windthrows.
The described approach can serve as a preliminary step toward detailed post-event surveys, but also as a preliminary “quick and dirty” mapping framework for local authorities especially when resources for ad hoc field surveys are not available.
Such a systematic potential change identification, in combination with regular expert-driven validation, can finally pave the way for a process of self-improvement in detection and classification accuracy: if classified changes are validated, machine-learning algorithms can be trained to learn and improve performance not only in change detection accuracy but also in single-scene classification. 
Future improvements of the described procedure could be finally devised for allowing a continuous operational activity and for maintaining an open-source software implementation.

How to cite: Crema, S., Marchi, L., Borga, M., and Cavalli, M.: Thunderslide - from rainfall to preliminary landslide mapping: implementing an open data-oriented framework for landscape management authorities, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-14666, https://doi.org/10.5194/egusphere-egu21-14666, 2021.

Land surface elevation changes can cause damage to infrastructure and other resources; thus, its monitoring is crucial for the safety and economics of the city. Long-term excessive extraction of underground water is one of the factors that causes ground to sink. Faridabad, the industrial hub of Haryana, a state in north India is staring a severe water crisis in the near future and has already been declared as a dark zone with regard to groundwater resources. At many places, the underground water table has dropped more than 150m. The plummeting groundwater levels and the geology of this region make it prone to subsidence.

Continuous monitoring of land surface elevations using traditional surveying techniques can be time-consuming and labor-intensive. Several studies have shown the potential of remote sensing techniques in monitoring the changes in topography to an mm level accuracy. In this study, we used the elevation change map (derived using 200+ sentinel -1 images), subsidence gradient, groundwater in-situ data, population, population density, land cover, and lithology. These information were then processed and analyzed in a geographical information system to perform a hazard vulnerability and risk assessment. The final risk map was classified into three different classes viz high, medium, and low risk pertaining to ground movement.

The results indicate that the high-risk zone covers an area of more than 2.5 square kilometers. New Industrial Town (NIT) in Faridabad with an estimated population of more than 1.5 million, is found to be at high risk of ground movement. Groundwater levels in this area are currently depleting by more than 5m/year. Some other areas which are under high risk are the Dabua colony, Sanjay Gandhi Memorial Nagar, and Gandhi colony. All these regions have a high population density and demand urgent government attention.

How to cite: Garg, S., Karanam, V., Motagh, M., and Jayaluxmi, I.: Risk of Ground Movement in Faridabad, India – Investigated using Remote Sensing and In-Situ Data, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-15694, https://doi.org/10.5194/egusphere-egu21-15694, 2021.

Climate change has increased the frequency of flood events globally. Floods cause massive loss of life and cause the expenditure of billions of dollars. While it is important to curb floods caused by anthropogenic factors in the first place, it is equally important to reduce the impact in the aftermath of floods. The extent of past flood events is crucial for developing disaster management plans and flood hazard modelling. Due to the lack of capacity and availability of the funds with local officials, many past disasters remain unmapped and the information is just limited to total life loss and damage estimates.

Satellite data has been widely hailed as an alternative to drone and aerial surveys. And recent advances in open Earth Observation (EO) data availability, for instance, the Sentinel-1 SAR data by the European Space Agency (ESA), and cloud processing platforms such as the Google Earth Engine (GEE) have opened unprecedented opportunities for using EO data for hazard and disaster response efforts. Recent literature in the field of EO is witnessing an increasing number of the Sentinel-1 and GEE combination for flood mapping.

In the present work, we demonstrate the utility of a recently developed tool, the Global Flood Mapper (GFM), which is an open GEE application for rapid mapping of flood inundation extent using Sentinel-1 data. GFM uses a pre-flood time period to analyse numerous Sentinel-1 scenes of the same study area, this accounts for seasonal variation and has lesser noise as compared to other methods that use just one pre-flood scene. We map a couple of flood events across the globe to demonstrate the scalability and ease of using GFM. In addition, we analyse the flood hazard vulnerability of the state of Bihar in India using flood extent for the year 2018, 2019 and 2020 by delineating frequently flooding areas. This showcases yet another crucial utility of the GFM tool. GFM can support the flood extent mapping of the past events in addition to the rapid flood mapping of the current events, that could aid researchers and disaster managers for better flood preparedness and response. 

We access GFM through the link available on this public repository: https://github.com/PratyushTripathy/global_flood_mapper

How to cite: Tripathy, P. and Malladi, T.: Global Flood Mapper: Democratising open EO resources for flood mapping, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-16194, https://doi.org/10.5194/egusphere-egu21-16194, 2021.