EO data have proven to be crucial for hazard, vulnerability, and risk mapping from small to large regions around the globe, during the occurrence of disasters and the pre/post hazard phases. In this framework, the Committee on Earth Observation Satellites (CEOS) has been working for several years on disaster management related to natural hazards (e.g., volcanic, seismic, landslides and floods), including pilots, demonstrators, recovery observatory concepts, Geohazard Supersites, and Natural Laboratory (GSNL) initiatives and multi-hazard management projects. Moreover, European Ground Motion Service (EGMS) has significantly improved the ability to monitor and analyse geohazards using Interferometric Synthetic Apeture Radar data. Data are available since mid-2022 from the Copernicus Land Monitoring Service (CLMS) under the responsibility of the European Environment Agency (EEA).
The session is dedicated to multidisciplinary contributions focused on the demonstration of the benefit of the use of EO for assessment of natural hazards and risk management.
The contributions may include:
- Innovative applications of EO data for rapid hazard/risk assessment
- Development of tools for assessment and validation of hazard/risk models
- Use of EGMS data/products to monitor and investigate different kinds of geohazards and their impact on both environment and infrastructure
The use of different types of remote sensing data (e.g. thermal, visual, radar, laser, and/or the fusion of these) or platforms (e.g. space-borne, airborne, UAS, drone, etc.) is highly recommended, with an evaluation of their respective pros and cons focusing also on future opportunities (e.g. new sensors or algorithms).
Early-stage researchers are strongly encouraged to present their research. Contributions demonstrating innovative, cross-disciplinary approaches and case studies with practical implications are particularly welcome. In addition, we invite contributions from international collaborations, such as CEOS, GSNL and GEO.
Orals: Fri, 8 May, 08:30–15:30 | Room 1.31/32
Algerian oasis ecosystems are among the few resilient ecosystems, that manage to maintain a delicate balance between resource availability and the needs of societies. They constitute a unique agroforestry system, providing essential services to the ecosystem and local communities, such as provisioning, regulation, and cultural services. Following climate change and repeated droughts, these ecosystems are undergoing significant environmental modifications, marked by changes in the nature and behavior of vegetation cover, a fundamental element of ecological stability. Therefore, in order to maintain the ecological balance of oasis ecosystems and combat their degradation, it is necessary to understand the relationship between climate change and its impacts on the transformation of these oasis ecosystems. For this study, the pre-Saharan oasis zone of ASLA (south of Naâma), in Algeria, was chosen because of the significant environmental changes it has undergone following a prolonged decrease in rainfall. Our study is based on exploring spatio-temporal variations and identifying changes in land cover. To do this, we extract classification categories representing soil conditions, which we subdivide into classical classification categories corresponding to plant groups. To this end, we classified land cover by combining several spectral indices, calculable from satellite data for each spectral band, to create multiband input data for a supervised classification approach based on a support vector method (SVM). This method was applied to Landsat 8 OLI images with 30-meter resolution, combined with images from the Algerian satellite Alsat-2B with 2.5-meter resolution for panchromatic images. Archive images from 2008, 2013, 2018, and 2023, at five-year intervals, were used to detect changes. We then studied the relationship between variations in land surface parameters and changes in land surface temperature (LST) and normalized difference vegetation index (NDVI) before and after drought. Using GIS, we integrated climatic parameters (precipitation and land surface temperature) combined with land cover and NDVI results. Then followed expert recommendations to determine the weights to be assigned to each parameter in the model. This method allowed us to clarify the situation regarding the degradation of the oasis ecosystem, to classify the study area according to its degree of vulnerability, and to determine the spatiotemporal changes that occurred over the 15-year period.
How to cite: Zegrar, A., Bentekhici, N., and Benshila, N.: Spatial and temporal trends in oasis ecosystems and their response to prolonged drought in an arid zone of Algeria, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-375, https://doi.org/10.5194/egusphere-egu26-375, 2026.
The Himalayan region is becoming increasingly susceptible to landslides due to unplanned construction and rapid urbanization, particularly in Uttarakhand and Himachal Pradesh, India. Road widening and infrastructure development on steep, unstable slopes have triggered land subsidence in Joshimath (Awasthi et al., 2024) and in the Char Dham Highway landslide. Landslides are also occurring in Himachal Pradesh's Solan district. In order to better understand how rapid changes in land-use and land-cover (LULC) are increasing the risk of landslides in this vulnerable and urbanizing district of Himachal Pradesh, this study proposes an integrated Remote Sensing (RS) and Earth Observation (EO). In Google Earth Engine (Gorelick et al., 2017), multi-temporal Sentinel-2 images from 2019 to October 2025 were analysed using cloud-masking and spectral indices (NDVI, NDBI, and NDWI) to precisely identify land cover types. Change detection analysis using this processed dataset showed that built-up areas increased by 11% between 2019 and 2024 and a remarkable 16% growth between 2024 and October 2025, indicating increased urbanisation during the most recent period (2024-2025). The analysis identifies a transition in urbanization areas. In the LULC change map, we observed that Baddi, Nalagarh and Barotiwala constitute established urban centers, however, the 2024-2025 duration shows maximum expansion within the Chamba (northeast), Arki (central-east), and Kasauli (southeast) regions. The northeastern and southeastern regions of the Solan district are emerging as the new urban expansion zones. Our ongoing research focuses on developing a Random Forest-based landslide susceptibility model that combines multi-sensor Earth Observation data with these LULC dynamics through an optical–SAR fused framework. In order to develop a framework for identifying high-risk areas, we are investigating Sentinel-1 SAR images (the GRD products) to measure coherence and backscatter change and combining with topographic parameters derived from SRTM (slope, curvature, aspect) (Sharma et al., 2024). To improve this optical-SAR fused model accuracy, additional geological data from the Geological Survey of India and rainfall data from CHIRPS (Climate Hazards Group Infrared Precipitation with Stations) will be incorporated. This integrated method allows for a quantitative assessment of susceptibility in this vulnerable Himalayan terrain by correlating land-use transitions with slope instability indicators.
How to cite: Dhayal, P., Banerjee, S., and Raman, B.: Mapping urban expansion and landslide susceptibility over Solan District of Himachal Pradesh, India, using Random Forest approach integrating LULC dynamics and Sentinel-1 data, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-1023, https://doi.org/10.5194/egusphere-egu26-1023, 2026.
The Committee on Earth Observation Satellites (CEOS) Working Group on Disasters (WGD) has coordinated multiple initiatives to enhance volcano disaster risk management through improved access to satellite data. Historically, access to high-resolution synthetic aperture radar (SAR) and optical imagery—critical for monitoring volcanic activity—has been restricted by costs or limited research-focused allocations. To overcome these limitations, CEOS-WGD launched the Volcano Pilot Project (2014–2017), which demonstrated the feasibility of systematic, integrated volcano monitoring in Central and South America using space-based observations. Through coordinated contributions from multiple space agencies, regional observatories gained unprecedented access to SAR and high-resolution optical datasets, enabling more effective monitoring of active volcanoes.
Building on this success, the Volcano Demonstrator Project (2019–2023) expanded coverage to Southeast Asia and Africa, further confirming the benefits of collaborative satellite data sharing for volcano monitoring. In 2023, CEOS approved the Global Volcano Early Warning and Eruption Response from Space (GVEWERS) initiative—a permanent, sustainable framework uniting international space agencies, academic institutions, and volcano observatories. GVEWERS aims to ensure timely, free, and low-latency access to critical satellite datasets for forecasting, detecting, and tracking volcanic activity worldwide. This capability is essential for mitigating hazards and providing early warnings of potential eruption impacts, as illustrated by the role of satellite data in monitoring unrest at Fentale, Ethiopia (2024–2025), and in tracking volcanic products emplacement and redeposition at Fuego Volcano, Guatemala, since 2024.
The success of GVEWERS depends on strong engagement from the global volcanology community. We invite international participation to advance this collaborative effort, which represents a transformative step toward reducing volcanic risk through space-based Earth Observation and fulfilling the vision of the United Nations’ Early Warning 4 All program.
How to cite: Bagnardi, M., Poland, M., Albino, F., Biggs, J., Dualeh, E., Ebmeier, S., Grandin, R., Pinel, V., Pritchard, M., Wauthier, C., Way, L., and Zheng, W.: Facilitating Satellite-Based Monitoring of Volcanic Unrest and Eruptions Through Global Cooperation: The GVEWERS Initiative, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-4075, https://doi.org/10.5194/egusphere-egu26-4075, 2026.
Forecasting the evolution of landslide displacements over time in order to have a plan for early warning and risk management is meant to be searched by a comprehensive look at the past and present. In recent years, machine learning techniques have made remarkable advancements in the investigation of natural hazards, specifically by harnessing data patterns and historical information to enhance prediction accuracy. This innovative approach not only improves the understanding of the natural phenomena but also empowers the efforts to reach informed decisions based on reliable forecasts. In particular, machine learning models have the power to describe sophisticated and nonlinear relationships concerning the complex evolution of phenomena. This study highlights the effectiveness of preprocessing and feature engineering techniques, such as transformations, Fourier series, and temporal lags, when applied to the analysis of the evolution of the displacement patterns of slow landslides with time. It emphasizes that, in some cases, even with straightforward methods, like linear regression and Prophet, reliable results can be achieved. A workflow for modeling time series forecasting has been specifically developed, with the aim of processing large volumes of data, as well as incorporating selected features derived from time indices and external inputs. The results from both models, optimized through careful feature engineering, showed high reliability and performance, especially when bolstered by well-designed regressors, lag structures, and seasonal markers. In terms of accuracy, the Prophet model exhibits higher performance. The study is deemed to show that engineered features significantly decrease prediction errors, and the key takeaway highlights the importance of feature richness over model complexity.
Keyword: Machine learning, Feature engineering, Pre-processing, Landslide, displacement, Time series.
How to cite: Sabaghi, M., Parise, M., Esposito, F., Del Buono, N., and Lollino, P.: Enhancing the performance of data-driven models through pre-processing and feature engineering for the forecasting of landslide displacement , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-5327, https://doi.org/10.5194/egusphere-egu26-5327, 2026.
Based on the general assumption that global climate change and anthropogenic interventions have alarmingly affected grasslands worldwide, this study aims to investigate the current state of grassland degradation by closely examining how degradation processes are perceived across different spatial and temporal scales.
Climate change and human impact are not the sole causes of grassland degradation. The climatic factor is often separated from other influences, such as hydrology or soil properties. Vegetation dynamics are closely tied to land-surface hydrology, with positive effects when water availability is adequate and adverse effects when water is scarce or excessive. Several studies have also emphasised the influence of landform on vegetation quality and distribution. The geomorphological factor is often linked to high rates of degradation, caused by landslides and gully erosion. In terms of both geological structure and soil properties, geomorphological features are challenging to define.
Even if most researchers choose to assess them together, several studies have highlighted the need to distinguish between triggers to ensure appropriate mitigation. By supporting this statement, our analysis focuses on identifying and separating the main drivers of grassland degradation.
Some of the most qualitative and widely applied methods are based on the Normalised Difference Vegetation Index (NDVI), which is considered a proxy for grassland degradation. Thus, to determine the current status of grassland degradation in the Moldavian Plateau, a method for analysing vegetation dynamics is proposed, using NDVI Landsat 8 OLI data from 2013 to 2020 (30m); MODIS data from 2000 to 2023 (250m), and AVHRR data merged with MODIS at 9.5 km spatial resolution, materialised through the PKU GIMMS NDVI dataset available from 1982 to 2022.
The method applied separates the multiannual NDVI trend from the seasonal component. The NDVI trend analysis is essential because it provides the information needed to investigate and identify the leading degradation agents.
The high-resolution analyses captured fine-scale features, while the medium- and low-resolution analyses provided a clear picture of the primary drivers of grassland degradation. The proper association of local (e.g., overgrazing) and regional (e.g., drought) factors contributes to a better understanding of the degradation phenomenon and supports sustainable measures.
Although the analysis is more qualitative than quantitative, it emphasises the importance of local analysis in the global process assessment. Moving from one level of spatial analysis to another, we found considerable differences that can affect perceptions of the impact induced by a specific phenomenon, in this case, global climate change. At the scale of the Moldavian Plateau, grasslands remain stable from a climatic perspective, while the primary problem is associated with anthropogenic interventions.
How to cite: Crețu-Văculișteanu, G., Necula, N., and Niculiță, M.: A multiscale analysis of grassland degradation. A case study from Northeastern Romania, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-12453, https://doi.org/10.5194/egusphere-egu26-12453, 2026.
Wildfires are an increasingly critical natural hazard, requiring rapid and reliable detection in order to support emergency measures. OroraTech operates a dedicated thermal-infrared satellite constellation to address this need. As of January 2026, this constellation comprises ten satellites, providing a swath of ~400 km and imaging at a ground sampling distance of 200 m. A key feature of the system is on-orbit fire detection, where thermal data is processed directly onboard the satellites. This onboard processing minimizes downlink requirements and substantially reduces detection latency, enabling the delivery of near-real-time wildfire hotspot alerts which improves situational awareness during rapidly evolving fire events.
The OroraTech constellation will be further expanded until a global revisit time of approximately 30 minutes is reached. Such high temporal resolution, combined with low-latency onboard processing, is expected to substantially improve the early detection of emerging fires, particularly during critical afternoon and evening hours. This is where many established Earth observation (EO) missions have limited coverage.
In this contribution, we present recent observations from the constellation and evaluate its wildfire detection performance across a range of fire events. We compare our results with fire products from established EO missions, including products from VIIRS onboard the Suomi-NPP, NOAA-20 and NOAA-21 satellites as well as from FCI onboard the MTG satellite. The analysis focuses on quantifying the detection accuracy of the OroraTech products. Finally, we discuss how such agile, high-revisit cubesat observations can complement traditional satellite systems to enhance the monitoring of wildfire hazards and operational risk management.
How to cite: Pörtge, V., Appalla, S. M., Wahbe, J., Seifert, M., Bereczky, M., and Gottfriedsen, J.: Wildfire Detection Performance of OroraTech’s Thermal Satellite Constellation, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-5334, https://doi.org/10.5194/egusphere-egu26-5334, 2026.
Differential land subsidence affects many world metropolises, impacting their public and private infrastructure, including housing, transport and utility networks, social, healthcare and education facilities and, in turn, causing socio-economic impacts. This work showcases an innovative workflow based on geospatial data for exposure-vulnerability rating, hazard quantification and risk assessment. The methodology integrates Interferometric Synthetic Aperture Radar (InSAR)-derived information on ground displacement from Copernicus European Ground Motion Service (EGMS), with land cover and settlement characteristics from freely and openly available global datasets including the Copernicus Global Human Settlement Layer (GHSL) and DLR’s World Settlement Footprint (WSF). Such an integrated approach represents a significant step forward from InSAR displacement velocity-based approaches that are nowadays common in the specialist literature, to actionable risk information that are still rare. Land subsidence-induced deformation and structural stress on urban assets are quantified within the 15 metropolitan cities of Italy, along with the distribution and amount of residential/non-residential infrastructure and population exposed. Deformation-induced risk is assessed via the implementation of a tailored risk matrix enabling the geospatial intersection of four hazard (H1 to H4) and four exposure-vulnerability (EV1 to EV4) classes into 16 combinations of likelihood and impact (or also, probability and severity), and the consequent classification of risk in three levels (R1 to R3). The analysis shows that a total of 1.44 out of 2665 km2 urbanised land within the 15 cities is at high risk (R3) due to significant angular distortions (and, sometimes, additive threat from horizontal strain) affecting very high exposure-vulnerability infrastructure. Moreover, it is estimated that, for more than 2700 buildings within the 15 cities, there is high likelihood of already occurred/incipient structural damage. The reference knowledge-base on present-day subsidence-induced risk can inform land and risk management at national scale, and provides a baseline for future assessments to build upon with a look to the next decades and sustainable urban development.
This work is funded by the European Union – Next Generation EU, component M4C2; project SubRISK+ (https://www.subrisk.eu/), 2023–2026 (CUP B53D23033400001). Value-added risk mapping outputs and statistics are openly available via the SubRISK+ ‘Control Room’ web platform (https://controlroom.subrisk.eu/).
Full details about the workflow and results are available in the full paper: Cigna F., Paranunzio R., Bonì R., Teatini P. 2025. Present-day land subsidence risk in the metropolitan cities of Italy. Scientific Reports, 15, 34999 (https://doi.org/10.1038/s41598-025-18941-8).
How to cite: Cigna, F., Paranunzio, R., Bonì, R., and Teatini, P.: A novel workflow to map differential land subsidence risk using EGMS InSAR and urban settlement data: national scale assessment in Italy, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-6608, https://doi.org/10.5194/egusphere-egu26-6608, 2026.
In a context of climate change, agricultural systems are increasingly exposed to natural hazards, ranging from prolonged droughts to extreme precipitation events (IPCC, 2023). Monitoring the resilience of high-value crops, such as vineyards, is therefore critical for effective risk management and adaptation strategies. While remote sensing has been widely employed to investigate vineyard phenology (Giovos et al., 2021), current approaches rely predominantly on optical data and UAV platforms, which are limited by weather conditions and often lack the structural sensitivity required for robust biomass estimation (Weiss et al., 2020).
This study addresses this gap by exploring the potential of Synthetic Aperture Radar (SAR) technology - specifically X-band data from the COSMO-SkyMed constellation - as a tool for assessing vineyard vulnerability and structural response to environmental stressors. A limited but still significant case study is reported in northern Italy, where the winemaking region of Oltrepò pavese is experiencing a drift in crop suitability; a sample of about 40 vineyards with specific azimuth orientations of rows was defined, to avoid possible anisotropy phenomena, then time series of single-pol (HH) CSK data were identified on each vineyard for year 2024. We posit that precise knowledge of vineyard biomass is not only relevant for carbon sink quantification but is also a key indicator of the crop's capability to withstand increasingly warm and dry conditions.
This research analyses the complex interaction between vegetation structure and meteorological hazards, specifically focusing on the influence of accumulated rainfall and dew formation on radar backscatter. Building on previous, multi-sensor SAR observation of vineyards (Bergamaschi et al., 2025) we present an ordinary least squares (OLS) modelling framework to quantify the relationship between hydrometeorological variables and SAR signal variability. Our preliminary results suggest that accumulated recent rainfall acts as a significant predictor of structural changes, with precipitation over the preceding 72 hours explaining over 70% of the local radar signal evolution. This strong correlation underscores the potential of X-band SAR to serve as a reliable proxy for monitoring crop status under hydrological stress. While the proposed OLS model successfully captures the primary drivers of backscatter variability, future developments will aim to enhance risk assessment capabilities through mixed-effects models and the integration of additional biophysical parameters.
This work contributes to making Earth-observation data a valuable resource to natural hazard studies, helping to build a pathway toward operational, all-weather monitoring of agricultural risks in a changing climate.
References
IPCC. (2023). Summary for Policymakers. In Climate Change 2023: Synthesis Report. Contribution of WGs I, II and III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change.
Giovos, R., Tassopoulos, D., Kalivas, D., Lougkos, N., & Priovolou, A. (2021). Remote Sensing Vegetation Indices in Viticulture: A Critical Review. Agriculture, 11(5), 457.
Weiss, M., Jacob, F., & Duveiller, G. (2020). Remote sensing for agricultural applications: A meta-review. Remote Sensing of Environment, 236, 111402.
Bergamaschi, A. Verma, A. Bhattacharya and F. Dell’Acqua (2025). "Joint Analysis of Optical and SAR Vegetation Indices for Vineyard Monitoring: Assessing Biomass Dynamics and Phenological Stages Over Po Valley, Italy", IEEE Access, vol. 13, pp. 153886-153895, 2025.
How to cite: Bergamaschi, A., Nihar, A., Verlanti, A., Verma, A., Bhattacharya, A., Dell'Acqua, F., and Nunziata, F.: Assessing Vineyard Resilience to Hydro-Climatic Hazards: An X-band SAR Approach for Agricultural Risk Monitoring, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-6671, https://doi.org/10.5194/egusphere-egu26-6671, 2026.
In 2008, the Hawaiian Volcanoes Supersite was established to make available large amounts of satellite and other data to study Hawaiian volcanism. The location was chosen to be the first of the Geohazards Supersites and Natural Laboratories initiative because of the history of volcanological research on the Island of Hawaiʻi and the need for hazards monitoring and mitigation. Ground-based data are collected by the U.S. Geological Survey Hawaiian Volcano Observatory, and national space agencies provide access to satellite synthetic aperture radar and other imagery that would not otherwise be freely obtainable. The vast quantity of open space-based data has contributed to: (1) development of new methodologies; (2) successful responses to volcanic crises; and (3) innovative multidisciplinary research. There remain opportunities for further growth, particularly regarding better coordination among supersite users and implementation of synergistic studies that make use of the full spectrum of available data, including for non-volcanology applications. Nonetheless, the Hawaiian Volcanoes Supersite demonstrates the importance of freely available, low-latency data, especially from satellites, to disaster risk management and reduction—a vision that has been articulated in numerous international agreements.
How to cite: Poland, M., Bagnardi, M., Salvi, S., Amelung, F., Paladino, T., Johanson, I., and McLay, M.: The Hawaiian Volcanoes Supersite: Demonstrating the benefits of open data for science and society, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8195, https://doi.org/10.5194/egusphere-egu26-8195, 2026.
A novel methodology for assessing the catchments most severely affected by extensive heavy rainfall events is presented to support post‑disaster recovery activities. The approach exploits pre‑ and post‑event optical Sentinel‑2 imagery to perform a dual change‑detection analysis. The first component targets land‑cover alterations (Land Cover Change Detection, LCCD), including slope denudation, debris deposition, and alluvial flooding. The second component focuses on variations in the optical properties of surface waters (Water Colour Change Detection, WCCD), such as colour shifts of lake and river waters associated with increased turbidity.
Integrating information on water‑colour change (WCCD) with a more traditional change detection analysis based on variations in vegetation spectral indices (LCCD) is advantageous in high‑altitude environments, where vegetation cover is sparse or absent. The combined change detection compensates for the individual limitations of each method and enhances overall performance by 6–11% and 31–38% compared with the standalone use of LCCD and WCCD, respectively. The final product is a severity map that classifies catchments into increasing levels of impact, derived from the aggregated magnitude of changes detected by both the LCCD and WCCD components.
The methodology relies entirely on freely accessible datasets (Copernicus Sentinel‑2 imagery, the TINITALY 1.1 Digital Elevation Model, and OpenStreetMap layers), and all processing steps are implemented using open‑source software (Google Earth Engine, QGIS, and R), ensuring its potential applicability at the global scale. The approach was tested on two distinct heavy‑rainfall events that affected the northwestern Italian Alps in June and September 2024. Across these case studies, the methodology achieved a correspondence rate of 59–65% between the catchments identified as severely affected and those containing documented natural instability events.
How to cite: Matta, E., Nigrelli, G., Alberto, W., Filipello, A., Zittlau, M., and Chiarle, M.: EO based assessment of geomorphological impacts caused by heavy rainfall events in high mountain areas, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-9143, https://doi.org/10.5194/egusphere-egu26-9143, 2026.
Wildfires are an emerging peril in traditional natural hazard risk assessment. Remote sensing data comprises the most comprehensive data source for their assessment.
However, scientists and practitioners in Disaster Risk Reduction are faced with several fire products from different satellite missions, whose differences, advantages and limitations can be difficult to access and understand, especially for users outside the remote sensing domain. This complicates the process of identifying the most appropriate dataset, making it a challenging and time-consuming endeavor, and in some cases can result in suboptimal or even erroneous results.
We address this issue by offering a concise overview of remote sensing fire products and clarifying terms that are interpreted differently across scientific communities, with a focus on their application in risk assessment. Moreover, we provide risk estimates based on different historic wildfire hazard sets. These are derived from MODIS satellite products for the years 2002–2024, leveraging burned area, fire radiative power and land use information. We join these hazard sets with exposure datasets (representing physical assets, population and forested area) and damage records to calibrate their vulnerabilities to wildfires. These form the basis for estimating wildfire impacts and risks, while quantifying uncertainties related to the chosen hazard representation. Such risk analyses find application in prioritising adaptation options and in designing insurance products.
How to cite: Steinmann, C. B., Koh, J., Kropf, C. M., Scholten, R. C., Bresch, D. N., and Hantson, S.: Quantifying global wildfire impacts to natural and human systems using remote-sensing data, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-9374, https://doi.org/10.5194/egusphere-egu26-9374, 2026.
How to cite: Dimitriou, I. C.: Vulnerability Index for croplands of Greece, with a twist, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-9563, https://doi.org/10.5194/egusphere-egu26-9563, 2026.
Groundwater overexploitation is a widespread problem that can lead to land subsidence, with significant impacts on infrastructure and ecosystems. Recent advances in satellite Earth observation allow the systematic monitoring of ground deformation and groundwater storage changes over large areas.
In this work, we explore the potential of combining satellite-based land deformation data with independent information on groundwater storage evolution to investigate groundwater-related subsidence at large spatial scales in the Spanish territory. Interferometric Synthetic Aperture Radar (InSAR) products are used to identify areas affected by significant ground motion, while satellite gravimetry data provide complementary insights into regional groundwater storage trends.
The spatial comparison of these datasets highlights areas where ground deformation and groundwater depletion signals coexist, suggesting a strong link between subsidence processes and intensive groundwater use. The results illustrate how multi-sensor satellite observations can support the identification of priority areas for groundwater management and risk assessment.
This study demonstrates the value of integrated satellite approaches as a screening tool to support sustainable groundwater management at regional to national scales.
How to cite: Guardiola-Albert, C., Bru, G., Béjar-Pizarro, M., and Ezquerro, P.: Large-scale assessment of land subsidence related to groundwater depletion using satellite observations, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-9642, https://doi.org/10.5194/egusphere-egu26-9642, 2026.
GNSS and InSAR observations have detected almost continuous inflation since October 2023 within the Svartsengi Volcanic System (SW Iceland). Inflation was interrupted by rapid deflation and concurrent dike intrusions, resulting in a total of nine eruptions (as of January 2026, the time of writing) within the Sundhnúkur crater row and its extension.
Here, we focus particularly on inflation events to improve our understanding of magma supply and the evolution of the magmatic system following each diking event/eruption. The InSAR observations were acquired from multiple spaceborne SAR satellite missions (e.g., Sentinel-1, TerraSAR-X, and COSMO-SkyMed) while the GNSS observations were obtained from the well-established geodetic network operating at and surrounding the Svartsengi volcanic system. These dataset were jointly modeled using a variety of source geometries (e.g., spherical, sill-type, and ellipsoidal) embedded in a homogeneous, elastic half-space, allowing us to assess how source shape affects the inferred depth and volume of inflation.
Analysis of the modeling results reveals a clear pattern. The earliest inflation episodes (up to February–March 2024) were relatively short, lasting several weeks, and exhibited strong variability in both inferred source depth and recharge volume from one inflation episode to another across all tested source geometries, before triggering a new dike intrusion or eruption. Across the three tested source geometries, inferred source depths ranged as follows: 3.5–4.5 km for spherical sources, 3–3.5 km for ellipsoidal sources, and 4–5.5 km for sill-type sources. Corresponding volumes are approximately ranging from 4 to 21 × 10⁶ m³, 3 to 19 × 10⁶ m³, and 5 to 24 × 10⁶ m³ for spherical, ellipsoidal, and sill-type sources, respectively.
Since March 2024, the system appears to have become more stable: although absolute depth and volume estimates still depend on the assumed source geometry, the inferred depth and volume for each individual geometry have remained fairly consistent across successive events. Specifically, depths have stabilized around 3.9 ± 0.2 km for spherical sources, ~3 ± 0.2 km for ellipsoidal sources, and 4.7 ± 0.1 km for sill-type sources, with corresponding volumes approximately ranging from 18 to 21× 10⁶ m³, 15 to 18 × 10⁶ m³, and 22 to 25 × 10⁶ m³, respectively.
A detailed study of the volcanic system and its temporal evolution can provide critical insights into the processes governing magma accumulation. Geodetic data form the cornerstone of this analysis, and when combined with seismic, petrological, and other multidisciplinary observations, they allow a more accurate interpretation of the system’s pre-eruptive behaviour. Linking system evolution with these multi-parameter observations enables better characterization of inflation episodes, supporting improved forecasting and more reliable assessment and mitigation of associated volcanic hazards.
How to cite: Lanzi, C., Parks, M., Drouin, V., Sigmundsson, F., Geirsson, H., Hooper, A., Ófeigsson, B. G., and Friðriksdóttir, H. M.: Inflation Dynamics and Magma Recharge within the Svartsengi Volcanic System (SW Iceland) Inferred from GNSS and InSAR Data , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-12710, https://doi.org/10.5194/egusphere-egu26-12710, 2026.
Mediterranean grasslands are strongly constrained by water availability and exhibit pronounced seasonal variability in productivity. Remote sensing provides valuable tools for vegetation monitoring, with the Normalized Difference Vegetation Index (NDVI) being one of the most widely used indicators. However, in Mediterranean environments the relationship between NDVI and actual aboveground biomass is often complex due to spectral saturation, summer senescence, and a decoupling between greenness and structural biomass. Identifying when NDVI reliably represents grassland production is therefore essential for its application in grazing management and climate impact assessments.
This study was conducted in three representative grassland areas of the Community of Madrid (central Spain): Piñuecar, Colmenar Viejo, and Tielmes, spanning a gradient from humid mountain environments to semi-arid lowlands. NDVI time series were derived from the MODIS MOD09Q1 product for the period 2000–2025, with 250 m spatial resolution and an 8-day temporal frequency. Aboveground biomass was estimated using the SIMPAST predictive grassland model, driven by climate data from six global climate models from the CMIP6 ensemble. NDVI was analysed both as instantaneous values and as temporally accumulated NDVI using simple integration. The relationship between NDVI and biomass was evaluated through linear regressions and coefficients of determination (R²) at annual, seasonal, and phenological scales.
Instantaneous NDVI showed almost no explanatory power for biomass variability, with R² values close to zero (≈ 0.00–0.03), indicating that punctual greenness indicators fail to represent accumulated grassland production. In contrast, temporally accumulated NDVI exhibited a strong relationship with annual biomass, with R² values ranging from approximately 0.60 to 0.75 across sites. Seasonal analyses revealed that the highest correlations occurred during autumn and spring, coinciding with periods of active growth. During summer and winter, NDVI–biomass relationships weakened considerably due to senescence, and reduced metabolic activity.
Segmenting the annual cycle into five eco-physiological periods further improved the coherence between the spectral signal and actual growth dynamics, reaching maximum R² values of up to 0.74–0.75 during peak growth phases. Piñuecar showed the strongest NDVI–biomass coupling, while Tielmes achieved high correlations during episodic humid pulses despite its generally arid conditions. Colmenar Viejo exhibited greater interannual variability, likely linked to heterogeneous water stress.
These results confirm that temporal integration of NDVI is essential to represent productivity in Mediterranean grasslands. Phenological segmentation allows identification of time windows in which NDVI acts as a reliable proxy for real growth, providing operational criteria for grassland monitoring under climatic variability..
References
Aragón Pizarro, M., Díaz-Ambrona, C. G. H., Tarquis, A. M., Almeida-Ñauñay, A. F., and Sanz, E.: Modelling Biomass Projections in Grasslands of Central Spain Under Climate Change Scenarios, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-10928, https://doi.org/10.5194/egusphere-egu25-10928, 2025.
Iglesias, E., Báez, K., H. Diaz-Ambrona, C. Assessing drought risk in Mediterranean Dehesa grazing lands. Agricultural Systems, 149, 65-74, 2016. https://doi.org/10.1016/j.agsy.2016.07.017
Acknowledgements
The first author acknowledges the support of Project “Garantía Juvenil” scholarship from Comunidad de Madrid. This research was partially supported by Universidad Politécnica de Madrid under project “Clasificación de Pastizales Mediante Métodos Supervisados – SANTOS” (RP220220C024).
How to cite: Berzal, A., Sanz, E., Diaz-Ambrona, C. G. H., Almeida, A. F., Martín, J. J., and Tarquis, A. M.: Relationship between NDVI and biomass in Mediterranean grasslands under climatic variability, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10894, https://doi.org/10.5194/egusphere-egu26-10894, 2026.
The Virunga volcanoes, including Nyiragongo and Nyamulagira, pose a continuous threat to approximately 2.5 million inhabitants in the cities of Goma (Democratic Republic of the Congo - DRC) and Gisenyi (Rwanda), as well as to the surrounding settlements situated at their base. The volcanic hazards include lava flows, permanent gases and ash emissions, mudflows, ground deformation and fissuring, in addition to the CO2 and CH4 dissolved in lake Kivu, one of the large African Rift lakes situated between the DRC and Rwanda. These hazards are exacerbated by the vulnerable living conditions of the local populations and the insecurity resulting from armed conflicts that have persisted over the last 3 decades. Furthermore, limited financial and technical resources, together with recurrent armed conflicts, have hindered the efforts of the Goma Volcano Observatory (GVO) - the government institution in charge of monitoring the volcanoes and lake Kivu to develop a reliable early warning system, especially ground-based networks. In 2017, a permanent Supersite was established over the Virunga to enhance geophysical scientific research and geohazards assessment, with the aim to assist emergency managers in making informed decisions during volcanic unrest and improve eruption forecasting. In addition to the freely accessible Earth Observation (EO) data (e.g. ASTER, Landsat, Sentinel), the CEOS (Committee on Earth Observation Satellites) guarantees the access -free of charge- to COSMO-SkyMed, Pleiades and SAOCOM images, and supports the production of hazard, risk and recovery maps through the Copernicus EMS services using EO data. The pool of voluntary scientific collaboration built around the Virunga Supersite supports, on a fair basis, the enhancement of the expertise of local scientists for EO data processing and interpretation to improve volcanic hazard assessment and produce effective risk reduction strategies. Hence, the EO data enabled the generation of lava flow hazard maps and the assessment of transportation network vulnerability in Goma. EO data played a crucial role in the emergency response to the May 2021 Nyiragongo eruption, specifically in mapping and modelling the associated dyke intrusion. Furthermore, maps of daily SO2 and ash dispersion are produced as well as the modelling of hazard forecasting. EO data also supports the routine monitoring of Nyiragongo and Nyamulagira volcanoes, enabling the GVO to estimate the daily rate of gas emissions and ground deformation. It provides critical oversight of Nyiragongo ongoing effusive activity inside the main crater which potentially holds a permanent lava lake, while monitoring Nyamulagira’s intermittent caldera overflows which could give rise to lava flows that threaten populations living south of the volcano. Overall, the availability of EO data and the collaborative effort to generate EO-based data products are key resources to monitor this high-risk volcanic region, overcoming the lack of ground-based networks, and hence are unique tools to promptly assess volcano hazard.
How to cite: Balagizi, C., Ciraba, H., Sambo, G., Iragi, K., Valade, S., Coppola, D., Nobile, A., Corradino, C., Cappello, A., Ganci, G., Proietti, C., Naranjo, C., Beccaro, L., Corradini, S., Tolomei, C., Polcari, M., and Trasatti, E.: Key role of earth observation data for monitoring and assessing volcanic hazard in resource-limited and conflict-affected settings: the case of DR Congo (Africa), EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-11182, https://doi.org/10.5194/egusphere-egu26-11182, 2026.
The Eastern Mediterranean is characterized by a complex geodynamic regime associated with the interaction between the Eurasian and African plates, giving rise to significant seismicity, active faulting, tectonic uplift, landslides, rockfalls, and subsidence processes. Cyprus, located at the transition from oceanic subduction to continental collision, represents a unique natural laboratory for investigating these processes and their societal impacts. To address long-standing gaps in geodetic and Earth Observation (EO)–based hazard monitoring in the region, the Cyprus Geohazard Observatory Supersite (CyGOS) has been established in 2025 as a Permanent Supersite within the GEO Geohazard Supersites and Natural Laboratories (GSNL) framework.
CyGOS builds upon the CyCLOPS strategic research infrastructure, integrating dense networks of Tier-1 GNSS permanent stations co-located with meteorological sensors, tiltmeters, and calibration-grade InSAR corner reflectors, together with multi-mission SAR data provided through Committee on Earth Observation Satellites (CEOS) support. The Supersite provides a coordinated framework for the acquisition, calibration, and integration of EO and in-situ data to deliver high-resolution ground deformation products relevant to seismic hazard assessment, landslide monitoring, subsidence detection, and long-term tectonic strain analysis.
CyGOS already contributes to and interoperates with regional and global research infrastructures and services, including the European Plate Observing System (EPOS) TCS-GNSS, the EUREF Permanent Network (EPN), and the European Ground Motion Service (EGMS). GNSS time series and velocity solutions from CyGOS are provided to EPOS and EPN, supporting reference-frame densification and long-term deformation monitoring in a tectonically active region where high-quality geodetic constraints remain sparse. In parallel, GNSS-calibrated InSAR products and nationwide velocity fields are developed to enhance the interpretation, validation, and regional relevance of EGMS products, particularly in areas affected by rapid or highly localized deformation.
Beyond its current contributions, CyGOS aims to further strengthen its role within European and global initiatives by (i) delivering a validated national ground motion service for Cyprus that is fully interoperable with EGMS, (ii) providing calibration and validation datasets for multi-mission SAR time series through its permanent and mobile corner reflector infrastructure, including potential contributions to the CEOS Working Group on Calibration and Validation (WGCV–SAR), and (iii) enabling cross-domain integration of geodetic, seismic, geological, and environmental datasets. These objectives are designed to support comparative studies across GSNL Supersites, improve the robustness of EO-based hazard products, and facilitate methodological benchmarking and reproducible research. All datasets and derived products are disseminated to the scientific community following FAIR principles, fostering open collaboration, reuse, and innovation in EO-based natural hazard research.
This contribution introduces the CyGOS Supersite concept, infrastructure, and initial activities, and discusses its role within the GSNL, CEOS and EPOS frameworks as a regional hub linking EO-based geohazard monitoring with European and global initiatives.
How to cite: Danezis, C., Zomeni, Z., Brcic, R., Kotsakis, C., Ganas, A., Kakoullis, D., Fotiou, K., Ibarrola Subiza, N., Chatzinikos, M., and Nikolaidou, T.: CyGOS: A Permanent GEO GSNL Supersite for EO-driven Multi-Hazard Monitoring in the Eastern Mediterranean, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-12948, https://doi.org/10.5194/egusphere-egu26-12948, 2026.
The Copernicus programme provides one of the most comprehensive Earth Observation (EO) data offers worldwide. Beyond the Sentinel missions, Copernicus Contributing Missions (CCM) complement spectral, spatial, and temporal gaps by delivering high-resolution optical, radar, three-dimensional digital elevation models, and methane hotspot data globally, ensuring that Copernicus user needs are effectively addressed. EO data is a critical enabler for informed decision-making, and CCM supports this through systematic and on-demand data delivery across a wide range of applications, including emergency rapid mapping, risk and recovery analysis, security monitoring, methane emission detection, and large-area coverage for marine and land domains.
Since the programme’s operational start in 2015, building on the GMES legacy, these datasets have been primarily available to the eligible Copernicus Services but can also be accessed by national public authorities in Europe. This contribution showcases CCM datasets used in emergency contexts, demonstrating how the combined use of systematic and on-demand acquisitions offers unique opportunities to investigate disasters and post-disaster dynamics beyond Sentinel capabilities. Access to these resources is facilitated through the Copernicus Data Space Ecosystem (CDSE) and dedicated on-demand services such as the Rapid Response Desk (RRD), enabling researchers to exploit multi-source data in an integrated environment.
In parallel with the operational data offer, ESA, in collaboration with DG DEFIS, has launched a tier of activities to introduce new EO commercial data domains into the Copernicus programme, including thermal infrared, hyperspectral, and radiofrequency, as well as innovative capabilities such as video collection, onboard processing, and AI-driven analytics. These developments aim to expand the Copernicus portfolio and support European industrial competitiveness while addressing emerging user needs. Examples of potential applications include monitoring land surface temperature, detecting urban heat stress, and improving hazard forecasting models.
Case studies presented in this contribution illustrate how Copernicus datasets, combined with new commercial capabilities, are reshaping opportunities for public authorities and researchers to exploit EO data for rapid hazard assessment, multi-hazard modeling, and resilience planning.
Keywords: Copernicus, Copernicus Contributing Missions, Natural Hazards, Copernicus Services, EO data legacy, EO Innovation.
How to cite: Manzo, C., Fischer, P., Esteve, R., Goor, F., Korzeniowska, K., Amans, V., Losco, P., and Di Ciollo, C.: Expanding Copernicus EO Capabilities for Hazard Monitoring: From Contributing Missions to Emerging Data Domains, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-13447, https://doi.org/10.5194/egusphere-egu26-13447, 2026.
The intensification of the hydrological cycle as a consequence of climate change is altering the distribution and intensity of hydro-climatic extreme events. Extremes at both ends of the hydrological cycle, dry events (droughts) and wet events (heavy rainfall), are increasing in frequency and intensity. When such events occur in cascades, their compounding impacts can increase in severity. However, datasets explicitly designed to study such cascades remain scarce. Within the ESA-funded ARCEME project (Adaptation and Resilience to Climate Extremes and Multi-hazard Events), we introduce a dataset tailored to study cascading droughts and heavy precipitation events. The events are identified using a dual sampling strategy relying on climatological anomaly detection, and on sampling from the footprints of disaster events reported by the Emergency Database (EM-DAT). The resulting dataset provides over 400 georeferenced datacubes with a spatial extent of 10 by 10 km, each covering one year before and one year after the cascading event, and sampling a wide range of climate zones and terrestrial biomes. Each datacube integrates (i) Sentinel-2 L2A optical imagery, (ii) Sentinel-1 radiometric Terrain Correction (RTC) radar data, (iii) Ancillary Landcover and topographic information. Optimized for the analysis of impacts on natural and managed vegetation, the dataset provides a standardized data collection suitable for data-driven studies of compound and cascading hydro-climatic extremes.
How to cite: Teber, K., Weynants, M., Gans, F., Kluczek, M., Bojanowski, J. S., and Mahecha, M. D.: A curated sentinel collection to study cascading droughts and extreme precipitation events, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-14030, https://doi.org/10.5194/egusphere-egu26-14030, 2026.
Subsidence in already emplaced lava can be caused by contraction as they cool, degassing and collapse. In this study, we present preliminary estimates of short-term volume change associated with the January (14 Jan, 07:58 UTC – 16 Jan, 01:08 UTC) and February (8 Feb, 06:03 UTC – 9 Feb, afternoon) 2024 eruptions at the Sundhnúkagígar crater row, Iceland. Surface elevation changes were derived from a series of multi-temporal pre- and post-eruptive Digital Elevation Models (DEMs) based on stereo imagery collected by UAV and manned aircraft, using a photogrammetric workflow in the Agisoft Metashape software. Volume changes were quantified from DEMs of Difference (DoDs) by integrating surface elevation changes over the mapped lava field, accounting for random, spatially correlated, and systematic errors. Positive and negative lava volume estimates represent the areal integration of surface uplift and subsidence respectively, as a result of the eruption, rather than strictly representing net mass addition or removal at a given location. Positive changes may reflect lava emplacement or internal redistribution of previously erupted material, whereas negative changes indicate thermal contraction, lava drainage, degassing, or collapse of the cooling lava surface. All reported volumes refer to changes integrated over the mapped lava field only. During the January eruption, rapid emplacement between 14 and 15 January resulted in a dominance of positive volume change, with 0.463 ± 0.0005 Mm³ of positive and −0.233 ± 0.0004 Mm³ of negative volume change. Between 15 and 17 January (approximately 48 h after eruption onset), volume changes were dominated by surface lowering, with −0.094 ± 0.0009 Mm³ negative versus 0.047 ± 0.0006 Mm³ positive volume change, reflecting contraction and internal redistribution as the dominating processes. From 17 January to 13 February, volume changes were minor, with 0.049 ± 0.001 Mm³ positive and −0.040 ± 0.001 Mm³ negative. For the February eruption, the analysis was constrained by the geological setting and the short repose time between eruptions, as rapid resurfacing of the lava field by subsequent eruptive activity limited the temporal persistence of measurable surface changes. On 8 February, comparison of DEMs acquired at 13:15 and 17:05 UTC shows a dominance of negative volume change, with −1.65 ± 0.01 Mm³ of negative versus 1.35 ± 0.01 Mm³ of positive volume change. Between 8 February (17:05 UTC) and 13 February, negative changes −2.07 ± 0.06 Mm³ exceeded positive changes 1.15 ± 0.04 Mm³. Ongoing work aims to further refine these results by quantifying vertical surface subsidence rates to better characterize post-eruptive surface change behaviour.
How to cite: Schwegmann, S., Belart, J. M. C., Sigmundsson, F., Rom, J., and Birkefeldt Moller Pedersen, G.: Post-eruptive Evolution of Lava Fields at the Sundhnúkagígar Crater Row, Svartsengi Volcanic System, Iceland: A Multitemporal Analysis Based on Aerial Stereo Imagery, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-14393, https://doi.org/10.5194/egusphere-egu26-14393, 2026.
The accelerated retreat of Andean glaciers is one of the most evident impacts of climate change, with direct implications for environmental stability and water security. In this context, the progressive exposure of sulfide‑rich lithological units promotes the generation of Acid Rock Drainage (ARD), a process that degrades water quality and poses a threat to downstream ecosystems and water uses. Despite its environmental relevance, the study of ARD in high-mountain regions remains limited due to terrain inaccessibility and the site-specific and high-cost nature of traditional methods, which are based exclusively on field sampling and laboratory analyses.
This study presents an innovative methodological framework, implemented in Google Earth Engine, for the probabilistic mapping of ARD in glacial retreat zones of the Cordillera Blanca (Áncash, Peru). We integrated Sentinel‑2 surface reflectance imagery, spectral ratios sensitive to iron oxides, topographic variables derived from a 12.5m ALOS PALSAR digital elevation model, and an ordinal geological classification based on ARD generation potential. In addition, field spectral signatures resampled to the satellite sensor were incorporated through the Spectral Angle Mapper (SAM), providing independent physical information on mineralogical alteration processes.
Variable selection was performed through correlation analysis and multicollinearity diagnostics (VIF ≤ 5), ensuring a parsimonious and physically interpretable set of predictors. The performance of three nonlinear algorithms (Random Forest, SVM, and XGBoost) was evaluated under a spatial cross-validation scheme using 5 km hexagonal blocks, designed to minimize biases associated with spatial autocorrelation. Results showed that Random Forest achieved the best performance, with an AUC of 0.96 and an F1‑score of 0.90 under spatial validation, demonstrating strong generalization capability. Model interpretability analysis using SHAP revealed that the ferric iron index and SAM spectral similarity were the most influential predictors, confirming the importance of integrating field data into remote‑sensing‑based approaches.
The resulting probabilistic map identifies ARD hotspots concentrated in recently exposed periglacial zones, consistent with field observations based on physicochemical parameters and heavy‑metal analyses in water. This study demonstrates the effectiveness of combining remote sensing, machine learning, and geological knowledge for monitoring ARD in glaciated mountain ranges, providing a cost-effective and scalable tool that contributes to environmental risk management in a changing climate.
How to cite: Santiago-Bazan, F., Herrera-Nizama, J., Montano, Y., Sánchez-Carrión, E., and Camacho-Hernández, M.: Probabilistic mapping of acid rock drainage in glacial retreat zones using multi-source remote sensing and machine learning, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-14422, https://doi.org/10.5194/egusphere-egu26-14422, 2026.
Differential Synthetic Aperture Radar (SAR) Interferometry (DInSAR) is a well-established remote sensing technique that enables to measure Earth surface displacements with centimeter-to-millimeter accuracy. In particular, this technique exploits the phase differences between two SAR images relevant to acquisition pairs carried out over the same area in different epochs, with (nearly) the same illumination geometry. DInSAR was initially developed to analyze single deformation events, such as earthquakes and volcanic unrests. More recently, multi-temporal DInSAR techniques have been developed to track the temporal evolution of detected surface deformation by retrieving displacement time series.
The large availability of spaceborne SAR systems is characterized by dawn-dusk, sun-synchronous systems. In this orbital design, the interferometric performance may exhibit some drawbacks related to the revisit time and/or the spatial coverage. Moreover, the low sensitivity to the North-South deformation component typical of sun-synchronous DInSAR systems is a limitation for investigating deformation phenomena.
In this context, constellations of small SAR satellites are increasingly becoming an effective solution. Compared with “conventional” SAR spaceborne systems, small satellites have reduced design, engineering, and management costs. Moreover, the possibility of launching multiple satellites on the same vector allows space agencies to deploy constellations in a single mission. However, due to their reduced size and weight, such systems have limited imaging performance, which could jeopardize their coverage capability and imaging performance. Accordingly, effective exploitation requires innovative mission configurations.
This work provides an update on the NIMBUS-SAR mission, part of the SAR component of the Italian IRIDE program, which will include two batches of 6 high-resolution X-band small satellites each, operating at altitudes between 490-550 km.
To achieve the goal of covering the Italian territory with high spatial resolution and short interferometric revisit time, the mission will employ a Medium Inclination Orbit (MIO) solution. This will allow to effectively cover the whole Italian territory in 6 days and, through the DInSAR exploitation, to measure also the North-South deformation component, thus permitting us to investigate the three-dimensional behavior of the detected displacements. More specifically, the NIMBUS-SAR constellation will be deployed in 49° right-looking (batch 1, to be launched at the end of 2026) and 43° left-looking (batch 2, to be launched at the end of 2027) inclination orbits.
In this contribution, we first provide an update on the expected DInSAR performance of the NIMBUS-SAR mission, with emphasis on the retrieval capability of the North-South deformation component. Moreover, to fully assess the MIO configuration DInSAR performance in a natural hazard scenario, we also present some results obtained as output of an experimental campaign conducted by Capella Space over the Campi Flegrei caldera (Italy), which is characterized by renewed uplift phenomena since 2005. Specifically, four Stripmap SAR datasets were collected through ascending and descending orbits and right- and left-looking directions, exploiting complementary satellite heading angles. The presented results demonstrate the feasibility of using MIO DInSAR data to high accurately retrieve 3D displacements, particularly of the North-South component, in an active volcano monitoring scenario.
How to cite: Lanari, R., Berardino, P., Bonano, M., Casu, F., Costa, G., Cotugno, F., Faccin, V., Gulino, M., Manunta, M., Minchella, A., Montuori, G., Renga, A., and Stella, C.: DInSAR performance of the NIMBUS-SAR medium inclination orbit mission for 3D surface displacements retrieval in natural hazards scenarios, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-14811, https://doi.org/10.5194/egusphere-egu26-14811, 2026.
Earth surface deformation due to volcanic and tectonic activities, and landslides have significant impacts on both human society and the natural environment. Satellite remote sensing, particularly Interferometric Synthetic Aperture Radar (InSAR), is a powerful tool to obtain extensive ground displacement spatio-. However, at large spatial scales, the monitoring data often contain mixtures of multiple deformation processes, making direct interpretation highly challenging. This complexity calls for data-mining approaches to make large-scale ground motion monitoring data readily interpretable for end users.
To address this issue, we investigate an unsupervised and explainable framework for large-scale analysis of InSAR displacement time series. We use the European Ground Motion Service (EGMS) Level 2b ascending and descending dataset over Reykjanes Peninsula, Iceland, as a case study. The proposed workflow integrates statistical source separation and deep learning-based clustering to extract, group, and interpret dominant deformation patterns.
First, a statistical analysis of large scale InSAR time series is conducted using independent component analysis to extract the dominant deformation sources. Second, these components are interpreted in terms of physical processes by integrating external geophysical and environmental datasets, such as geological maps, tectonic structures, and topographic features. Third, a deep clustering network is applied to the time series data to group deformation patterns into interpretable categories that reflect distinct ground motion behaviours.
This work contributes towards the development of scalable and explainable ground motion classification analysis tools from massive InSAR time series data, offering valuable support for decision-making and early warning systems in relevant management and disaster response agencies.
How to cite: Dong, Y., Van Wyk de Vries, M., Nava, L., Gualandi, A., Floris, M., and Catani, F.: Unsupervised Analysis of Large-Scale EGMS PS-InSAR Time Series for Ground Deformation Processes: A Case Study in Iceland, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-15699, https://doi.org/10.5194/egusphere-egu26-15699, 2026.
The China Seismo-Electromagnetic Satellite (CSES) mission provides in-situ measurements of plasma parameters, electromagnetic fields, and energetic particles in the topside ionosphere, with the primary objective of characterizing ionospheric disturbances associated with seismic activity and solar–terrestrial interactions. In this context, we present CSESpy (Papini et al., 2025), a Python package that offers streamlined access to CSES Level 2 data products and expedites higher-level analysis and visualization across multiple payloads and both CSES-01 and CSES-02 spacecraft. Here, we illustrate the capabilities of CSESpy through a typical use case: the characterization of coseismic ionospheric electromagnetic anomalies associated with the 14 August 2021 Haiti earthquake (Recchiuti et al., 2023). Building on this case study, CSESpy is then used to extend the search for ionospheric electromagnetic anomalies to all geographic locations sampled by CSES, exploiting the full data set. The results demonstrate the potential of CSESpy as a powerful tool for systematic earthquake studies and for the investigation of complex events involving coupled variations across multiple physical observables in the near-Earth electromagnetic environment.
References
[1] Papini, E., Follega, F. M., Battiston, R., & Piersanti, M.: CSESpy: A unified framework for data analysis of the payloads on board the CSES satellite, Remote Sensing, 17(20), 5070, 2025, doi:10.3390/rs17205070
[2] Recchiuti, D., D’Angelo, G., Piersanti, M., Di Ruzza, S., Cicone, A., & Battiston, R.: Detection of electromagnetic anomalies over seismic regions during two strong (Mw > 5) earthquakes, Frontiers in Earth Science, 11, 1152343, 2023, doi:10.3389/feart.2023.1152343.
How to cite: Papini, E., Follega, F. M., Giordano, D., Recchiuti, D., D'Angelo, G., Piersanti, M., Battiston, R., Parmentier, A., Diego, P., Ubertini, P., and Picozza, P.: CSESpy: A unified framework for data analysis of the payloads on board the CSES-01 and CSES-02 satellites with applications to Earthquake studies., EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-17429, https://doi.org/10.5194/egusphere-egu26-17429, 2026.
In 2021, an eruptive episode started on the Reykjanes peninsula in the southwestern part of Iceland. To date (January 2026), at least 14 dike intrusions have occurred along this part of the mid-Atlantic plate boundary, 12 of which reached the surface in eruptions. The intrusive activity is accompanied by seismic activity and has also triggered earthquakes as large as M 5.6 along the boundary. All this activity has led to widespread surface faulting with associated hazards. Previous studies indicate that during such episodes, most of the six volcanic systems in the peninsula become activated, on the timescale of tens to a few hundreds of years. Some of these volcanic systems extend into cities and towns, including the eastern part of the capital area of Reykjavík.
There are different types of hazards and risks due to fault movements: a) Damage to houses and buildings on or near fault ruptures. b) Damage to infrastructures that cross active fault scarps, such as roads, pipelines, and powerlines. c) Fault movements can cause opening and dislocations of faults, which can be hazardous for people and livestock. d) Fault movements can cause the formation of sinkholes above the faults, which is also hazardous for people and livestock. e) Grabens can subside, sometimes below water-level, causing inundation of previously dry land. f) Fault movements can cause changes in borehole pressure, either increase or decrease, causing lack or overflow of water.
The Icelandic Meteorological Office currently works on a volcanic hazard and risk assessment for the entire Reykjanes peninsula. Communities and stakeholders can use it for planning in order to minimize societal disruptions due to such unrest periods. This includes a hazard assessment for fault movements, which are often associated with volcanic unrest, such as the one currently ongoing. This assessment is built upon work where multiple types of remote-sensing data, including aerial photographs, digital elevation models (DEMs), and InSAR images have been used for fault mapping, including faults that have recently been activated. The project also includes an assessment of how active different parts of the volcanic systems are. Such an assessment can be complicated, as the fractures and faults are located in different types of material (loose soil or lavas) of different ages. The long-term hazard assessment for fault movements will thus be a valuable tool to increase the resiliency of the society and its infrastructures due to such events.
How to cite: Hjartardóttir, Á. R.: Using remote-sensing data for long-term hazard assessment due to fault movements in the Reykjanes peninsula, Iceland, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-18734, https://doi.org/10.5194/egusphere-egu26-18734, 2026.
In Asturias (NW Spain), landslides are among the most significant geological hazards, causing major economic losses and fatalities every year. This study analyzes ground movements along the entire Asturian coast using Advanced Differential SAR Interferometry (A-DInSAR) techniques. Data provided by the European Ground Motion Service (EGMS; Crosetto et al., 2020) for the period 2018-2022 were acquired in ascending and descending orbits (Level 2b) and vertical and horizontal components (Level 3). These products were subsequently processed using ADAtools (Navarro et al., 2020) software to obtain deformation velocity maps and Active Deformation Area (ADA) maps. Overall, the results show a notable concentration of ADAs along the central Asturian coast, whereas the western and eastern coasts show isolated ADAs. Most ADAs are associated with to deformations in port and road infrastructure, as well as coastal landslides. The EGMS has proven to be a very useful tool for detecting and characterizing ground deformations with millimetric precision, as well as for monitoring large coastal areas free of charge and accessible to both expert and non-expert users.
Crosetto, M.; Solari, L.; Balasis-Levinsen, J.; Bateson, L.; Casagli, N.; Frei, M.; Oyen, A.; Moldestad, D.A.; Mróz, M. Deformation monitoring at European Scale: The Copernicus Ground Motion Service. In Proceedings of the International Archives of the Photogrammetry, Remote Sensing and Spatial Information Science, XXIV ISPRS Congress 2021, Nice, France, 5–9 July 2021; Volume XLIII-B3-2021, pp. 141–146.
Navarro, J. A., Tomás, R., Barra, A., Pagán, J. I., Reyes-Carmona, C., Solari, L., Vinielles, J. L., Falco, S., and Crosetto, M. (2020). ADAtools: Automatic Detection and Classification of Active Deformation Areas from PSI Displacement Maps. ISPRS International Journal of Geo-Information 9, 584
Research funding: RETROCLIFF (PID2021-122472NB-100, MCIN/AEI/FEDER, UE) and GEOCANTABRICA (IDE/2024/000753, SEK-25-GRU-GIC-24-072, Principado de Asturias).
How to cite: Cuervas-Mons, J., Domínguez-Cuesta, M. J., Rodríguez-Méndez, N., Rodríguez-Rodríguez, L., Carrillo, J., and Jiménez-Sánchez, M.: Ground deformation monitoring on the asturian coast (NW Spain) using the European Ground Motion Service, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-19128, https://doi.org/10.5194/egusphere-egu26-19128, 2026.
Earth Observation (EO) data play a crucial role in the assessment and management of natural hazards, particularly in post-disaster contexts where rapid, reliable, and spatially consistent information is required to support emergency response and disaster risk reduction strategies. Landslides triggered by major earthquakes represent a typical cascading hazard, often affecting mountainous regions crossed by key infrastructure and occurring under adverse observational conditions such as cloud cover, strong illumination variability, and complex terrain geometry, which limit the effectiveness of conventional optical-based mapping approaches [1].
In this study, we demonstrate the benefit of combining multimodal EO data and vision foundation models for rapid mapping of earthquake-triggered landslides. We exploit the complementary information provided by Sentinel-2 optical imagery and Sentinel-1 Synthetic Aperture Radar (SAR) data within a prompt-free adaptation of the Segment Anything Model (SAM) [2]. The proposed MultiModal SAM (MM-SAM) framework integrates early fusion of optical and SAR observations with a lightweight domain adaptation strategy, enabling the transfer of SAM’s general visual representations to the geohazard mapping domain while keeping most of the pre-trained parameters frozen. This design allows fully automatic, pixel level landslide segmentation with limited labelled data, addressing key limitations of conventional Deep Learning approaches in operational post-disaster scenarios [3].
The approach is evaluated on the 2021 Haiti earthquake case study, that was the focus of a dedicated activation in CEOS Recovery Observatory Demonstrator project [4]. The analysis is conducted using a publicly available multimodal Sentinel-1 and Sentinel-2 dataset specifically developed for earthquake-triggered landslide detection [5]. Result show that the integration significantly enhances mapping robustness under challenging conditions such as cloud cover, complex topography, and heterogeneous surface characteristics. The MM-SAM framework produces accurate and spatially consistent delineation of landslide-affected areas and demonstrates stable performance across independent training, validation, and test subsets.
Overall, this work highlights the added value of multimodal EO data and foundation model-based approaches for scalable and rapid hazard mapping. The proposed MM-SAM framework represents a step toward transferable and operational tools for post-disaster landslide assessment, with potential applications in emergency response, disaster risk reduction strategies, and future multi-hazard monitoring systems.
References
[1] Meng, Shaoqiang, et al. TLSTMF-YOLO: Transfer learning and feature fusion network for earthquake-induced landslide detection in remote sensing images. IEEE Transactions on Geoscience and Remote Sensing, 2025.
[2] A. Kirillov, E. Mintun, N. Ravi, H. Mao, et al., “Segment Anything,” arXiv:2304.02643, 2023.
[3] Yu, Junchuan, et al. Landslidenet: Adaptive Vision Foundation Model for Landslide Detection. In: IGARSS 2024-2024 IEEE International Geoscience and Remote Sensing Symposium. IEEE, 2024. p. 7282-7285.
[4] Tapete, D., et al., “SAR-based scientific products in support to recovery from hurricanes and earthquakes: lessons learnt in Haiti from the CEOS Recovery Observatory pilot to the demonstrator”, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5803, https://doi.org/10.5194/egusphere-egu22-5803, 2022
[5] Bralet Antoine, et al. "Multi-modal Remote Sensing Dataset for Landslide Change Detection in Haiti", IEEE Dataport, July 14, 2024, doi:10.21227/4heb-7h07
How to cite: Di Stasio, P., Tapete, D., Gamba, P., and Ullo, S. L.: Mapping Earthquake-Triggered Landslides Through Multimodal Sentinel-1 and Sentinel-2 Earth Observation Data, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-20527, https://doi.org/10.5194/egusphere-egu26-20527, 2026.
Clay shrink–swell is a major geohazard affecting buildings and infrastructure in France, driven by seasonal soil moisture variations and exacerbated by recurrent droughts. Interferometric Synthetic Aperture Radar (InSAR) time series provide spatially dense measurements of ground deformation, but the robustness of shrink–swell indicators and their relevance for hazard assessment depend on data sources and processing strategies.
This study investigates the contribution of multi-source InSAR Earth Observation data to the characterization of shrink–swell dynamics at scales relevant for hazard mapping. It builds on a detailed urban-scale analysis over the Toulouse metropolitan area, where two independent Sentinel-1 InSAR pipelines (an academic, fully transparent workflow (Flatsim) and an industrial operational product (SatSense)) were harmonised within a common time-series framework. Identical analyses were applied to both datasets, including trend estimation, harmonic modelling of seasonal deformation, sliding-window analysis, clustering, and the derivation of a composite shrink–swell indicator (RGA index).
Although the two pipelines differ substantially in processing strategy and noise characteristics, they retrieve consistent deformation patterns: weak long-term subsidence combined with spatially coherent seasonal signals controlled by clay-rich formations. Differences mainly affect spatial smoothness and noise levels and do not alter hazard-relevant metrics. These results indicate that shrink–swell signals derived from Sentinel-1 time series are robust to processing choices.
Finally, we discuss how this approach can be extended by integrating European Ground Motion Service (EGMS) products and complementary InSAR datasets to support national-scale screening of shrink–swell hazard in France.
How to cite: Le Corvec, N.: Multi-source InSAR Earth Observation for mapping clay shrink–swell hazard in France: from urban-scale time series to risk-relevant indicators, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-21004, https://doi.org/10.5194/egusphere-egu26-21004, 2026.
Remote sensing technologies provide effective means for monitoring and analyzing the environmental impacts of natural hazards. Among Earth observation approaches, optical remote sensing has remained a fundamental data source for decades. In recent years, unmanned aerial vehicles (UAVs), also referred to as drones, have emerged as a flexible and cost-effective platform for acquiring high-resolution geospatial data, including optical imagery, thermal data, and LiDAR point clouds. Owing to their high operational flexibility, UAVs are particularly suitable for collecting first-hand spatial data shortly after disaster events, supporting rapid damage assessment. This study employs UAV-based optical imagery acquired after the Noto earthquake, which occurred on 1 January 2024 in Japan, to support post-disaster geovisualization and spatial analysis. Structure-from-motion (SfM) and multi-view stereo (MVS) techniques are applied to reconstruct three-dimensional (3D) geoinformation from the UAV images. Two key products—a textured triangulated irregular network (TIN) model and an orthophoto—are generated to visualize affected areas and support geospatial analysis. The study focuses on interpreting earthquake-induced damage by integrating 3D models and texture information, with particular emphasis on road damage assessment. Texture features are extracted from orthophotos and represented using indicators derived from the grey-level co-occurrence matrix (GLCM). These texture descriptors, combined with geospatial attributes, are used as inputs to an extreme gradient boosting (XGBoost) model for semi-automatic road damage prediction. The predicted damage results are subsequently correlated with the 3D TIN models to identify locations where road damage is likely to occur. By integrating texture-based analysis with 3D geovisualization, this workflow improves the interpretation of earthquake-related damage across virtual and real-world contexts. The results indicate that UAV-derived optical imagery, combined with machine learning and 3D reconstruction techniques, can support efficient post-disaster damage assessment. This approach enables advanced simulation of decision-making processes and rescue operations, reducing unnecessary costs while improving the effectiveness and timeliness of hazard response.
How to cite: Cheng, M.-L. and Kuramoto, Y.: Earthquake-Induced Terrian and Road Damage Analysis Using UAV-Derived Geospatial and Texture Information, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-4132, https://doi.org/10.5194/egusphere-egu26-4132, 2026.
Thermal anomalies in volcanic hydrothermal systems provide an early and spatially explicit proxy for changes in permeability, fluid pathways, and magmatic–hydrothermal coupling. However, routine volcano monitoring still faces a critical scale gap: ground-based thermometry and gas surveys provide high-quality point data but limited spatial coverage, whereas satellite thermal products often lack the spatial resolution needed to resolve structurally controlled steaming ground and diffuse degassing structures, where many precursory signals localize. This limitation becomes particularly acute at quiescent calderas and rift-related volcanoes, where small-to-moderate thermal changes can occur over metre-scale fracture networks without producing detectable satellite-scale signals. Pantelleria Island (Sicily Channel Rift, Italy), an active volcanic system with persistent fumaroles, steaming ground, and diffuse degassing, represents an ideal natural laboratory to test high-resolution, repeatable thermal monitoring strategies. Here we present a reproducible workflow for radiometric UAV thermal-infrared (TIR) monitoring that converts centimetre-scale orthomosaics into georeferenced products suitable for operational surveillance: (i) surface temperature maps, (ii) heat-flux density rasters, and (iii) volcanic radiative power (VRP) distributions. We acquired calibrated UAV-TIR imagery under calm conditions (low wind; relative humidity ~80%) at altitudes of 60–100 m and processed the data using structure-from-motion photogrammetry to generate co-registered TIR orthomosaics. We then quantified pixel-wise surface heat loss using the ground-surface energy-balance framework widely applied to geothermal terrains (Sekioka–Yuhara approach), combining net longwave radiative emission (Stefan–Boltzmann law) and sensible convective transfer (Newton cooling). Results reveal a strong, multi-megawatt hydrothermal output concentrated within the main fumarolic sector (Q_pos ≈ 8.44 MW over 0.176 km²; core steaming-ground fluxes ~10¹–10² W m⁻²), whereas the second sector exhibits weak, spatially limited anomalies and an order-of-magnitude lower output (Q_pos ≈ 0.206 MW over 0.031 km²). This quantitative contrast supports a permeability-controlled discharge model in which heat and mass transfer focus along discrete upflow pathways and alteration domains, consistent with independent degassing evidence reported for Pantelleria’s hydrothermal areas. By generating operationally usable heat-flux and VRP baselines at the scale of individual vent fields, this approach strengthens volcano monitoring by enabling (i) objective ranking of thermal anomalies, (ii) structural interpretation of upflow pathways, and (iii) time-lapse detection of subtle hydrothermal changes that may precede or accompany unrest. The workflow is readily transferable to other volcanic islands and caldera systems where hydrothermal signals are spatially focused and temporally variable.
How to cite: Pisciotta, A., Battaglia, A., Bellomo, S., D'Alessandro, W., and Müller, D.: Mapping Hydrothermal Heat Output with Radiometric UAV-TIR: A New Workflow for Volcanic Geothermal Targeting (Pantelleria, Italy) , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-7923, https://doi.org/10.5194/egusphere-egu26-7923, 2026.
The acquisition of landslide inventories is the first step in landslide susceptibility assessment. Inventories indicate the geographical coordinates and the morphological and geological characteristics of areas where landslides have occurred, providing essential information for susceptibility analysis. Traditionally, the construction of landslide inventories is performed manually, relying on expert experience and requiring considerable time. Remote sensing techniques offer an alternative for faster mapping through automatic and semi-automatic approaches. Thus, this study evaluates the applicability of a semi-automatic landslide inventory within three susceptibility models: Logistic Regression (LR), Support Vector Machine (SVM), and eXtreme Gradient Boosting (XGBoost). The study area is the Palmital–Gurutuba watershed located in the municipalities of Itaóca and Apiaí, São Paulo State, Brazil. The inventory was constructed following a single extreme rainfall event that triggered landslides on January 15, 2014. The results indicate good applicability of the semi-automatic shallow landslide inventory across all three models. For LR, the AUC-Success and AUC-Prediction were 0.77 and 0.80; for SVM, 0.88 and 0.82; and for XGBoost, 0.94 and 0.85. The Cohen’s Kappa index (k) was employed to evaluate the level of agreement among the susceptibility maps. The results showed an overall mean k value of 0.5; this constitutes a moderate level of agreement. These findings reinforce the potential of semi-automatic landslide inventories as a reliable basis for susceptibility modelling, particularly in scenarios where rapid responses are required after extreme events. Although semi-automatic approaches may still present limitations related to classification errors or the need for expert validation, they substantially reduce the time and effort needed to produce consistent inventories. Their integration with machine learning models demonstrates that, when properly constructed and validated, semi-automatic inventories can effectively support susceptibility assessments and contribute to more efficient hazard mapping and risk management strategies.
How to cite: Dias, H. C., Hölbling, D., and Grohmann, C. H.: Evaluating a semi-automatic landslide inventory for machine learning-based shallow landslide susceptibility assessment, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8030, https://doi.org/10.5194/egusphere-egu26-8030, 2026.
Remote sensing techniques, including optical and Synthetic Aperture Radar (SAR) imagery, offer an effective and rapid method for emergency disaster monitoring, particularly in areas that are difficult to access. In this study, multi-sensor satellite observations from Pleiades, TerraSAR-X, Capella, and Sentinel-1 are utilized to monitor the evolution of a landslide dam formed in Wanrung Township, Hualien, Taiwan, following intense rainfall associated with Tropical Storm Wipha. The landslide was initially detected by seismic monitoring on July 21, 2025. Subsequently, a high-resolution TerraSAR-X image acquired on July 30 revealed a landslide area of approximately 16 ha. Stereo optical images from Pleiades were used to generate a digital surface model (DSM), which enabled the estimation of landslide volume. Additionally, the water volume of the barrier lake was also derived from the lake surface elevation relative to the DSM. As the barrier lake gradually expanded, multi-temporal Pleiades imagery was applied to monitor changes in lake area. In parallel, Interferometric SAR (InSAR) analysis based on deep learning–assisted scatterer selection was conducted using Sentinel-1 data to investigate slope deformation around the landslide body. On September 23, 2025, additional heavy rainfall induced by Typhoon Ragasa caused a rapid rise in water levels, resulting in dam failure and subsequent downstream flooding. Owing to the all-weather, day-and-night imaging capability of active SAR systems, TerraSAR-X and Capella continued to acquire post-event data, providing critical information on embankment and bridge failures. The resulting inundation and sediment deposition affected approximately 382 ha in Guangfu Township. These results demonstrate that integrated multi-sensor satellite observations not only enable rapid tracking of landslide-dam evolution but also provide an operational and transferable monitoring framework covering dam formation, stability assessment, failure detection, and post-event impact evaluation. Such a comprehensive remote sensing strategy is particularly valuable for emergency management under increasing extreme rainfall events and can be applied to future landslide-dam crises worldwide.
How to cite: Tai, Y., Huang, C.-M., Yang, Y.-C., Liu, C.-L., Tseng, K.-H., Tsai, F., and Chang, C.-P.: Rapid Disaster Response to A Landslide Dam Using Multi-Sensor Optical and SAR Satellite Observations in Hualien, Taiwan, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8680, https://doi.org/10.5194/egusphere-egu26-8680, 2026.
Volcanoes are geodynamic systems strictly connected to a wide range of phenomena affecting their surroundings during both eruptive and unrest phases. Among these we focus on the degassing processes from fumarolic fields at Vulcano island. This represents the southernmost island of the Aeolian archipelago (Italy) that was recently affected by an unrest phase started in 2021. In this context, a significant increasing in degassing level around the island and in thermal energy release in the La Fossa cone was observed. Although the unrest period ended in 2022, the observed level of degassing remained higher than those measured prior to the unrest. The aim of this work is to identify the fumarolic field and its temporal evolution before, during and after the unrest phase. To this end, a series of measurement campaigns and fieldwork were conducted using different ground-based instruments, such as field spectroradiometer. Moreover, hyperspectral data were by spaceborne sensors, including ASI-PRISMA and DLR-ENMAP were acquired. In this work, the adopted methodology and obtained results are presented.
How to cite: Spinetti, C., Colini, L., Palombo, A., and Santini, F.: Volcanic unrest at Vulcano Island by hyperspectral remote sensing, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10248, https://doi.org/10.5194/egusphere-egu26-10248, 2026.
Accurate and timely detection of landslides is crucial for effective response strategies. Traditionally, landslide detection has relied on supervised learning models that require extensive labeled datasets or on expert labour to delineate landslide bodies in remote sensing imagery. Our explorative study uses zero-shot learning models, specifically the RemoteCLIP framework, for detecting landslides without the use of task-specific training data. We focused on rainfall-triggered shallow landslides in Slovenia that occurred in August 2023 to test this model’s efficacy.
RemoteCLIP, a variant of the CLIP (Contrastive Language–Image Pre-training) model, leverages visual and textual semantic similarities to classify land surface features based on metadata and environmental cues extracted from remote sensing imagery.
The method involved processing post-event aerial orthoimages with 1 m spatial resolution to identify potential landslides. Raw heat-map outputs from RemoteCLIP were compared against recorded landslide incidents for evaluating detection capabilities. The results revealed that RemoteCLIP effectively identified several major landslide sites, though performance fluctuated with prompt input and terrain complexity.
Our findings highlight the dependence of text-image foundation model performance predominantly on prompt design. Furthermore, they suggest opportunities for model refinement through inclusion of multimodal data inputs, like topographic information, as well as constraining the model output to physically plausible domains.
While being a first step, this research indicates the potential for zero-shot learning to transform landslide detection tasks by reducing dependence on large curated datasets, accelerating deployment in emergency situations, and enhancing responsiveness in data-scarce regions. Through further development, such AI-driven methodologies could provide valuable additions to current geohazard monitoring systems. We propose future work to focus on the integration of multimodal data beyond optical remote-sensing imagery and on the development of frameworks combining zero-shot models with more domain-specific classifiers and physical constraints.
How to cite: Strohmaier, F., Tempel, J., and Brenning, A.: Can we use a Zero-Shot Learning Model for Shallow Landslide Detection? An Application of RemoteCLIP, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-12073, https://doi.org/10.5194/egusphere-egu26-12073, 2026.
Thermal imaging of the Milos volcano (Cyclades, Greece) is used to monitor its active hydrothermal system, specifically focusing on the Eastern part of Milos. Satellite data is essential for tracking Land Surface Temperature (LST) anomalies and radiative heat flux in this dormant but active volcanic field. Milos hosts a well-known shallow geothermal field mainly developed beneath the eastern part of the island. In this work we aim to establish the background surface temperature level for this volcano and observe possible fluctuations related to seasonal effects or changes in the shallow hydrothermal activity.
Thermal sensors Landsat 8 and Landsat 9 (8-day sampling interval) at 100-m resolution during the year 2025 were used for this study. The final dataset contained 65 satellite images, each with cloud and shadow coverage below 40%. Initially, a pixel-based geostatistical analysis was done, where 12 monthly mean LST maps and 12 monthly standard deviation maps were produced to investigate the surface thermal conditions of the island. To mitigate climate change's influence, a further investigation was followed by producing 3 more maps, to detect and locate the accurate annual spatial distribution of valid clear-sky Landsat LST observations, derived by each pixel’s counts over 40oC and its relevance to the normalized annual frequency. The analysis was done completely on Google Earth Engine.
The results showed that the monthly analysis of land surface temperature imagery consistently detects temperatures inside the Zephyria depression (eastern Milos) that are 5–25 °C warmer than the surrounding terrain, which can reach up to 58oC. Additionally, the final analysis succeeded in mitigating the external weather conditions and revealed that 34 observations (out of 65) present a land surface temperature over 40oC inside the Zephyria depression, with a clear spatial correlation to the shallow geothermal field of the island. Another important outcome was that despite the limitations due to atmospheric interference, the limited land-coverage of the island and the small scale of fumaroles onshore Milos, Landsat 8 and Landsat 9 are both able to detect the thermal anomalous pixels by using one year as a referenced period.
How to cite: Peleli, S. and Ganas, A.: Satellite Thermal imaging of the Milos volcano, Cyclades, Greece, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-13686, https://doi.org/10.5194/egusphere-egu26-13686, 2026.
The southern region of Figueira da Foz (western coast of Portugal) faces a severe sediment deficit and high coastal vulnerability, largely due to the port jetties, which have partially blocked the southward longshore sediment transport, causing substantial shoreline retreat. To mitigate these effects, Cova-Gala beach (immediately north of the Figueira da Foz port) has undergone several small scale beach nourishment interventions. In July–August 2025, the Portuguese Environment Agency (APA) carried out the largest intervention in this area, depositing approximately 3.3 million m³ of sand, distributed between the subaerial beach (1.8 million m³) and the shoreface (1.5 million m³).
This study focuses only on the geomorphological and volumetric analysis of the Cova-Gala beach section between groynes 3 and 5 (cells 3-4 and 4-5). In this area, approximately 180,000 m³ of sediment was initially deposited in the dry beach above Chart Datum, representing the maximum retention capacity of the local groyne field. Digital Elevation Models (DEM) were generated from UAV and GNSS-RTK surveys to analyze coastal retreat and volumetric erosion. The analysis focused on the period between August 22, 2025 (post-filling) and January 7, 2026 (post-storm), using GNSS-RTK to ensure positional accuracy and to validate the January data through topographic profiles. This period presented a particular intense succession of storms. In December alone, three major storms occurred (Davide, Emilia and Francis). Under these synoptic conditions, atmospheric pressure ranged from 978 to 994 hPa, with peak winds of 90-124 km/h and hs ranging from 5 to 11 m. Analysis of the UAV data quantified the sediment volume loss to approximately 90,000 m³. Subsequently, surface interpolation in QGIS, comparing the August DEM with the January GNSS survey, determined a loss of approximately 88,000 m³, a value that shows strong convergence with the UAV. To ensure estimate reliability, vertical accuracy was assessed using an independent GNSS-RTK control point. This validation yielded a maximum cumulative error of 5 cm and a Root Mean Square Error (RMSE) of approximately 2.7 cm. The observed dry-beach erosion suggests a rapid cross-shore sediment transport mechanism, moving sand from the beach face/berm to the foreshore/shallow nearshore in response to wave action. Although the subaerial beach experienced significant retreat, partial recovery is expected in the coming months, driven by milder wave action that favors onshore sediment transport. These results demonstrate the effectiveness of UAV photogrammetry for high-precision, rapid assessment of artificial nourishment performance under storm wave conditions.
This work is supported by FCT, I.P./MCTES through a PhD scholarship (2024.03765.BDANA) and national funds through (PIDDAC): LA/P/0068/2020 - https://doi.org/10.54499/LA/P/0068/2020 , UID/50019/2025 and https://doi.org/10.54499/UID/PRR/50019/2025, UID/PRR2/50019/2025. Finally this work is a contribution to project iCoast (project 14796 COMPETE2030-FEDER-00930000).
How to cite: Nunes, A., Costa, P. J. M., Davies, S., and Pinto, C. A.: Morphological Monitoring of Artificial Nourishment at Cova-Gala Beach (western coast of Portugal), EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-14726, https://doi.org/10.5194/egusphere-egu26-14726, 2026.
Geomorphological transformations represent one of the most significant outcomes of high-magnitude flood events, as intense hydraulic forces have the capacity to rapidly reshape river channels, redistribute sediments, and modify the connectivity and functionality of adjacent floodplains. Understanding these processes is crucial for both hazard assessment and sustainable river management. In this context, the present study employs a multi-temporal approach using Unmanned Aerial Systems (UAS) combined with Structure-from-Motion (SfM) photogrammetry to detect, visualize and quantify geomorphological changes induced by flooding along selected sections of the Lilas River, located on Evia Island in Central Greece. These particular river reaches were strongly affected by the extreme flash flood that occurred in August 2020, an event that caused significant geomorphic disruption.
High-resolution aerial surveys were carried out both before the flood event, and shortly thereafter, in June 2018 and in September 2020 respectively. These surveys enabled the generation of highly detailed Digital Surface Models (DSMs) and orthomosaics, with a ground sampling resolution of approximately 2.5 cm. By performing differential analyses of the DSMs, the study was able to capture detailed patterns of erosion and deposition along the river corridor. The results indicate a pronounced spatial variability, with areas of intense erosion exhibiting local vertical lowering exceeding 7 meters, while zones of sediment accumulation showed depositional aggradation of up to approximately 5 meters after corrections for vegetation cover. Such extreme geomorphic changes highlight the uneven distribution of flood-induced forces along the river channel.
One of the most striking findings of the study is the substantial channel widening that occurred in response to the flood. At specific locations, cross-sectional widths expanded by factors ranging from three to nine, primarily as a result of lateral bank erosion. These findings underscore the complex interactions between natural geomorphic processes, extreme hydrological forcing, and anthropogenic landscape modifications, demonstrating that flood impacts cannot be understood without considering the coupled effects of these factors.
Overall, the study illustrates the capability of repeatable UAS–SfM workflows to provide high-resolution, quantitative evidence of flood-driven geomorphic change. Such data are invaluable for supporting post-event assessments, informing river restoration planning, and guiding the design of infrastructure adaptation strategies. Moreover, the results contribute to broader efforts in flood risk management, particularly in Mediterranean catchments that are highly susceptible to extreme weather events. By integrating detailed topographic measurements with hydrological and ecological considerations, the methodology presented here represents a powerful tool for anticipating and mitigating the consequences of future floods.
How to cite: Spyrou, N.-I., Mavroulis, S., Vassilakis, E., Andreadakis, E., Diakakis, M., Stamatakopoulos, P., Kotsi, E., Konsolaki, A., Faliagka, V., Parcharidis, I., and Lekkas, E.: Photogrammetric Analysis of Post-Flood Geomorphological Changes along the Lilas River (Evia Island, Central Greece) Using UAS Data, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-17271, https://doi.org/10.5194/egusphere-egu26-17271, 2026.
This work demonstrates a validation framework applied to InSAR.Hungary, designed to evaluate its results against independently implemented scientific datasets, including other InSAR solutions and GNSS-based deformation fields over Hungary. The InSAR.Hungary validation was conducted with respect to the European Ground Motion Service (EGMS) and the D2200 realization of the EPN Densification, focusing on L3 product-level deformation rates in both the East-West and Vertical directions. The framework enables quantitative and statistical comparison between InSAR.Hungary and the reference datasets, including the use of statistical methods, spatially structured resampling strategies, and hypothesis testing procedures. Its main goal is to determine whether observed differences are attributable to random spatial processes or are influenced by structural bias. While both InSAR.Hungary L3B and EGMS L3 products include low-frequency deformation components derived from least-squares collocation velocity models (which are also based on EPN Densification), we further validated the L3B results against the D2200 realization of EPN Densification to assess how well the L3B components reflect observed GNSS deformation rates. The results are illustrated for qualitative interpretation, while numerical and statistical analyses provide quantitative support for the validation findings.
How to cite: Magyar, B.: Validation framework of InSAR.Hungary using EGMS and EPND: Methodology and Results, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-19474, https://doi.org/10.5194/egusphere-egu26-19474, 2026.
How to cite: Fiorucci, F., Ardizzone, F., Pisano, L., and Cavalli, R. M.: Detecting and monitoring the Shallow Landslides in Agricultural Environments using Google Earth and Sentinel-2 images: two case studies, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-20463, https://doi.org/10.5194/egusphere-egu26-20463, 2026.
Submarine volcanic eruptions are one of the most common yet least observed forms of volcanic activity, as their unpredictability and remote locations limit opportunities for direct observation and monitoring. The observability of surface manifestations depends on vent depth and associated hydrostatic pressure, magma properties, eruptive intensity and surrounding oceanographic conditions that control plume ascent and dispersion.
In shallow water environments these events can have variable manifestations at the sea surface (mostly in the first 500mbsl), ranging from ejection of material and explosions to volcanic plumes and pumice rafts. Among these, discolouration plumes are one of the most frequently observed surface expressions, as the result of the interaction between hot volcanic fluids and cold seawater. Their colour can vary depending on the concentration of certain elements (such as Fe, Al and Si) in the volcanic fluids. Therefore, these plumes change the optical properties of the upper ocean and are often detected by satellite observations.
However, interpretation of these signals is challenged by natural ocean colour variability that can generate optically similar features. Additionally, data limitations related to sensor resolution, cloud contamination, and scarce in-situ observation used for validation, further aggravate these interpretations. This study explores how this type of activity manifests in satellite ocean colour time series across different regions and eruptive events, with the aim of distinguishing volcanic related signals from background variability. An updated global database of documented submarine eruptions was used to extract multi-sensor ocean colour observations and to examine their temporal and spectral behaviour before, during and post eruption. By analysing these changes in sea surface reflectance relative to non-eruptive conditions, we are able to investigate the characteristics and consistency of volcanic ocean colour anomalies and evaluate their detectability within different marine environments.
As a next step, this framework provides a basis for integrating physical oceanographic data, such as temperature and surface currents, to improve discrimination between volcanic discolouration plumes and optically similar features of different origin, including phytoplankton blooms, river discharge and sediment resuspension. This approach will support more robust interpretation of ocean colour anomalies and feed the development of future automated detection, classification and monitoring tools for submarine volcanic eruptions.
How to cite: Domingues, J. R., Ebmeier, S., Suleiman, I., and Martins, A. M.: Interpreting volcanic ocean colour anomalies in a variable ocean, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-22065, https://doi.org/10.5194/egusphere-egu26-22065, 2026.
How to cite: Necula, N. and Niculiță, M.: Ground deformation analysis of the Danube Delta using European Ground Motion Service products, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-17319, https://doi.org/10.5194/egusphere-egu26-17319, 2026.
Peatlands are the most effective terrestrial carbon sinks and play a key role in hydrological regulation and ecosystem services. Their spatial extent, condition and level of degradation are not fully mapped, particularly at a regional scale. This study used Interferometric Synthetic Aperture Radar (InSAR) data extracted from the European Ground Motion Service (EGMS) to characterise the typical vertical surface displacement behaviour of peatland environments. The main goal is to identify peculiar ground displacement signatures to peatlands for refining existing inventories and detecting of previously unclassified peat areas. A two-step unsupervised methodology combining Principal Component Analysis (PCA) and K-means clustering was applied across Great Britain (GB). First, vertical displacement data from EGMS, covering the period from January 2019 to December 2023, were filtered using national land cover datasets in order to remove Measurement Points (MPs) located in urban areas, roads and infrastructure. This filtering allowed the analysis to focus on natural environments where peatlands are expected to occur. PCA was applied to the InSAR time series to reduce its dimensionality and extract the dominant modes of variability in vertical displacement. The resulting principal components were clustered using the K-means algorithm to identify distinct classes of temporal deformation behaviour. Well-documented peatland sites, such as Hatfield Moors, were used as reference areas to interpret and refine the clustering results. Two PCA clusters were identified as representing the typical deformation behaviour of peatlands in GB. This behaviour is characterised by distinct seasonal oscillations related to “breathing” processes in the peat and long-term subsidence trends associated with peat compaction and degradation. The reference behaviour, together with seasonal indicators, was used to screen the full EGMS dataset and identify MPs with similar dynamics. A spatial clustering analysis was then applied to group MPs with high spatial density while excluding isolated or scattered points that are less likely to represent coherent peatland areas. The resulting clusters were then compared with the Copernicus CORINE Land Cover (CLC) 2018 to evaluate their alignment with recognised peatlands and to detect areas potentially affected by peat soils that are not currently classified as peatlands. Cross-correlation analyses were performed between vertical displacement time series and climatic variables, including precipitation, temperature, and a moisture index, in order to validate the identified displacement patterns and investigate their driving mechanisms. These analyses helped to identify the dominant atmospheric controls on peatland seasonality and support the discrimination between different peat types and hydrological conditions. The methodology was validated through three case studies: (i) Hatfield Moors, a well-studied peatbog with ongoing restoration efforts; (ii) New Forest, an extensive peatland complex in southern GB; and (iii) an area not previously classified as peatland but showing comparable behaviour. The results reveal widespread negative vertical displacement across most peatland areas between 2019 and 2023, showing prevailing subsidence and limited rewetting. The study demonstrates the potential of using InSAR data for large-scale peatland monitoring and identification, supporting improved peatland management and restoration strategies also at regional scale.
How to cite: Del Soldato, M., Fibbi, G., Poggi, F., and Medici, C.: Regional scale identification of peatland environments using European Ground Motion Service data, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-19543, https://doi.org/10.5194/egusphere-egu26-19543, 2026.
Intense seismicity started on 27 January 2025 in the Aegean sea about 10 km NE of Santorini and lasted for over a month. This event was preceded by a period of elevated seismicity within the Santorini caldera since September 2024. Continuous GNSS stations on Santorini and neighboring islands as well as Sentinel-1 InSAR acquisitions over Santorini recorded significant deformation during this time period. Ocean-bottom pressure sensors also recorded subsidence after 27 January.
Here, we focus on the inversion of the geodetic data to infer the potential sources behind the deformation. We find a source of inflation at 3.8 km depth within the Santorini caldera between July 2024 and January 2025. Its location matches the location of the source behind the previous unrest in 2011-2012. Between 27 January and end of February, the deformation pattern is found to be consistent with a deflating source at 7.6 km depth below Kolumbo volcano and 13-km long opening dislocation between Kolumbo and Anhydros. Using these results, we were also able to divide this latter episode into smaller time intervals to study the propagation of the opening dislocation upward and to the NE. These results, in combination with the seismicity, lead to the conclusion that there is a coupling between the Santorini and Kolumbo volcanoes and that a dike was intruded in the crust on 27 January, coming from a low velocity anomaly body at 18 km depth.
This study shows that even in difficult settings (deformation occurring underwater with sparse islands around), remote sensing and earth observations can provide essential information to explain an on-going unrest crisis. It is therefore critical to ensure that such data collection is secured onward to help with the understanding of future volcanic and tectonic crisis.
How to cite: Drouin, V., Parks, M., Anastasiou, D., Raptakis, K., Haghighi, M. H., Karstens, J., Isken, M. P., and Nomikou, P.: Using remote sensing and earth observation data to determine the sources behind the 2024-2025 unrest at Santorini and Kolumbo volcanoes., EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-18944, https://doi.org/10.5194/egusphere-egu26-18944, 2026.
Jökulhlaups in wide range of net volume and discharge drain from Mýrdalsjökull ice cap in S-Iceland. The largest ones are related to eruptions of the underlaying volcano Katla and can drain cubic kilometres of flood water with peak discharge on the order of 105 m3 s–1. The last such jökulhlaup occurred during the eruption in 1918. Most of the jökulhlaups from Mýrdalsjökull are however related to geothermal activity beneath ice cauldrons located near the rim of the Katla caldera. These jökulhlaups are much smaller, typically with flood volume between 105 and 107 m3 and peak discharge between few m3 s–1 and few hundred m3 s–1. Larger events, with net flood volume exceeding 107 m3 and peak discharge exceeding 1000 m3 s–1, occurred in 1955, 1999, 2010 and 2024. The source of these jökulhlaups was in all cases beneath known ice cauldrons but what exactly causes them is debated. They have both been attributed to small subglacial eruptions and to powerful events in the geothermal system beneath the glacier. The jökulhlaup in 2024, drained on July 27th from ice cauldrons located in the northeast part of the caldera. These cauldrons had since the start of regular surface elevation monitoring in 1999, been very stable features in the glacier surface. The lack of topographical surface changes indicated insignificant storage of meltwater beneath these cauldrons, likely caused by persistent leakage from them. This approved also with repeated radio echo sounding (RES) profiling carried out annually over these cauldrons since 2012 with the aim of detecting water chambers beneath the cauldrons. Water draining these cauldrons was expected to drain southwards into the river Múlakvísl, but the jökulhlaup in 2024 drained towards east into the river Leirá at the glacier margin, and from there into the river Skálm. The only hydrological monitoring in the flood path were at the bridge passing Skálm on the primary road in S-Iceland. The swift jökulhlaup had already flooded over the bridge and the road next to it before the road had been officially closed. Here we analyse the cause of this jökulhlaup by studying: a) Elevation changes, deduced from Pleiades satellite images, at the jökulhlaup’s source and path both before and during the July 2024 jökulhlaup as well as during the subsequent period of repeated smaller jökulhlaups still ongoing. b) Extensive and detailed glacier bed mapping with RES carried out in the spring 2025 over the jökulhlaup’s source and flood path, as well as comparison with RES profiles measured in this area prior to the 2024 jökulhlaup.
How to cite: Magnússon, E., Pálsson, F., and Belart, J. M. C.: The jökulhlaup Mýrdalsjökull ice cap, S-Iceland, in July 2024. Insights from radio echo sounding and mapped surface elevation changes, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10466, https://doi.org/10.5194/egusphere-egu26-10466, 2026.
The European Ground Motion Service (EGMS) provides continental-scale InSAR ground deformation time series, offering new opportunities for investigating slow-moving geohazards such as landslides. However, the direct use of EGMS products in landslide hazard and susceptibility modelling remains challenging due to the large data volumes, the temporal structure of the time series, and the need to integrate deformation data with environmental and climatic variables.
In this contribution, we present a scalable workflow for transforming EGMS data into analysis-ready inputs for dynamic landslide susceptibility studies. The methodology was developed using GRASS GIS and PostgreSQL/PostGIS, exploiting a high-performance multicore computing infrastructure (tens of CPU cores) to efficiently manage very large datasets while preserving the temporal information required for robust interpretation. To optimise computational performance and validate the robustness of the pipeline, the workflow was first tested on a reduced pilot area and subsequently extended to the entire Province of Salerno (southern Italy), a region characterised by complex geomorphology and widespread slope instability.
EGMS Level-2a ascending Persistent Scatterer displacement time series were imported into GRASS GIS and reorganised into complete time series, resulting in a database exceeding 600 million displacement observations. To reduce data dimensionality while retaining physically meaningful information, Persistent Scatterers were spatially associated with slope units and filtered based on extreme displacement values. The deformation observations were therefore integrated with geomorphological, geological and climatic variables, including hourly precipitation data and surface temperature, aggregated at the slope-unit scale.
The resulting spatio-temporal database provides a consistent and comprehensive foundation for training machine learning models aimed at dynamic landslide susceptibility assessment and future early warning applications. The proposed workflow demonstrates how EGMS products can be systematically transformed into scalable and integrated inputs for regional-scale geohazard analysis.
How to cite: Monte, N., Marchesini, I., Di Martire, D., Reichenbach, P., and Lombardo, L.: A scalable GIS–database workflow for processing EGMS InSAR time series for landslide susceptibility studies, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-11473, https://doi.org/10.5194/egusphere-egu26-11473, 2026.
Posters virtual: Mon, 4 May, 14:00–18:00 | vPoster spot 3
EGU26-3065 | Posters virtual | VPS12
Experimentation with the use of EGMS and IRIDE satellite dataMon, 04 May, 14:00–14:03 (CEST) vPoster spot 3
The objectives of the experiment were the following:
- Evaluation of continuous ground motion from satellite data (European Ground Motion Service for the period 2019-2023 and IRIDE for 2024);
- Analysis of different types of landslides (active landslides, dormant landslides, landslide-prone areas, subsidence);
- Identification of elements for the Emergency Limit Condition (CLE) analysis near areas affected by specific ground motions derived from satellite data in the period 2019-2024
- Identification of buildings in the municipality of Perugia near areas affected by specific ground motions derived from satellite data in the period 2019-2024
- Submission of the results to all regional offices that authorize, evaluate, design, or schedule interventions on the territory and to the regional civil protection agency.
QGIS version 3.42 software was used for the experiment.
The following databases were imported into QGIS:
- European Ground Motion Service satellite data.
- IRIDE satellite data – Cross Monitoring of Ground Motion and "Hot Spots" of Cover Change.
- PAI geomorphological landslide hazard maps.
- Local Seismic Hazard Map of the Umbria Region.
- Umbria Region Emergency Limit Condition Analysis (CLE) maps.
- Geological Map of the Umbria Region.
- Building database of the Umbria Region's land registry system for the Municipality of Perugia.
- Administrative Boundaries of the Umbria Region and base maps such as the Regional Technical Map and Google Satellite.
Spatial analyses were performed using the GIS on the collected data to homogenize and select specific information useful for subsequent processing.
Multiple analyses werw performed for 2 specific case studies.
All objectives were achieved: assessment of continuous ground motion from satellite data (two different databases: IRIDE for 2024 and EGMS 2019-2023); subsequent analysis using different types of landslides (active landslides, dormant landslides, landslide-prone areas, subsidence); subsequent assessment using the CLE (Emergency Limit Condition) elements; subsequent assessment using buildings in the Municipality of Perugia.
How to cite: Motti, A. and Natali, N.: Experimentation with the use of EGMS and IRIDE satellite data, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-3065, https://doi.org/10.5194/egusphere-egu26-3065, 2026.
EGU26-18022 | ECS | Posters virtual | VPS12
Landslide Hazard, Vulnerability, and Risk Analysis (HVRA) Using Machine Learning and AI: A Case Study of the Darma Valley, Kumaun Himalaya, IndiaMon, 04 May, 14:03–14:06 (CEST) vPoster spot 3
Landslides have become one of the most destructive geological hazards in the Himalayan region, exhibiting a significant increase in both occurrence and intensity in recent decades. This increasing trend poses serious threats to human life, infrastructure, and essential public assets, underscoring the need for comprehensive risk evaluation in these highly vulnerable mountainous terrains. The present study offers an extensive assessment of landslide hazard, vulnerability, and associated risk in the Darma Valley of the Kumaun Himalaya, India. Landslide susceptibility was modelled using a Multilayer Perceptron (MLP) neural network, and the model’s predictive performance was validated through ROC–AUC analysis. Vulnerability was quantified by integrating land-use/land-cover categories with their respective economic valuations. Furthermore, rainfall and seismic intensity maps were combined with the susceptibility outputs to derive a detailed landslide hazard map. The results indicate that roads are the most vulnerable elements, followed by settlements and dam infrastructures, largely due to their substantial reconstruction costs and higher exposure levels. The final risk map, produced by integrating hazard and vulnerability layers, reveals that approximately 9% of the study area falls within high to very high risk zones, 22% within moderate risk, 26% within low risk, and 43% within very low risk zones. These findings offer essential guidance for promoting sustainable development and supporting land-use planning that accounts for environmental risks. They also contribute to more informed and effective decision-making aimed at strengthening the resilience of the fragile and sensitive Himalayan landscape.
How to cite: Shawez, M., Kumar, S., Gupta, V., Kumar, P., and Rawat, G.: Landslide Hazard, Vulnerability, and Risk Analysis (HVRA) Using Machine Learning and AI: A Case Study of the Darma Valley, Kumaun Himalaya, India, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-18022, https://doi.org/10.5194/egusphere-egu26-18022, 2026.
