Earth Observation (EO) data are increasingly accessible beyond specialist communities, enabling earlier detection of environmental crimes and more timely interventions. Public online platforms that publish geospatial products—such as global deforestation, fisheries, and mining trackers—offer standardized datasets with regular updates, but they frequently lack the spatial detail, temporal resolution, or contextual layers needed to inform national enforcement and policy responses. At the same time, many countries hold rich, domain‑specific datasets (concessions, environmental monitoring, socio‑economic records) that remain siloed, inconsistently formatted, or not tailored to detect illegal activities. Closing this gap requires coordinated planning across national agencies to share, harmonize, and process data so that evidence‑based, operational information can guide integrated responses. The United Nations Office on Drugs and Crime (UNODC) develops customized applications and capacity‑building programs that combine Sentinel series imagery with very‑high‑resolution commercial images and national datasets to monitor illegal mining and related crimes. UNODC has been working with authorities in Indonesia and Ecuador, and expanding efforts across Amazon countries, illustrates practical approaches to integrating global EO products with local data. Partnerships with the European Space Agency and private sector imagery providers have supported tailored workflows and technical exchanges. This presentation will showcase innovative geospatial workflows that deliver timely indicators of new mining activity, summarize lessons learned from deployments with national partners, and highlight technical and institutional challenges encountered. Emphasis will be placed on practical data‑integration strategies, and building national capacity to produce a comprehensive geospatial picture of illegal activities to support enforcement and policy decision‑making.

The Office of the Auditor General of Norway published a performance audit in April 2026 on mapping of nature values and climate impacts as a basis for decision-making in land-use planning (Document 3:9 2025–2026). The audit addressed two questions: (1) the extent of nature loss in Norway due to land conversion, and (2) whether existing mapping of nature values and climate impacts at national and local level provides a sufficient knowledge base for sustainable land-use decisions. Our presentation will focus on the methodology developed to address the first question. Prior to the audit, national estimates of nature loss existed, but only at an aggregate level. To provide more detailed and audit-relevant evidence, we developed a spatial methodology combining national geospatial datasets on land-use change with detailed nature and biodiversity data. This made it possible not only to quantify land conversion, but also to identify which types of nature were affected, including red-listed nature types and other ecologically valuable habitats. The approach demonstrates how geospatial data can be transformed into audit evidence through transparent assumptions, validation, and reproducible workflows. While developed in the context of a performance audit rather than environmental crime investigations, the methodology may also be relevant in contexts where spatial evidence is needed to identify and document environmental impacts of illegal or harmful land-use activities.

The ENFORCE project (Horizon Europe) is developing the ENFORCE Plaza, a pan-European platform integrating citizen science, earth observation (EO) data and services, and AI to support environmental compliance assurance across eight case studies addressing different challenges covering illegal fires, waste dumping, deforestation, marine litter, and soil and water contamination. This presentation focuses on efforts to integrate satellite derived data and products into the Plaza's Tools component, which combines a Data Acquisition & Ingestion layer with a Data Governance & Processing layer to harmonise satellite-derived indicators alongside in-situ sensor data and citizen reports. Within this framework, data processing and analytical methods, including machine learning approaches that combine EO-derived indicators, in-situ sensors, and citizen reports, are being developed to support environmental compliance monitoring. We present the state of the Plaza dashboard, including how EO-derived variables are utilised to detect critical environmental condition changes (e.g., fire and burn-area, ocean colour anomalies and river pollutant levels) and displayed alongside case-specific reports and authoritative datasets, supporting situational awareness and case management. The platform is designed around distinct user personas, each with tailored access to reports, evidence, and case dashboards, reflecting different evidentiary and operational needs across the compliance chain from initial detection to potential judicial use. We will introduce the Data Readiness Level (DRL) framework, an ENFORCE concept assessing data maturity across dimensions that integrate data science, scienyific validity and legal admissibility for the use of participatory data for environmental compliance, currently we are in the conceptualization phase. The framework help determine whether EO derived and citizen-generated information is sufficiently documented, validated and traceable to support regulatory or judicial use. As the DRL is still being developed, it will be presented as a forward-looking direction rather than a validated result. The presentation aims to share practical lessons on EO integration into operational, multi-stakeholder compliance workflows.

Although legislation, international regimes and respective local policies to manage environmental resources are largely in place, environmental compliance assurance (ECA) efforts continue to fall short in terms of data availability, awareness and action. Citizen science (CS) initiatives, through the citizen-generated data (CGD) they produce, offer innovative Earth Observation sources alongside new forms of joint data and knowledge production. However, deepening the role of citizens and communities in ECA - and positioning them as active partners in combating environmental crimes - constitutes a transformative change that will not occur spontaneously. It requires more than technical alignment: it demands visioning and experimenting with new roles and interactions to open up possibilities for collaboration not previously imagined, and to explore what operational readiness for CGD actually means in practice. To catalyse such transformative change in current ECA practices, the EU co-funded more4nature project assists 40 cases in Europe and beyond with a process of social learning and socio-technical experimentation, combined with the provision of algorithms, models, and interoperable systems to public authorities. In each case, a facilitated learning journey supports CS initiatives and institutional actors in each case to advance towards meaningful levels of collaboration. This process is grounded in a comprehensive needs assessment that integrates behavioural, technical, and socio-political dimensions, generating a nuanced understanding of which mode of ECA is currently practiced - and envisaged, as well as the factors that enable and inhibit effective implementation of collaborative ECA. Targeted efforts are then undertaken to strengthen the technical, social and behavioural capacity of civil society and institutional actors to engage in new forms of collaboration, including - but not limited to - the effective integration and uptake of CGD in ECA. This presentation concludes with lessons learned and recommendations from the more4nature cases - spanning the domains of related to zero pollution, biodiversity protection and deforestation prevention - on how to strengthen the capacities of CS initiatives and institutional actors to enable them join forces combatting environmental crimes through collaborative ECA.

Compliance assurance, the promotion, monitoring and enforcement of adherence to EU law, increasingly depends on verifiable, repeatable and spatially explicit evidence. The Knowledge Centre on Earth Observation (KCEO) of the European Commission has completed its third Deep Dive assessment, Earth Observation in Support of Compliance Assurance across EU Policies (to be published on 1 September 2026), examining how Copernicus and complementary Earth Observation (EO) resources support this function along the full chain from detection to enforcement. Co-designed with four policy Directorates-General (DG ENV, DG CLIMA, DG ENER and DG GROW) and drawing on contributions from some 140 experts, including DG DEFIS, the Copernicus Entrusted Entities and external specialists, the assessment applies a fitness-for-purpose methodology comparing policy needs against product specifications on four criteria: spatial resolution, observation cadence, temporal extent and spatial extent. It covers five established application areas where EO has reached operational maturity (the common agricultural policy, maritime surveillance and marine environmental monitoring, security and environmental-crime investigations, terrestrial ecosystems under the Nature Directives and the EU Deforestation Regulation, and land use emissions reporting under LULUCF) and six emerging use cases addressing building thermal performance, urban greenhouse gas emissions and air quality, energy-sector emissions, peatland restoration, water quality and critical raw materials, while cross-cutting chapters examine data quality, legal admissibility, operational deployment and the integration of citizen and NGO reporting. EO is operational, not experimental: remote sensing already represents around 80 % of the CAP's area-based checks in many Member States, and satellite-based evidence has been accepted by the European Court of Justice in infringement proceedings. Across the emerging use cases, however, products mostly return partial matches: the building blocks largely exist but few reach compliance practitioners in usable form, and the binding constraint is rarely sensor resolution alone. Documented uncertainty, legal admissibility and institutional capacity increasingly decide whether EO evidence can be used; only around 5 % of Copernicus Land Monitoring Service datasets carry documented uncertainty layers of the kind needed for legal use. The presentation will set out the recommendations along three pathways: direct implementation and last-mile applications, Copernicus programme evolution, and priority research, including legal-admissibility standards.

Environmental crimes are often spatially dispersed, time-sensitive and difficult to investigate through single-source information alone. Illegal dumping, hazardous waste handling, illegal burning or pollution incidents may leave both physical traces on the ground and digital traces in open-source information streams. This contribution presents a transferable intelligence framework developed within the ESA EO_LAW context, where Earth Observation-derived information, ICT analytics and non-EO data fusion were demonstrated across law-enforcement domains, including environmental crimes, and terrorism/organized crime. The framework couples Earth Observation image intelligence (IMINT), open-source intelligence (OSINT), geospatial analysis and AI-assisted analytics to support monitoring, prioritisation and operational decision-making. The workflow includes four main steps: extraction of activity localisations from media mining and OSINT processes; detection of anthropogenic changes from EO data; spatial analysis using geospatial datasets; and the linking between EO-derived change points and OSINT-derived activity points. Although originally demonstrated in security and organized-crime applications, the approach is directly applicable to environmental challenges. In particular, for a case study on terrorist groups’ operations in Burkina Faso, the focus was on the WAP complex of natural reserves. The integration of OSINT analysis with EO and other geospatial data (mainly road networks and buildings) identified potential hideout locations within the complex. Inspection of such locations (using interpretation of Very-High resolution Imagery) further highlighted and localized a well-known impact of fires in the region: turning parks and resorts into abandoned areas that may serve terrorist groups’ safe heavens. Rather than treating satellite products as standalone evidence, the framework demonstrates how multi-source intelligence can help authorities screen large areas, prioritise suspicious locations and prepare targeted inspections. The approach can strengthen both monitoring and enforcement functions, especially where environmental crime overlaps with security risks, weak territorial control or rapidly changing land-use conditions.

The State Audit Institution of Montenegro presents the results of its cooperation with the European Space Agency under the MNE Illegal Deforestation project, implemented with Geodata (Hungary). The initiative explored how Earth Observation data can strengthen environmental auditing and support national authorities in addressing illegal forest logging and cross-border timber trafficking. Satellite-derived information was used to identify and assess changes in forest cover, highlight areas potentially affected by unauthorised logging, and provide independent evidence to support audit findings and institutional follow-up. By combining geospatial analysis with administrative and regulatory information, the project demonstrated how environmental audits can move beyond retrospective compliance checks and contribute to more targeted and timely enforcement action. The activity generated concrete operational outcomes. The case illustrates the practical value of Earth Observation as an accountability and enforcement tool, while also highlighting the importance of cooperation among audit institutions, technical experts and national authorities. It offers a replicable model for using satellite-based evidence to strengthen forest governance and counter environmental crime.

Advances in geospatial data and tools produced through the World Resources Institute’s (WRI) Global Nature Watch (GNW, formerly Global Forest Watch) have supported a wide variety of stakeholders in the fight against forest related environmental crimes. WRI data supports new laws such as EU Deforestation Regulation which ensures that certain products which are placed on the EU market do not contribute to deforestation and forest degradation, and are produced legally. We will showcase key datasets such as the new Google-WRI Global Forest Types data (1), produced in support of EUDR deforestation and degradation assessments (2, 3). We will also explain how WRI data on forest concessions, and forest disturbance information are currently being used the police, food authorities and financial institutions as well as companies in Europe for timber legality assessments. For many years, geospatial data and tools from WRI have also been used directly by concerned groups around the world to protect their land from illegal activities. We will demonstrate how WRI’s near-real-time integrated deforestation alerts, provide quick and confident information on forest disturbances (4) and related tools such as the Forest Watcher mobile phone app translate to legal action and successful outcomes for forest protection – for example arrests and seizures related to habitat destruction in the Massai Mara (5). We will showcase recent guidance and case studies use of near-real-time forest disturbance data from around the world (6). Finally, we will reflect on the use of AI in the new WRI tool Global Nature Watch Horizon to support the fight against environmental forest crimes: https://landcarbonlab.org/ai-global-nature-monitoring/ Neumann et al. 2026 Global forest typology at 10-meter resolution for forest and land-use monitoring Carter et al. www.globalforestwatch.org/blog/data/satellite-data-eu-regulation-deforestation-free-supply-chains/ Carter et al. www.globalforestwatch.org/blog/data-and-tools/monitoring-forest-degradation-eudr Reiche et al. 2024 https://iopscience.iop.org/article/10.1088/1748-9326/ad2d82#erlad2d82s4 Nussbaum 2023 https://www.globalforestwatch.org/blog/users-in-action/protecting-forest-habitats-mara-elephant-project/ Carter et al. 2024 https://www.reddcompass.org/mgd/resources/GFOI-DeforestationAlertModule-20240909.pdf

Illegal logging and deforestation rank among the fastest-growing environmental crimes, yet they remain notoriously hard to detect. Much of the activity, such as timber extraction, machinery movement and sub-canopy clearing, occurs beneath the forest canopy, beyond the reach of optical and high-frequency radar sensors, and often across remote borders. This contribution presents ongoing work on low-frequency, long-wavelength Synthetic Aperture Radar (SAR) for Foliage Penetration (FOPEN) object and change detection, applied to environmental compliance monitoring. Using L-band SAOCOM imagery cross-referenced with 50 cm optical data for labelling, we built an end-to-end, open-source processing pipeline and evaluated a spectrum of change-detection methods, including coherent change detection, CFAR, and robust-PCA approaches alongside supervised and self-supervised deep learning (e.g., YOLO-Siamese, Metric-CD). During the inception phase we specified operational use cases spanning illegal logging and forestry, deforestation, and cross-border security. Results show that long-wavelength SAR can reveal sub-canopy timber extraction, forest-cover loss, and concealed vehicles and machinery that are invisible to conventional sensors, while exposing practical challenges — temporal decorrelation and vegetation clutter — that drive false alarms. We report validation across supervised and unsupervised metrics (F1, Precision, Recall, AUC) and release curated benchmark datasets and labelling tools to support reproducibility and wider uptake. Looking to the NISAR and ROSE-L missions, we outline an operationalization roadmap for near-continuous, all-weather forest monitoring — directly supporting the workshop's aim of integrating Earth Observation into evidence-based, data-driven detection and characterisation of environmental crime.

The effective implementation of geolocation-based biodiversity monitoring to ensure compliance with environmental requirements is often hampered by technical complexity, excessive data volumes, and fragmented institutional workflows. To overcome these obstacles, we have developed and tested a comprehensive capacity-building ecosystem specifically designed for national water management agencies to monitor compliance with legislation on the protection of water bodies. This ecosystem consists of: (i) optimized data exchange between the scientific community and law enforcement; (ii) a modular training program that illustrates ecological interrelationships; and (iii) the creation of a platform for sharing best practices. This program has been implemented at the Skolivski Beskydy National Nature Park. We plan to scale it up to all water bodies. We also aim to learn how to use artificial intelligence for data processing, apply methods for assessing damage, and automatically calculate the extent of the damage caused, which underscores the importance of criminal proceedings in this category of cases.

Environmental compliance assurance relies on geospatial intelligence to promote laws, monitor risks, and prove breaches. However, integrating cross-sectoral spatial data remains challenging and compliance professionals sometimes struggle to fully leverage complex geodata to in risk monitoring and evidence collection Using the EU’s Habitats Directive—specifically protected forest habitats vulnerable to logging—we demonstrate an R-software package to streamline this process. The tool combines authoritative habitat maps with satellite-derived tree cover loss data to detect hazards where forests are compromised. Through an interactive web interface, users can filter tree cover loss by area, habitat type, and timeframe to focus on major compliance risks. The system then offers two complementary views: Regional Assessment: Summarizes findings across whole areas using interactive charts. Local Assessment: Allows detailed spatial mapping alongside satellite imagery to plan ground interventions. Compatible with SpatioTemporal Asset Catalogs (STAC), the platform can integrate near-real-time satellite updates and airborne LiDAR. The software package allows users to upload their own data to assess compliance risk, be it in protected habitats, or other areas.

Illegal timber trade is the most profitable natural resource crime, costing 30-157billion USD a year globally. Complex supply chains obscure timber sources, which often leavestraders’ declaration of origin as the sole evidence of provenance, despite the possibility of fraud.Recently, there has been a growing interest in using chemical markers, such as Stable IsotopeRatio Analysis or Trace Elements to verify origin claims by matching levels ofmarkers within wood tissue to location-specific values predicted from reference data.Unfortunately, chemical tracing is currently limited by the need for large reference data sets from all species of interest and by not taking advantage of other types of relevant data.We present two approaches for improving the scalability and accuracy of chemical timber tracing. First, we introduce a model that integrates data from multiple tree genera while accounting for statistical differences between them. Using a dataset of 1020 samples from four economically important genera (Betula, Fagus, Pinus, Quercus) across 12 Eastern European countries, we show that it is possible to accurately predict the harvest location even when relevant data are unavailable in some areas, provided that data from other genera are available. Our approach could lower reference sampling costs and enable tracing in situations where new samples cannot be collected, such as during armed conflict.Second, we explore how existing models can be improved by incorporating environmental constraints that determine whether a location is ecologically viable or operationally feasible for commercial harvesting. This is achieved using open-access geospatial datasets of key environmental attributes—land cover, topography, and tree cover fraction. For 323 eastern European birch samples, we found that incorporating land-cover restrictions can halve the harvestable range within the predicted harvest area without compromising prediction accuracy. Our results demonstrate the value of open-access datasets in advancing rigorous methods for timber traceability.

Illegal artisanal and small-scale gold mining (“galamsey”) in Ghana exemplifies a rapidly evolving environmental crime and supply-chain due-diligence challenge. Galamsey drives deforestation, encroachment into forest reserves, sedimentation and mercury-related water pollution, while mined gold may enter formal trade through opaque intermediaries. Documentary audits alone are therefore insufficient: they need to be complemented by independent, spatially explicit evidence on where extraction is likely occurring, how it changes over time, and whether reported sourcing information is plausible. This contribution presents an emerging GeoCompliance workflow developed at the University of Salzburg’s Department of Geoinformatics – Z_GIS. The approach fuses multi-temporal Earth Observation, geospatial context layers and open-source information to generate auditable risk signals for environmental compliance and enforcement support. For the Ghana use case, Sentinel-2 time series are used to detect vegetation loss, exposed soil and mining-related land-cover transitions; Sentinel-1 SAR backscatter dynamics are explored to improve monitoring under persistent cloud cover. EO-derived activity cues are combined with protected-area boundaries, river buffers, cocoa-growing regions, administrative units, conflict-affected and high-risk area logic, and selected media/OSINT signals. The resulting “plausibility engine” flags spatial inconsistencies between claimed supply-chain information and observable mining activity, rather than claiming to replace legal investigation or field verification. A core design principle is information fusion with traceable provenance: each risk flag should be explainable, reproducible and accompanied by uncertainty information. Validation is envisaged through stratified expert review, comparison with high-resolution reference imagery, local partner knowledge and targeted ground/desk verification. By linking galamsey hotspot detection with supply-chain plausibility checks, the workflow aims to support earlier screening, prioritisation of inspections, and evidence-based dialogue among enforcement bodies, regulators, companies and local partners.

Illegal gold mining, particularly artisanal and small-scale gold mining (ASGM), presents a severe environmental and public health crisis across gold-bearing regions of the Amazon basin. A major driver of this damage is the use of mercury, which contaminates local watersheds and ecosystems. In areas such as French Guiana, monitoring these operations is extraordinarily complex due to their small scale, illegal nature, and remote locations deep within dense rainforests. Traditional remote sensing techniques are constrained by persistent cloud cover, which severely limits optical satellite imaging, while the small-scale partial deforestation associated with these sites can hardly be detected using standard SAR images such as Sentinel-1, which lacks the required spatial resolution. To address these monitoring gaps, this study demonstrates a novel approach utilizing Single Look Complex (SLC) images from wide-swath altimetry missions, such as the current CNES/NASA Surface Water and Ocean Topography (SWOT) mission and the future Sentinel-3 Next Generation Topography (S3NG-Topo). These platforms use near-nadir Ka-band SAR imaging in a bistatic across-track configuration to estimate surface water elevation through interferometric processing. We exploit the very high reflectivity of water surfaces in these images to detect the characteristic water-filled excavated pits generated during gold extraction. The proposed framework implements a multi-temporal image processing pipeline combined with specific denoising algorithms designed for Ka-band interferometric images. This method highlights areas with temporal changes in near-nadir Ka-band reflectivity, which often indicate ongoing gold mining activities. Consequently, previously unobservable small-scale gold mining operations can now be detected. This approach offers law enforcement and environmental agencies combating illegal gold mining a targeted tool to prioritize their efforts regarding the tasking of very high-resolution satellite imagery and on-the-ground interventions.

Sand and gravel are basic raw materials used in construction. Due to transportation costs, they are not moved over long distances. Where geological conditions allow, sand and gravel are extracted in the vicinity of investments. Obtaining a mining licence is always tied to conditions for land reclamation after the end of operation. The area should be restored to its pre-mining state, which means that the pit should be backfilled. Clean materials that pose no threat to the environment should be used for this purpose. Additional license conditions apply to situations in which water appears in the pit. Due to the possibility of groundwater contamination, exploitation must usually be discontinued. According to environmental inspection authorities and local governments, the greatest environmental hazards are during the reclamation phase. Backfilling a pit with a capacity of several hundred thousand cubic meters using clean materials is expensive. As a result, cases of filling excavations with municipal waste and other materials that threaten the environment are revealed. Prohibited materials are most often transported at night and, once dumped into the pit, are covered with a layer of clean soil. Identifying the type of material used to fill excavations from Sentinel-2 (S-2) images is a very difficult task. However, S-2 can be used to monitor water in excavation pits, determining the time when water appears and when the pit is filled in. For this purpose, a modified automatic land cover classification method S2GLC (Sentinel-2 Global Land Cover), was used. Results of change detection combined with the terms of the sand or gravel mining licence can indicate an environmental crime. The research was conducted in the Masovian Voivodeship (Central Poland).

Large-scale soil movements are difficult to monitor using conventional inspection methods alone. In the Netherlands, more than half a million soil movement activities take place each year, making field inspections impractical. TerraVisie is an operational Earth Observation information service that supports environmental inspectors by providing continuous satellite-based monitoring of soil movement activities. The service combines high-resolution optical imagery and radar data, and uses automated change detection, and geospatial analytics to identify areas where significant changes have occurred. Rather than replacing field inspections, TerraVisie enables inspectors to focus their efforts on locations with a higher likelihood of non-compliance, improving the efficiency and effectiveness of enforcement activities. TerraVisie provides a historical image archive, providing monthly high-resolution imagery that allows inspectors to look back in time and reconstruct the evolution of a site. This historical context supports investigations by revealing when changes occurred and providing visual evidence that is often unavailable during on-site inspections after activities have been completed. The presentation will demonstrate the value added by TerraVisie for the operational activities of Dutch enforcement authorities for environmental crime investigations, remote validation of reported activities, and intelligence-driven inspection planning. It will illustrate how operational Earth Observation services can strengthen environmental enforcement by combining historical satellite observations with automated change detection to deliver actionable information for inspectors.

In the EU, soils are under significant pressure from contamination, with growing concern associated with improper disposal of waste and harmful substances and the displacement of polluted soil masses - activities that frequently fall within the scope of environmental crime. While satellite technologies have undeniable advantage for large scale continuous monitoring, recent development of more lightweight and compact sensors opened new possibilities for UAV-based surveying, which can serve as complementary means for data collection and verification. Interest in airborne hyperspectral (HS) sensing applications has been growing rapidly, as it enables more detailed spectral analysis of areas of interest. On the other hand, Gamma-ray sensing (GRS) is a niche technique developed mainly for radiological and geological applications, that has just begun to find uses in other domains. Combining several modalities such as GRS and HS may offer new opportunities for detecting contamination and other anomalies in the topsoil layer. We present the first results of exploratory research aimed at detecting several types of 'foreign' material (construction debris, household waste mix, etc.) on and just below the earth's surface. A small-scale experimental setup was realised under real conditions: samples were placed at three depths relative to the soil surface, and repeated measurements were taken over time using Medusa MS-350 GRS sensor (handheld mode) and Cubert Ultris X20 Pro HS sensor in the wavelength range of 350-1000 nm (UAV mode), with an addition of ground-penetrating radar (GPR) data. The presentation outlines the first analytical findings on the spectral and radiometric 'fingerprints' of the tested materials, discusses the limitations of each modality individually, and provides an outlook on the potential for data fusion across these sensors in the context of detection of illegal dumpings, unauthorised burial of hazardous materials or concealment of polluted soil.

AMA presents UCRONIA, Rome’s Digital Twin for the Circular Economy, developed to support environmental monitoring, operational planning and evidence-based enforcement. The platform integrates more than 500 million geospatial records, with satellite data playing a central role in identifying environmental risks and potentially illegal activities. This advanced integration stems from the Memorandum of Understanding signed by AMA and the European Space Agency (ESA) in July 2025. The cooperation enables four use cases within UCRONIA: detecting and estimating illegal waste accumulations through very-high-resolution multispectral imagery; mapping urban heat islands using Sentinel-2 data to improve operational planning and mitigate climate impacts; monitoring ground deformation through Synthetic Aperture Radar data to protect critical infrastructure and more than 100 AMA properties; and assessing urban vegetation conditions using Sentinel-2 surface reflectance data. Particular attention is given to illegal waste dumping, where space-based observations help identify suspicious sites, prioritise inspections and support field verification. By combining satellite-derived evidence, predictive monitoring and targeted on-site checks, UCRONIA demonstrates how Earth Observation can strengthen urban environmental governance, improve enforcement capacity and provide a scalable model for preventing and addressing environmental offences.

Within the framework of ESA EO4SECURITY Environamental Crimes project, four use cases were previously developed in cooperation with International organizations including UNODC and Interpol. Building on these results a new project phase focused on developing customized EO information products to support operational and strategic decision-making of AMA Roma S.p.A. one of Europe's largest municipal waste management companies. The project focuses on addressing environmental crimes and urban vulnerabilities through advanced analytics, specifically targeting illegal waste discharge, urban heat islands, and ground instability affecting critical infrastructure. Illegal waste disposal is one of the fastest-growing environmental crimes in Europe, generating significant environmental, economic, and public health impacts while posing substantial challenges for enforcement authorities. Effective monitoring requires timely, objective, and repeatable information over large territories, capabilities that are increasingly enabled by Earth Observation (EO). The project resulting products, including illegal waste maps, historical analysis, periodic updates of heat indicators and ground motion. These products were delivered through WMS/WFS services to enable seamless integration into AMA's operational U.C.R.O.N.I.A. platform, enabling their direct use within routine environmental management activities and to merge with other geospatial layers, as drone or ground-based derived analytics. The project demonstrates how EO can evolve beyond research applications to become an effective operational component for the end users of environmental compliance workflows, and how EO-derived geospatial intelligence can strengthen environmental compliance by supporting the prioritization of field inspections, enabling systematic monitoring of extensive urban areas, and providing objective evidence of environmental changes over time. The operational experience highlights the importance of combining automated EO processing with expert validation and close collaboration between technology providers and end users to facilitate the operational adoption of EO services within public-sector environmental management. The proposed approach represents a scalable model for supporting environmental monitoring and future environmental crime investigations across European municipalities

Illegal waste dumping is among the most widespread and economically damaging categories of environmental crime in Europe. Nevertheless, dumping site detection still relies on manual photo-interpretation of Earth Observation imagery, a time-consuming, expertise-intensive process that limits the territory coverage achievable by environmental protection agencies. This work presents an integrated platform based on satellite imagery for analyzing waste sites, developed as an operational screening and diagnostic tool to support the detection and characterization of illegally dumped waste and the assessment of its potential environmental impact. The system is composed of three components. First, a Deep Learning (DL) waste detector processes very-high-resolution optical imagery and returns, for each analyzed tile, a confidence score and an interpretable saliency map localizing the waste. The detector was trained and validated on AerialWaste, an open dataset of professionally annotated sites (Torres & Fraternali, Scientific Data, 2023). In the published benchmark, the developed DL model attained 92.0% F1-score and 94.6% accuracy (Gibellini et al., Waste Management Bulletin, 2025). Second, a risk-estimation module fuses geospatial datasets, such as proximity to water bodies, protected areas, and population, to prioritize detected sites by their environmental severity. Third, a unified database of waste-treatment authorizations enables automated cross-referencing of detected sites against authorized facilities, supporting permit verification and the flagging of potentially unauthorized activity for further investigation. The proposed pipeline, delivered as a QGIS plugin, converts wide-scale imagery into a ranked, auditable shortlist of high-priority waste sites. Rather than manually scanning vast territories, agency operators can concentrate their expertise on validating and characterizing the highest-priority candidates, significantly reducing the effort required to move from broad-area screening to field-ready leads. The benefits of this approach include a drastic increase in the area that can be monitored with the existing staff and improved traceability of the evidence collected.

In Lombardy, the land application of livestock effluents for agronomic purposes is regulated by regional policies, which mandate compliance with specific operational and management requirements, alongside a prohibition during designated winter periods. This ban aims to mitigate groundwater pollution risks arising from the biological dormancy of crops, which restricts the uptake of nitrogen and phosphorus, while concurrently preventing atmospheric ammonia and secondary PM10 emissions. Due to the high density of intensive livestock farming in the Po Valley, approximately 800 facilities operate under the Integrated Environmental Authorization (IPPC code 6.6, Directive 2010/75/EU). These installations are subject to periodic oversight by ARPA, supported by the operational use of Earth Observation (EO) tools from various platforms. Although ARPA lacks direct statutory competence over smaller, non-IPPC farms, the Agency frequently conducts targeted inspections into alleged irregular spreading activities triggered by citizen complaints or requests from Judicial Police bodies. To overcome the logistical constraints of traditional on-site inspections, ARPA has implemented a remote monitoring system utilizing high-temporal-frequency satellite imagery. This system tracks environmental criticalities and provides robust evidence of illicit spreading operations. In addition to satellite data, Unmanned Aerial Vehicles (UAVs/drones) are also deployed to monitor livestock facilities, particularly when sites are inaccessible due to sanitary restrictions. This paper presents ARPA’s practical operational experience in integrating Earth Observation data to monitor livestock effluent spreading and farming activities alongside conventional field inspections, demonstrating how these technologies significantly enhance the efficiency, scope, and comprehensiveness of environmental enforcement.

Excessive or improper use of nitrogen fertilizers is a major driver of diffuse water pollution, eutrophication, and environmental degradation. Monitoring compliance with nitrogen management regulations remains challenging, as enforcement authorities largely rely on farmer declarations and costly field inspections that provide limited spatial and temporal coverage. This work presents a novel Earth Observation service for the systematic detection of nitrogen fertilizer applications using Sentinel-2 imagery. The service processes Sentinel-2 acquisitions in near real time to identify fertilized areas and delivers the results through CGI's Insula cloud-based platform, where users can visualize, download, and integrate the generated maps with environmental, administrative, and agricultural datasets for further analysis. The underlying methodology has been validated over the Italian territory, demonstrating strong performance in detecting nitrogen fertilizer application events, and is expected to be transferable across Europe with little or no adaptation. Future developments will further extend the service by identifying the most likely type of nitrogen fertilizer applied. The service systematically provides independent, large-scale evidence of N-fertilizer applications without requiring field inspections, ground sensor networks, or farmer self-reporting. By combining fertilization maps with environmental constraints, protected areas, and water vulnerability layers, authorities can identify potential hotspots of non-compliance, prioritize inspections, and allocate limited enforcement resources more effectively. This evidence-based approach strengthens compliance assurance while reducing operational costs and staffing requirements. The generated information can also support environmental monitoring and sustainability reporting. Finally, the proposed methodology enables a shift from reactive, sample-based inspections to proactive, evidence-based monitoring. It demonstrates how Copernicus Sentinel-2 data can strengthen environmental protection by improving the efficiency, scalability, and cost-effectiveness of compliance monitoring and enforcement.

Liquid discharges from wastewater treatment plants, industrial activities, agricultural runoff and accidental events can significantly impact water quality, aquatic ecosystems and public health. Timely detection supports competent authorities in identifying pollution sources, prioritising field inspections, strengthening environmental legislation enforcement and enabling data-driven decision making. Earth Observation (EO) provides a cost-effective approach for large-scale monitoring by detecting optical and thermal anomalies associated with water pollution events, contributing to SDG 6 (Clean Water and Sanitation) and SDG 3 (Good Health and Well-being).The Multisensor Water Discharges project, developed within the ESA Sentinel User Preparation Initiative, prepares the operational exploitation of the future Copernicus Expansion Missions CHIME and LSTM for monitoring liquid discharges in inland and coastal waters. While these missions are not yet operational, the project combines simulated hyperspectral and thermal observations with EO services supporting environmental surveillance, regulatory compliance and evidence-based policy, complementing the Copernicus Marine and Land Monitoring Services.A key element of the project is the active engagement of competent public authorities responsible for environmental protection, water quality monitoring and water resource management, such as ARPA Puglia, ensuring that the proposed EO services address operational and regulatory needs. The project addresses wastewater treatment plant outfalls, industrial and urban coastal discharges, illegal or unknown pollution events, and agricultural runoff reaching coastal waters through rivers and canals. User feedback supports calibration and validation of the methodologies, ensuring operational relevance and facilitating adoption by environmental authorities. The complementary use of hyperspectral and thermal observations demonstrates how CHIME and LSTM improve the detection and characterisation of pollution events often missed by conventional in situ observations. Beyond preparing for future Copernicus missions, the project highlights the potential of next-generation EO services to strengthen environmental surveillance, inspection planning, policy making, environmental legislation enforcement and environmental compliance monitoring by providing objective spatial evidence.

This paper presents two operational use cases developed in the context of EO4SECURITY Environmental Crimes project. The first use case, developed in cooperation with UNODC, focuses on detecting and tracking illicit crop cultivation and illegal mining activities in Perù , with the goal to demonstrate the added value of geospatial intelligence by means of Earth Observation and OSINT data sources. The integrated use of EO and OSINT proved to be insightful, providing added value information for the identification of environmental degradation and unauthorized land use in vulnerable ecosystems enabling the timely detection . The second use case, realized in cooperation with INTERPOL, addresses the monitoring of illegal fishing activities within protected marine areas in East Africa, leveraging satellite imagery and AIS vessel tracking data to detect suspicious maritime behavior and support conservation and law enforcement effort, generating daily reports provided to the end users to be integrated into their operational workflows. Together, these use cases illustrate how geospatial technologies can enhance situational awareness, strengthen environmental governance, and provide actionable intelligence for combating transnational environmental crimes in different phases, as post-event assessment ,historical analysis and indicators generation for early warnings.

The marine ecosystems are constantly being added with more pressure, as the globally increasing maritime traffic results in increased pollution from oil spills, vessel emissions, sewage and litter. Illegal fishing, discharges and waste dumping activities, enhance the need to develop and use innovative tools and methods for detecting, characterizing and attributing to their source such environmental offences. The Horizon Europe POSEIDON project, Pollution Observation from Space: Environmental Imaging for Detections in the Oceans & Nearshore, is funded by the European Union and comprises ten partners from Europe, Canada and Asia, in order to exploit the recent advances in Earth Observation and Artificial Intelligence, and creates solutions for supporting and strengthening the implementation of the MARPOL regulatory framework, the Ship Sourced Pollution Directive and the European Union Environmental Crime Directive. The services for identifying and detecting oil spills and chemicals in the oceans, sewage, garbage, gas emissions and other pollutants released by ships at sea, including sea ice conditions, are improved by leveraging and integrating various technologies, such as multi-sensor satellites, vessel tracking information, in-situ measurements, environmental models and AI-based analytics. The five international test sites that have been selected, Elefsis Gulf in Greece, Strait of Gibraltar, Maldives, Singapore Strait and Arctic regions, demonstrate these capabilities as they not only face unique and complex environmental conditions, but they also address various challenges that may have societal and economic impact if not identified and tackled in time. POSEIDON showcases how the utilization of different data sources can be incorporated into operational workflows to strengthen the prompt environmental crime detection and improve the situational awareness. Moreover, it identifies technical, operational, and institutional challenges and supports developing a stronger collaboration between enforcement authorities and the scientific community, aiming to contribute to the wider uptake in decision-making.

The detection of ships plays a significant role in supporting navigation safety, fisheries monitoring, maritime law enforcement, and the identification of vessels potentially involved in suspicious maritime activities. This paper presents the outcomes of a proof-of-concept utilising dual-polarimetric C-band SAR data acquired by Sentinel-1 to detect ships. Unlike conventional intensity-based or single-polarisation Constant False Alarm Rate (CFAR) detectors, the proposed method exploits the reflection symmetry property to differentiate between sea surface and vessels. While distributed sea clutter approximately satisfies reflection symmetry, the multiple scattering mechanisms of man-made vessels break this symmetry, providing a physically meaningful contrast that enhances target discrimination, particularly in moderate sea states. Candidate targets are refined using an iterative CFAR thresholding procedure, while persistent stationary structures such as oil platforms are removed using a multi-temporal reference mask built from historical Sentinel-1 acquisitions over the same area, reducing false alarms. The pipeline has shown promising results, identifying ships from 15 m to 300 m in length both in busy maritime environments, such as the Gulf of Mexico and the Singapore Strait, and in the open ocean, as in the Mediterranean Sea. Detections were validated against the Global Fishing Watch dataset, which provides vessel detections matched with Automatic Identification System (AIS) signals at over 99% accuracy. The final goal of this work is to match detected ships against AIS signals to identify non-broadcasting ("dark") vessels, potentially indicating illegal activity and prompting further investigation by maritime authorities. Full integration into Insula, an Earth Observation platform operated by CGI Italy, will enable this matching as part of a systematic, scalable monitoring system for maritime activities. Future work will extend the analysis to the same regions using dual-pol L-band data from NISAR, to assess the impact of higher resolution on detection performance.

Illegal groundwater drilling and abstraction remain among the least visible forms of environmental crime, yet under Article 2(3)(m) of the EU Environmental Crime Directive they can cause substantial damage to the ecological status of surface waters and the quantitative status of groundwater bodies. The IMPEL project TIGDA (Tackling Illegal Groundwater Drilling and Abstractions, 2022(VI)WG2) set out to close this gap by collecting and synthesising detection, permitting and enforcement practices from sixteen respondents across ten European countries, and by running four field workshops in Portugal, Slovenia, Spain and Italy to test detection techniques in situ — from drone-supported live inspections to satellite-based subsidence monitoring. Building on these findings, and on a dedicated technical study carried out with the Technical University of Civil Engineering of Bucharest and the Romanian Space Agency, the project evaluated Earth Observation (EO) methods — including free Copernicus and European Ground Motion Service data — against four escalating levels of detection capability, from broad awareness screening to court-ready evidence. The project's central output is a six-stage guidance framework moving authorities from Awareness and Detection, through Initial Response and Investigation, Compliance Promotion and Inspection, to Monitoring and Capacity Building, integrating drones, GIS geo-databases and cross-referenced permitting records at each stage. This contribution presents the TIGDA stepped-approach methodology and the practical lessons drawn from the four national workshops, with particular attention to how low-cost, open-source EO tools can extend inspection capacity in resource-constrained authorities. It also outlines the project's next phase, linking with the IMPEL GIEDA (Geospatial Intelligence for Environmental Damage Assessment) project to strengthen the evidentiary link between detected groundwater activities and the Environmental Crime Directive's substantial-damage threshold.

Skylight talk

Environmental crimes are on the rise, but we finally have tools to counteract them – political, economic, but above all technological. Among these, satellite Earth Observations (EO) is the most widely applied globally, and the one that witnessed the quickest and most dramatic evolution over the past decade, especially for monitoring purposes and to trigger investigations in sectors such as forestry crime or illegal water abstraction. Yet many are still unaware of its significant potential and much could be gained from a more systematic and structured uptake. All this has been explored in a recent, wide-ranging study entitled “Earth Observation Against Environmental Crimes: A Review of Users and Usages Worldwide”, led by the Rome-based Istituto Affari Internazionali (IAI) under an assignment from ESA. As part of the study, the IAI reviewed the status of EO usage against environmental crimes with a global focus and considering all major sectors, particularly forestry crime, illegal mining, pollution, and waste trafficking. It considers global and regional trends regarding in environmental crimes and EO data availability and usage, analysing the main elements favouring the uptake of satellite-based solutions, and the main obstacles such applications are facing. The project is divided into a general analysis, a sectorial analysis (also including minor sectors, such as sand mining and wildlife crime), and on ten case studies. The study report is expected to be published right before the event. This presentation will be held by Lorenzo Colantoni, senior researcher at IAI, and will include content from the several missions on the field done in, among the others, Indonesia, Mexico, and South Africa. The expected length of the event is 30 minutes – 15 minutes of presentation, and 15 minutes dedicated to Q&A.

In summary, the following are the regulations governing mining activities. •1. If the mining area is located within a forest zone, a forest area borrowing-use permit must be obtained from the Minister of Forestry. •2. An environmental impact analysis must be conducted prior to commencing mining activities. •3. Replacement land must be provided. •4. Taxes must be paid. •5. Mining activities must be conducted with due regard for environmental concerns and in compliance with existing regulations issued by the Ministry of Energy and Mineral Resources. •6. Upon completion of mining operations, reclamation must be carried out in accordance with established procedures. •7. Mining activities may only be conducted based on the Work Plan and Budget issued by the Ministry of Energy and Mineral Resources. •8. Failure to comply with regulations issued by the Ministry of Forestry and the Ministry of Energy and Mineral Resources renders the mining activities illegal.

Indigenous communities play a strategic role as the first line of defense against environmental crimes. Their knowledge of the territory, their ability to identify early illegal activities, and their direct interest in ecosystem protection make them key actors in the fight against environmental crimes. Since 2016, Rainforest Foundation U.S. (RFUS) and the World Resources Institute (WRI) have trained and supported Indigenous communities and organizations in the Peruvian Amazon to use Earth Observation data (Global Nature Watch Integrated Alerts,(1) satellite imagery) and geospatial tools (e.g. Forest Watcher, LocusPro, MapBuilder and Story Maps) to strengthen Indigenous governance and counter environmental crimes in their territories. To date, more than 300 Indigenous men and women in the Loreto Region use Earth Observation data and tools to patrol their territories. In addition to cutting deforestation by half in their territories(2) and helping assert geospatial data sovereignty,(3) the use of Earth Observation data and tools has enabled Indigenous monitors collaborate directly with enforcement authorities to prosecute environmental crimes,(4) identify geospatial deforestation patterns associated with illegal mining, illegal logging and wildlife trafficking modus operandi unique to their territories, and focus on-the-ground control activities to prevent the expansion of nature crime. The presentation will discuss outcomes, insights, best practices and lessons learned in these 10 years of work. Among others, topics will include: Design and implementation of up territorial monitoring workflows including setting up the technical infrastructure required in remote locations with limited connectivity and securing communities’ buy-in. Deployment of training programs through culturally sensitive strategies. Emerging strategies to deliver Earth Observation data and associated field intelligence to government authorities to ensure confidentiality and mitigate threats to community patrols. References: (1) Global Nature Watch, formerly known as Global Forest Watch (2) Slough, T., J. Kopas and J. Urpelainen. 2021. Satellite-based deforestation alerts with training and incentives for patrolling facilitate community monitoring in the Peruvian Amazon | PNAS. (3) Sanchez, N., R. Nogueron and W. Sanchez-Hidalgo. 2024.Indigenous Communities in Peru Protect their Forests with MapBuilder. Global Forest Watch blog. (4) Noriega. 2019. These indigenous Peruvians are fighting deforestation with satellites and drones. Vice News.

EO applications against environmental crimes have increased exponentially in the past decade, consolidating uses in sectors such as forestry crime, while expanding also to other, new areas, such as wildlife crime or illegal water abstraction. As a result, a variety of players is now employing EO, from enforcement agencies to NGOs, judges, and prosecutors, and uses ranges from evidence in court, to monitoring and even pre-emptive action. The event will consist in a roundtable with two representatives of different categories of users of EO data, which are actively involved in the global fight against environmental crimes: Julieta Calvo, director of the Centro Nacional de Información Geoambiental (CENIGA) in Costa Rica, and Bambang Hero Saharjo, expert witness for the Indonesian police and professor at the University of Bogor, Indonesia. Julieta Calvo coordinates all the preliminary EO analysis Costa Rican environmental courts are now obliged to perform. Bambang Saharjo is instead one of the most relevant global experts on EO uses against wildfires and illegal mining, and has been involved in some of the biggest environmental investigations in Indonesia in recent years. The event will be moderated by Lorenzo Colantoni, senior researcher at the Rome-based Istituto Affari Internazionali (IAI), and main author of the study “Earth Observation Against Environmental Crimes”, produced under an assignment from ESA (also proposed for a presentation in a separate abstract). Ms. Calvo and Mr. Saharjo have both been involved in the study, and represent some of the most interesting case studies uncovered by the research. The expected duration of the event is 30 minutes. Both participants will participate in person.

With the official designation of DG ENV as SESA Requesting Party in 2025, SESA has played an increasingly important role in supporting the implementation of the Environmental Crime Directive (ECD) and strengthening cooperation on environmental compliance across the European Union and beyond. The presentation will provide an overview of the collaboration between SESA and DG ENV, highlighting key milestones achieved during the first two years of implementation. These include SESA's participation in the 4 Networks Conference on the implementation of the Environmental Crime Directive, held in Bratislava in May 2025 and the establishment of a dedicated community of environmental crime and compliance users, bringing together Points of Contact from all 27 EU Member States. The session will also present the user uptake activities celebrated with the ENV crime community of users, including a series of information sessions to facilitate the access to SESA products and services. Finally, it will outline SESA's ongoing work in monitoring the implementation of the Environmental Crime Directive across Member States (showing some use cases). The presentation will demonstrate how coordinated operational support, structured user community and sustained engagement with national authorities contribute to more effective implementation of EU environmental legislation and enhanced cooperation in addressing environmental crime/compliance.

Environmental crime increasingly intersects with organised crime, illicit financial flows, smuggling, conflict and serious violations of international law. Addressing these interconnected threats requires moving beyond environmental monitoring towards integrated environmental security approaches that combine scientific evidence, governance, law enforcement and international cooperation. This keynote will explore how Earth Observation and geospatial intelligence can strengthen decision-making, support investigative and enforcement activities, and improve accountability for environmental harm. Drawing on the experience of the Institute for Environmental Security, it will highlight the use of satellite data and complementary information sources to detect and document environmental degradation, identify patterns linked to criminal activities, support investigations into money laundering and illicit trafficking, and contribute to the assessment of potential ecocide and crimes against humanity. The presentation will also address the institutional and operational conditions required to transform geospatial information into actionable knowledge, including data integration, capacity building, legal frameworks, cross-sector cooperation and access to scientifically robust decision-support tools. Particular attention will be given to bridging the gap between environmental policy and security policy, enabling authorities and international organisations to respond more effectively to complex and transnational environmental threats.