Résumé
Forest health plays a critical role in human wellbeing, providing air purification, climate regulation, and ecosystem services relevant to public health. The growing availability of remote sensing data and advances in AI have enabled substantial progress in forest monitoring and modeling. However, as climate change accelerates forest decline worldwide, comparatively little research has addressed how this decline, particularly when driven by air pollution, may affects the vegetation on which people depend. This gap is especially pressing in Luxembourg, where forests cover 92,150 ha (35% of the country) and are increasingly exposed to tropospheric ozone (O₃) and nitrogen oxides (NOₓ). The Nature and Forest Agency’s inventory highlight this growing vulnerability, with only 14.5% of trees classified as undamaged in 2023 compared with 79% in 1984.
This study aims to assess the potential impact of air pollution on forest tree health across Luxembourg, and to develop a methodological framework for identifying areas at higher risk, paving the way for future mitigation strategies. The proposed workflow integrates multi-source remote sensing data, including high-resolution aerial RGB/CIR orthophotos, LiDAR-derived canopy height models, and satellite-based air quality data (e.g., Sentinel-5P, CAMS). AI-based models will be used to detect and characterize potential signs of canopy stress and decline in relation to ambient pollutant concentrations, ozone-related exposure metrics, including AOT40 and, where feasible, phytotoxic ozone dose (POD) indicators, will be explored.
The expected outcomes include a spatially explicit assessment of forest areas potentially vulnerable to air-pollution-related stress, together with an evaluation of how AI-based remote sensing methods can support the detection of tree-level canopy condition. By focusing on observable indicators of forest stress rather than directly estimating human health outcomes, the study provides a methodological basis for future research linking forest degradation, ecosystem-service loss, and human wellbeing. The poster will present the proposed analytical workflow, data integration strategy, and expected outputs.
This study aims to assess the potential impact of air pollution on forest tree health across Luxembourg, and to develop a methodological framework for identifying areas at higher risk, paving the way for future mitigation strategies. The proposed workflow integrates multi-source remote sensing data, including high-resolution aerial RGB/CIR orthophotos, LiDAR-derived canopy height models, and satellite-based air quality data (e.g., Sentinel-5P, CAMS). AI-based models will be used to detect and characterize potential signs of canopy stress and decline in relation to ambient pollutant concentrations, ozone-related exposure metrics, including AOT40 and, where feasible, phytotoxic ozone dose (POD) indicators, will be explored.
The expected outcomes include a spatially explicit assessment of forest areas potentially vulnerable to air-pollution-related stress, together with an evaluation of how AI-based remote sensing methods can support the detection of tree-level canopy condition. By focusing on observable indicators of forest stress rather than directly estimating human health outcomes, the study provides a methodological basis for future research linking forest degradation, ecosystem-service loss, and human wellbeing. The poster will present the proposed analytical workflow, data integration strategy, and expected outputs.
| langue originale | Anglais |
|---|---|
| état | Publié - 7 juil. 2026 |
| Evénement | Climate Change and Health Symposium : Challenges and Opportunities for Luxembourg - Neimënster Abbey, Luxembourg, Luxembourg Durée: 7 juil. 2026 → 7 juil. 2026 https://opc-luxembourg.lu/en/event/climate-change-and-health/ |
Une conférence
| Une conférence | Climate Change and Health Symposium |
|---|---|
| Pays/Territoire | Luxembourg |
| La ville | Luxembourg |
| période | 7/07/26 → 7/07/26 |
| Adresse Internet |
Contient cette citation
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver