As wildfires intensify under changing climatic and socio-ecological conditions, Earth Observation (EO) has become indispensable for monitoring, modelling, and understanding wildfire dynamics across spatial and temporal scales. Advances in remote sensing have enabled the development of wildfire-related models. However, the reliability and transferability of such models depend fundamentally on the quality and accessibility of the datasets upon which they are trained. Over the past decades, rapidly expanding wildfire-related datasets have emerged from airborne and spaceborne observations. These datasets span diverse sensor modalities, spatial resolutions, temporal coverages, and geographic domains. Nevertheless, many remain highly application-specific and regionally constrained. Such fragmentation hinders cross-comparison, interoperability, benchmarking, and reproducibility, ultimately limiting the broader scientific and operational uptake of both datasets and downstream modelling approaches.
This session focuses on the challenges associated with the development, maintenance, validation, and reuse of remote sensing-derived wildfire datasets. We welcome contributions drawing upon airborne and spaceborne observations, including data acquired from polar-orbiting and geostationary platforms, and encompassing optical, thermal infrared, hyperspectral, SAR, and LiDAR sensors. Contributions addressing multi-sensor and multi-platform integration are particularly encouraged.
Topics of interest include, but are not limited to:
• Methodologies for the generation of wildfire-related datasets (data acquisition, annotation, semantic segmentation).
• Scalable data dissemination frameworks (FAIR-compliant data management, STAC-based catalogues, geospatial data cubes).
• Scientific reproducibility
• Dataset validation
This session seeks to advance best practices in the development, validation, stewardship, and dissemination of wildfire remote sensing datasets, while promoting interoperability, transparency, reproducibility, and long-term reusability. Contributions presenting openly accessible resources are especially encouraged. Early career scientists are particularly welcome to contribute.