Alberta 10 m Vegetation-Based Wildfire Fuel-Type Map and Supporting Geospatial Data

This dataset provides a province-wide 10 m vegetation-based wildfire fuel-type proxy map for Alberta, Canada, generated using L2HFuelNet, a weakly supervised resolution-enhancement framework. The model was trained using spatially coarse 30 m fuel labels and fine-resolution multi-source remote sensing inputs. The input feature set comprised Sentinel-2 multispectral observations, Sentinel-1 C-band SAR backscatter, and topographic variables derived from the Canadian Digital Elevation Model. The remote sensing inputs were accessed and processed through Google Earth Engine. Raw third-party input datasets are not redistributed in this deposit; their official sources and access links are documented in the accompanying README. The map represents seven aggregated categories: Spruce, Pine, Aspen, Grass, Mixedwood, Non-fuel, and Water. These categories are vegetation-based fuel-type proxies intended for regional fuel characterization, analysis of broad fuel continuity and fragmentation, landscape-scale planning, and methodological research. They do not preserve every subclass distinction in the operational Canadian Fire Behavior Prediction system, including the conifer–deciduous proportions represented by the original Mixedwood subclasses. The dataset should therefore not be used as a standalone input for quantitative fire-behavior simulation without supplementary forest-inventory, field, or operational FBP information. The deposit includes the 10 m GeoTIFF map, study-area boundary data, fuel-class definitions and color legend, a study-area overview image, documentation of the multi-source inputs, and machine-readable metadata. Detailed file descriptions, spatial-reference information, data provenance, recommended uses, and limitations are provided in the accompanying README.