A predictive model for Canadian carbonatite-hosted REE ± Nb deposits is presented herein. This model was developed by integrating diverse data layers derived from geophysical, geochronological, and geological sources. These layers represent the key components of carbonatite-hosted REE ± Nb mineral systems, including the source, transport mechanisms, geological traps, and preservation processes. Deep learning algorithms were employed to integrate these layers into a comprehensive predictive framework. Here is a link to the publication that describes this product: https://link.springer.com/article/10.1007/s11053-024-10369-7

This model is derived from geological, geophysical and other forms of geodata. Feature extraction used deep learning. Predictive modelling made use of the deep ensemble method. Displayed is a Pan-Canadian probability map of mineral potential of graphite. This map was generated using known graphite deposits and occurrences and their associated features. Higher probability values highlight areas with an increased probability of graphite mineral systems.

STE_TEI_ATTRIBUTE_POLYS_SP contains Terrestrial Ecosystem Information (TEI) polygons with key and amalgamated (concatenated) attributes derived from the RISC (Resource Inventory Standards Committee) standard attributes. These describe the physical and biological characteristics of ecosystems at a landscape level. TEI currently includes Terrestrial Ecosystem Mapping (TEM), Predictive Ecosystem Mapping (PEM), Sensitive Ecosystems Inventory (SEI), Terrain Mapping (TER), Wetland Mapping (WET) and Soil Mapping (SOIL). Mapping methods include manual air photo interpretation and modeling supported by limited field checking. **Current version**: v12 (2026-06-22)

Broad Ecosystem Units were mapped using predictive modeling methods from various data sources (ranging from 1:50,000 to 1:250,000 in scale) and are referenced to the CanVec digital spatial framework (1:50,000). Broad Ecosystem Units (BEU) are a level in the Yukon bioclimate ecosystem classification system that represents areas with similar broad vegetation communities, terrain type (soils and topography) within bioclimate zones. Broad Ecosystem Units are described in the accompanying report "Regional Ecosystems of West-Central Yukon, Part 1: Ecosystem descriptions ".The intended application for mapped broad ecosystem units is 1:100,000 or smaller (1:100,000 - 1:250,000 scale) - interpretations derived from the map products should not be applied at more detailed scales, even though the resultant 30m raster map allows users to view results at more detailed resolutions. With new information, boundaries and designations of Broad Ecosystem Units can change. Updates to Broad Ecosystem Units occur only periodically. For the most current information, or if you have questions, please contact the Ecological and Landscape Classification Program (ELC@yukon.ca).Distributed from [GeoYukon](https://yukon.ca/geoyukon) by the [Government of Yukon](https://yukon.ca/maps) . Discover more digital map data and interactive maps from Yukon's digital map data collection.For more information: [geomatics.help@yukon.ca](mailto:geomatics.help@yukon.ca)

This model is derived from geological and geophysical data, which is processed using deep learning and natural language processing techniques. Displayed is a Pan-Canadian probability map indicating the likelihood of discovering next-generation lithium-cesium-tantalum (LCT) pegmatites. This map was generated using known Canadian LCT pegmatites and their associated geospatial features, incorporating geological and geophysical data analyzed through deep learning and natural language processing techniques. Higher probability values highlight areas with an increased likelihood of hosting next-generation deposits, making this map a valuable tool for decision-making.

This dataset is not currently available in the BCGW but may be downloaded in geodatabase format. Please refer to: [Terrestrial Ecosystem Information (TEI) Data Distribution Packages](http://catalogue.data.gov.bc.ca/dataset/terrestrial-ecosystem-information-tei-data-distribution-packages) PEM_Polygons contains Predictive Ecosystem Mapping polygons with key and amalgamated (concatenated) attributes derived from the RISC (Resource Inventory Standards Committee) standard attributes. PEM uses modeling to divide the landscape into units according to a variety of ecological features including climate, physiography, surficial material, bedrock geology, soils and vegetation. This layer is derived from the STE_TEI_ATTRIBUTE_POLYS_SP layer by filtering on the PROJECT_TYPE attribute. COMMENTS - this spatial view replaces the older STE_PEM_SHORT_TABLE_SP.

Dominant Species Map 2015The data represent dominant tree species for British Columbia forests in 2015, are based upon Landsat data and modeling, with results mapped at 30 m spatial resolution. It is developed within the framework of Canada’s National Terrestrial Ecosystem Monitoring System (NTEMS). The map was generated with the Random Forests classifier that used predictor variables derived from Landsat time series including surface reflectance, land cover, forest disturbance, and forest structure, and ancillary variables describing the topography and position. Training and validation samples were derived from the Vegetation Resources Inventory (VRI), from a pool of polygons with homogeneous internal conditions and with low discrepancies with the remotely sensed predictions. Local models were applied over 100x100 km tiles that considered training samples from the 5x5 neighbouring tiles to avoid edge effects. An overall accuracy of 72% was found for the species which occupy 80% of the forested areas. Satellite data and modeling have demonstrated the capacity for up-to-date, wall-to-wall, forest attribute maps at sub-stand level for British Columbia, Canada.BC Species Likelihood 2015The tree species class membership likelihood distribution data included in this product focused on the province of British Columbia, based upon Landsat data and modeling, with results mapped at 30 m spatial resolution. The data represent tree species class membership likelihood in 2015. The map was generated with the Random Forests classifier that used predictor variables derived from Landsat time series including surface reflectance, land cover, forest disturbance, and forest structure, and ancillary variables describing the topography and position. Training and validation samples were derived from the Vegetation Resources Inventory (VRI) selecting from a stratified pool of polygons with homogeneous internal conditions and with low discrepancies when related to remotely sensed information. Local models were applied over 100x100 km tiles that, to avoid edge effects, considered training samples from the 5x5 neighbouring tiles. An overall accuracy of 72% was found for the species which occupy 80% of the forested areas. As an element of the mapping process, we also obtain the votes received for each class by the Random Forest models. The votes can be understood as analogous to class membership likelihoods, providing enriched information on land cover class uncertainty for use in modeling. Tree species class membership likelihoods lower than 5% have been masked and converted to zero.When using this data, please cite as: Shang, C., Coops, N.C., Wulder, M.A., White, J.C., Hermosilla, T., 2020. Update and spatial extension of strategic forest inventories using time series remote sensing and modeling. International Journal of Applied Earth Observation and Geoinformation 84, 101956. DOI: 10.1016/j.jag.2019.101956 ( Shang et al. 2020).

This deep water substrate bottom type model was created to aid in habitat modeling, and to complement the nearshore bottom patches. It was created from a combination of bathymetrically-derived layers in addition to bottom type observations. Using random forest classification, the relationship between observed substrates and bathymetric derivatives was estimated across the entire area of interest. The raster is categorized into: 1) Rock, 2) Mixed, 3) Sand, 4) Mud

Seamounts have been identified as Ecologically or Biologically Significant Areas (EBSAs) due to their unique oceanography and ecology; they frequently serve as sites for fisheries and as habitat for a number of species of conservation concern. A mix of isolated seamounts and seamount complexes are distributed throughout Canada’s Pacific offshore waters, although only a subset of these are named. We used several pre-existing spatial databases and predictive models to map all named seamounts within Canada’s Exclusive Economic Zone (EEZ), all named seamounts fished by Canada in international waters, and any predicted (modelled) unnamed seamounts in the EEZ. These data are intended to inform marine planning initiatives in BC by providing collaborative, peer-reviewed scientific data at scales relevant to a BC coast-wide analysis.

Landscape Units (Subdivisions) were developed as a part of the 2002 Northern Caribou Strategy and are used for landscape level planning within Wildlife Habitat Area 5-086. For details please see: [Apps, C. D., T. A. Kinley, and J. A. Young. 2001. Multi-scale habitat modeling for woodland caribou in the Itcha, Ilgachuz, and Rainbow mountains of west-central British Columbia. Wildlife Section, Ministry of Water, Land and Air Protection, Williams Lake, British Columbia, Canada.](http://www.env.gov.bc.ca/cariboo/env_stewardship/wildlife/inventory/caribou/northcar/hmi/hsi06-01.pdf)