Groundwater Dependent Ecosystems (GDEs) mapping deliverables, 2024–2025

The Groundwater Dependent Ecosystems (GDE) Mapping dataset provides geospatial information on the predicted distribution and probability of groundwater-dependent ecosystems within selected watersheds in Alberta's Oil Sands Region (OSR). Developed collaboratively by the Alberta Biodiversity Monitoring Institute (ABMI) and InnoTech Alberta under the Oil Sands Monitoring (OSM) Program, the dataset integrates remote sensing, hydrogeological, topographic, climatic, vegetation, wetland, and field observation data to identify potential aquatic and terrestrial ecosystems that depend partially or entirely on groundwater. The dataset was produced using an ensemble machine learning approach combining Extra Trees, Gradient Boosting, K-Nearest Neighbours, and Random Forest algorithms. Key explanatory variables include wetland classifications, topographic wetness index, soil drainage, elevation, depth to groundwater, height above nearest drainage, and synthetic aperture radar (SAR) measurements. The mapping products include predicted GDE probability surfaces and model classifications identifying potentially groundwater-dependent areas, including wetlands, riparian zones, streams, and terrestrial ecosystems associated with shallow groundwater. The 2024–2025 mapping covers portions of the McKay, Steepbank, Beaver, and Clearwater River watersheds, building upon the aquatic GDE mapping completed during 2023–2024. The dataset supports identification of potential groundwater-dependent ecosystems, groundwater–surface water interaction studies, monitoring site selection, cumulative effects assessment, and planning of long-term groundwater and ecosystem monitoring activities in the Oil Sands Region.

Data and Resources

Additional Info

Field Value
Alternate Title/Name
Creator Alberta Biodiversity Monitoring Institute (ABMI)
Contributor OSM Integrated Geospatial Program
Publisher Alberta Environment and Protected Areas
Subject Groundwater, Groundwater Dependent Ecosystems, Surface Water, Wetlands, Hydrology, Hydrogeology, Aquatic Ecosystems, Terrestrial Ecosystems, Vegetation, Environmental Monitoring, Cumulative Effects
Group/Subgroup Geospatial
Parameters GDE Probability, GDE Presence and Absence Classification, Aquatic GDE, Terrestrial GDE, Wetland Class, Wetland Form, Depth to Water, Topographic Wetness Index (TWI), Height Above Nearest Drainage (HAND), Elevation, Slope, Soil Drainage, Permeability, Aquifer Hosting Sediment, Normalized Difference Vegetation Index (NDVI), Normalized Difference Moisture Index (NDMI), SAR Backscatter, SAR Coherence, Aridity Index, Groundwater Indicators
Data Type Discrete, Continuous, Geospatial Raster Data, Modelled Probability Surfaces, Remote Sensing Derived Data, Field Observations
Frequency Annual Project Updates (2024–2025); Modelled Spatial Snapshot
Temporal Coverage Begins 2024-04-01
Temporal Coverage Ends 2025-03-31
Spatial Coverage thabasca Oil Sands Area, Northeastern Alberta, Canada, including portions of the McKay River, Steepbank River, Beaver River, and Clearwater River watersheds north of Fort McMurray.
Spatial Projection NAD 1983 10TM AEP Forest (EPSG:3400)
Usage Considerations or Constraints The dataset represents modelled probabilities of groundwater-dependent ecosystem occurrence and should not be interpreted as field-confirmed groundwater dependence at every location. GDE predictions were generated using ensemble machine learning models based on environmental explanatory variables, remotely sensed data, and available field observations. The models achieved internal area-under-the-curve (AUC) values of approximately 0.86–0.90; however, these results do not necessarily represent predictive accuracy in independent geographic areas. Training data for terrestrial GDEs were partly generated using rule-based ecological and hydrogeological criteria because of limited field observations. Further independent field validation and spatial cross-validation are recommended to assess model transferability and uncertainty. Predictions may be influenced by input data limitations and modelling artefacts. For example, some mining areas were assigned high GDE probabilities because low elevations in mine pits and tailings areas resembled topographic conditions associated with groundwater-dependent environments. The model combines aquatic and terrestrial GDEs into a single prediction class; therefore, the probability surfaces should not be interpreted as independently validated classifications of individual GDE types. Spatial coverage is limited to the selected study area and surrounding analysis extent. The model should be further validated before application to other portions of the Oil Sands Region. The dataset is intended to support regional screening, research, environmental monitoring, and monitoring program planning. It should not be used as the sole basis for site-specific groundwater assessments, regulatory decisions, or confirmation of ecosystem groundwater dependence. Users should consult the accompanying technical report for detailed modelling methods, field data, explanatory variables, model performance, uncertainty, and interpretation guidance
Source/Citation Alberta Biodiversity Monitoring Institute. (2026). Mapping Groundwater Dependent Ecosystems in Alberta’s Oil Sands Region – 2024/25. Prepared for the Oil Sands Monitoring Program. Alberta Biodiversity Monitoring Institute and InnoTech Alberta, Alberta, Canada.
Relation
Extent
Format .gdb
Resource Type Geodatabase
Language English
OSM Asset? Yes
Contact Email [email protected]