EOWetMet: Annual Wetland and Lake Classification Dataset for NWT and Sweden (2021-2025)
- Source
- DataCite
- Published
- 2026
- License
- Creative Commons Attribution 4.0 International
- Themes
- Water & Sanitation, Fresh Water, Remote Sensing & Satellites
About this dataset
The EOWetMet Annual Wetland and Lake Classification Dataset (2021-2025) provides annual land cover maps distinguishing six wetland and lake ecosystem classes across two Arctic 500 km × 500 km study areas in the Northwest Territories (NWT), Canada, and northern Sweden, for the period 2021 to 2025. The dataset forms part of the EOWetMet Community Product Database and was developed to capture the spatial and temporal characterisation of wetland and lake inundation at high resolution, in support of bottom-up methane emission modelling and a broad range of Arctic science applications including hydrological monitoring, permafrost and thermokarst change detection, and biodiversity assessment. The dataset provides high-resolution wetland and lake classification products in the Arctic-boreal region that integrate temporal inundation information to produce annual classification maps reflecting...
How the data was collected
The classification was produced using an Object-Based Image Analysis Random Forest classification framework implemented in Google Earth Engine, integrating 35 input features across multiple data categories including Sentinel-1, Sentinel-2, ALOS PALSAR-2, ArcticDEM, ERA5, MERIT Hydro and annual inundation information derived from Sentinel-1. Each pixel is classified as one of six classes identified by the following pixel ids:1 – Bog, 2 – Fen, 3 – Swamp, 4 – Marsh, 5 – Water, 6 – Upland (non-wetland and lake).