Spatial Estimates of Carbon Combustion from 2015 Saskatchewan Wildfires
Updated 4mo ago
1filesISO
Available on 1 platform
Sign in to view source links and access this dataset
Description
47 post-fire field measurements from 2015 burned stands in Saskatchewan, Canada, were used to train a radial support vector machine model. The National Aeronautics and Space Administration produced this 30-meter resolution grid dataset, which estimates carbon combustion (kg C/m2) and uncertainty for all 2015 wildfire areas in the province. Model predictions incorporate environmental variables across two ecozones, three soil moisture classes, and three stand dominance types.
Use Cases
Modeling regional carbon emissions from boreal wildfires based on spatial combustion estimates.
Assessing wildfire impact on carbon stocks based on ecozone and stand type classifications.
Validating remote sensing fire products using field-derived combustion measurements.
Analyzing combustion uncertainty across landscapes using the provided per-pixel standard deviations.
Strengths
Derived from 47 field-measured burned stands and 32 unburned control stands.
Includes per-pixel uncertainty estimates generated via Monte Carlo analysis.
Covers two ecozones (Boreal Plains and Boreal Shield) and multiple stand types.
Limitations
Column-level documentation is absent; field semantics must be inferred after download.
Data is specific to 2015 fires in Saskatchewan, Canada, limiting temporal and geographic generalizability.
Provenance
Source
National Aeronautics and Space Administration
Collection Method
Predictive radial support vector machine model trained on field measurements and applied with geospatial predictors.
Time Range
2015-2016
Freshness
Last updated 2026-03-12 19:46:42.333985; freshness should be verified.
Geography
Saskatchewan, Canada
License is listed as 'other-license-specified'; specific terms must be checked before use.