Arctic Winter Soil CO2 Flux Gridded Estimates 2003-2100

Gridded estimates of daily average soil CO2 flux for the non-growing season across pan-Arctic and Boreal permafrost regions north of 49 degrees. The dataset provides monthly data at 25 km resolution for a historical baseline (2003-2018) and future climate scenarios (2018-2100) under RCP 4.5 and 8.5. It was produced by NASA using a Boosted Regression Tree machine learning model trained on in situ observations.

Use Cases

• Projecting future winter carbon emissions from permafrost regions under different climate scenarios.
• Calculating annual non-growing season soil CO2 flux budgets from 2003 to 2100.
• Validating and improving process-based terrestrial carbon cycle models.
• Analyzing spatial patterns of soil respiration in Arctic and Boreal zones.

Strengths

• Covers a long temporal range from 2003 to 2100, including historical and future projections.
• Provides data at a consistent 25 km spatial resolution for the entire pan-Arctic region (>49 Deg N).
• Derived using a machine learning approach (Boosted Regression Tree) calibrated with in situ observations.

Limitations

• Specific column names, row counts, and dataset size are not provided by any source.
• Conflicting 'last updated' dates exist across platforms (2100-11 30 and 2026-03-13).
• The license is ambiguously listed as 'other-license-specified' without explicit terms.

Provenance

Source

National Aeronautics and Space Administration

Collection Method

Boosted Regression Tree machine learning model applied to in situ observations of non-growing season fluxes.

Time Range

2003-2100

Freshness

2026-03-13 01:47:37.791559

Geography

Pan-Arctic and Boreal permafrost regions (>49 Deg N)