# Coarse land cover datasets bias Arctic-Boreal wetland methane budgets

**Authors:** Josh Hashemi, Aleksi Räsänen, Tarmo Virtanen, Sari Juutinen, Guido Grosse, Mika Aurela, Annett Bartsch, Laura Chasmer, Scott J. Davidson, Mika Korkiakoski, McKenzie A. Kuhn, Mark J. Lara, Miska Luoto, Pekka Niittynen, David Olefeldt, Oliver Sonnentag, Anna-Maria Virkkala, Carolina Voigt, Claire C. Treat

PMC · DOI: 10.1038/s43247-025-02963-1 · 2025-11-14

## TL;DR

Coarse land cover maps lead to inaccurate methane emission estimates in Arctic and boreal wetlands, highlighting the need for high-resolution data.

## Contribution

The study demonstrates that high-resolution land cover data is essential for accurate methane budget estimation in Arctic and boreal regions.

## Key findings

- CH4 flux estimates remain within 13% error at resolutions ≤25 m.
- Resolutions coarser than 1 km often misclassify wetlands as methane sources instead of sinks.
- Fens are underrepresented in coarse-resolution maps, affecting emission estimates.

## Abstract

Accurate methane (CH4) emission estimates from Arctic and boreal wetlands are essential for reducing global budget uncertainties but are hindered by poorly constrained wetland distribution and classification. We assessed how land cover map resolution and thematic detail influence these estimates. Using very high spatial resolution land cover maps (≤2.5 m) with five to seven harmonized classes and 4–50% wetland coverage, we estimated CH4 emissions across seven Arctic and boreal sites in North America and Eurasia. Resampling to coarser resolutions (up to 5 km) revealed that CH4 flux estimates remained within 13% error when resolution was ≤25 m pixel size. At resolutions coarser than 1 km, four of seven sites shifted from net CH4 source to sink, due to misrepresentation of wetland extent in heterogeneous landscapes with small, fragmented wetlands. Thematic detail also proved critical, as fens—high CH4 emitters—were disproportionately underrepresented in coarse (>1 km) maps relative to other wetland types. We also show that existing global or circumpolar land cover maps tend to misrepresent wetlands, either overlooking smaller features or overestimating coverage in wetland dominated areas. Our findings indicate that coarse-scale land cover datasets are unsuitable for estimating CH4 budgets in these regions, where high spatial resolution and biogeochemically relevant land cover classes are essential for reliable CH4 emission upscaling.

Estimation of methane flux remained within 13% error at a resolution within 25 m, but resolutions coarser than 1 km often misclassified sites as methane sources instead of sinks, based on high-resolution analysis across seven Arctic and boreal sites in North America and Eurasia.

## Linked entities

- **Chemicals:** methane (PubChem CID 297), CH4 (PubChem CID 297)

## Full-text entities

- **Chemicals:** CH4 (MESH:D008697)

## Figures

5 figures with captions in the complete paper: https://tomesphere.com/paper/PMC12618245/full.md

---
Source: https://tomesphere.com/paper/PMC12618245