Qing et al. (2025) Delayed formation of Arctic snow cover in response to wildland fires in a warming climate

Identification

• Journal: Nature Climate Change
• Year: 2025
• Date: 2025-09-23
• Authors: Yamin Qing, Shuo Wang, Amir AghaKouchak, Pierre Gentine
• DOI: 10.1038/s41558-025-02443-6

Research Groups

• State Key Laboratory of Climate Resilience for Coastal Cities (SKL-CRCC), The Hong Kong Polytechnic University, Hong Kong, China
• Department of Land Surveying and Geo-Informatics, The Hong Kong Polytechnic University, Hong Kong, China
• Research Institute for Land and Space and Research Institute for Sustainable Urban Development, The Hong Kong Polytechnic University, Hong Kong, China
• Department of Civil and Environmental Engineering, University of California, Irvine, CA, USA
• Department of Earth System Science, University of California, Irvine, CA, USA
• United Nations University, Institute for Water, Environment and Health, United Nations University, Hamilton, Ontario, Canada
• Department of Earth and Environmental Engineering, Columbia University, New York, NY, USA

Short Summary

This study reveals that from 1982 to 2018, Arctic wildland fires significantly increased, causing a substantial delay in snow cover formation primarily due to fire-induced albedo reduction and temperature increases. Projections under a high-emissions scenario indicate a 2.6-fold increase in burned area and an 18-day decrease in annual mean snow cover duration by 2100.

Objective

• To investigate the causal effects of wildland fires on Arctic snow cover changes, specifically focusing on delayed snow cover formation.
• To assess the response of snow cover duration to major wildland fires.
• To examine future trends in wildland fires and snow cover under different Shared Socioeconomic Pathways (SSPs).

Study Configuration

• Spatial Scale: Arctic region (latitude ≥60° N).
• Temporal Scale: Historical analysis: 1982–2018. Projections: 1950–2100 (historical 1950-2014, future 2015-2100).

Methodology and Data

• Models used:
  • XGBoost (extreme gradient boosting) for predicting delayed snow cover formation and early disappearance.
  • Structural Equation Modeling (SEM) for exploring causal mechanisms.
  • Convergent Cross Mapping (CCM) for testing causal relationships in nonlinear time-series systems.
  • Superposed Epoch Analysis (SEA) for analyzing snow cover response to major fire events.
  • Least-squares linear regression for trend analysis and burned area projections.
  • Multiple linear regression for burned area projections.
• Data sources:
  • Satellite-based Burned Area (BA) products: FireCCILT11 (1982–2018), FireCCI51 (2001–2018), MCD64A1 v.6 (2001–2018).
  • Daily gridded Snow Water Equivalent (SWE) reconstruction (1982–2018).
  • ERA5-Land reanalysis: 2-meter air temperature (Ta), 2-meter dewpoint temperature, potential evaporation, vegetation transpiration (for VPD and Def calculation).
  • Polar Pathfinder Extended Climate Data Record: Albedo, Land Surface Temperature (LST), Surface Downwelling Shortwave Flux (SW), Surface Downwelling Longwave Flux (LW).
  • Coupled Model Intercomparison Project phase 6 (CMIP6) simulations for future projections (SSPs 2-4.5, 3-7.0, 5-8.5).

Main Results

• From 1982 to 2018, the burned area in the Arctic significantly increased (approximately 20,000 km² increase, P < 0.01), while snow cover duration in fire-affected areas significantly decreased (shortening by more than 15 days, P < 0.01).
• The average snow cover duration from 2001 to 2018 was 205 days, which is 10 days shorter than the 1982–2000 average.
• Post-fire factors, particularly post-fire albedo, air temperature (Ta), and land surface temperature (LST), had the strongest influence on delayed snow cover formation.
• Causal analysis confirmed that summertime burned area significantly increased post-fire LST and simultaneously decreased post-fire albedo, which in turn increased LST and Ta, ultimately leading to delayed snow cover formation (R² = 0.53, P < 0.01).
• Major wildland fires (burned area > 1 standard deviation above mean) caused a significant delay in snow cover formation, exceeding 5 days relative to the 3-year pre-fire average in the immediate snow year following the event. This delay scaled positively with fire severity.
• No significant advance in the snow cover end date was detected following major wildland fires.
• Under a high-emissions scenario (SSP 5-8.5), the annual burned area in the Arctic is projected to increase by a factor of 2.6, and the annual mean snow cover duration is projected to decrease by nearly 18 days between 2015 and 2100 compared to the historical average.

Contributions

• Provides the first comprehensive analysis of the causal effects of wildland fires on snow cover changes in the Arctic, specifically identifying the mechanisms (albedo reduction, temperature increase) leading to delayed snow cover formation.
• Quantifies the significant delay in snow cover formation (over 5 days) following major wildland fires and demonstrates its scaling with fire severity.
• Offers future projections of increased burned area and decreased snow cover duration under various climate change scenarios, highlighting amplified interactions.
• Utilizes a robust combination of machine learning (XGBoost), causal inference (SEM, CCM), and event-based analysis (SEA) with long-term satellite data to establish these relationships.

Funding

• Research Grants Council of the Hong Kong Special Administrative Region, China (project no. PolyU/RGC 15232023)
• The Hong Kong Polytechnic University (project no. P0052737)

Citation

@article{Qing2025Delayed,
  author = {Qing, Yamin and Wang, Shuo and AghaKouchak, Amir and Gentine, Pierre},
  title = {Delayed formation of Arctic snow cover in response to wildland fires in a warming climate},
  journal = {Nature Climate Change},
  year = {2025},
  doi = {10.1038/s41558-025-02443-6},
  url = {https://doi.org/10.1038/s41558-025-02443-6}
}