García‐Pereira et al. (2025) Permafrost sensitivity to soil hydro-thermodynamics in historical and scenario simulations with the MPI-ESM

Identification

• Journal: ˜The œcryosphere
• Year: 2025
• Date: 2025-11-20
• Authors: Félix García‐Pereira, J. Fidel González‐Rouco, Nagore Meabe-Yanguas, Philipp de Vrese, Norman Julius Steinert, Johann Jungclaus, Stephan Lorenz
• DOI: 10.5194/tc-19-5959-2025

Research Groups

• Geosciences Institute, IGEO (CSIC-UCM), Madrid, Spain
• Faculty of Physical Sciences, Complutense University of Madrid (UCM), Madrid, Spain
• Max Planck Institute for Meteorology, Hamburg, Germany
• CICERO – Center for International Climate Research, Oslo, Norway

Short Summary

This work shows that changing the hydrological state of permafrost produces differences of up to 3 °C in the annual ground temperature, 1–2 m in the active layer thickness, and 5 million km² in the permafrost extent. Including a deeper vertical thermal scheme reduces the extent decline by more than 2 million km² in the highest radiative emission scenario, demonstrated for the first time in fully-coupled Earth System Model experiments.

Objective

• To assess the sensitivity of permafrost dynamics (temperature, active layer thickness, and extent) to changes in soil hydro-thermodynamics, including soil hydrology configurations (dry/wet) and land surface model depth/vertical resolution, using fully-coupled Earth System Model simulations.

Study Configuration

• Spatial Scale: Circumpolar Arctic and subarctic land areas (Northern Hemisphere, 45–90° N).
• Temporal Scale: Historical (1850–2014) and future climate change scenarios (2015–2100) under Shared Socioeconomic Pathways (SSP1-1.9, SSP2-4.5, SSP5-8.5).

Methodology and Data

• Models used:
  • Earth System Model: Max Planck Institute Earth System Model (MPI-ESM1.2-LR)
  • Atmosphere Model: ECHAM (T63/L95 resolution)
  • Ocean Model: MPIOM (GR1.5/L40 resolution) with HAMOCC
  • Land Surface Model (LSM): JSBACH (version 3.2), including a modified version JSBACH-HTCp (Hydro-Thermodynamic Coupling for permafrost regions).
  • Hydrological Configurations: REF (standard JSBACH), DRY, WET (JSBACH-HTCp variants).
  • Vertical Discretizations: 5-layer (5L, 9.83 m depth), 11-layer (11L, 9.98 m depth), 18-layer (18L, 1391.48 m depth).
• Data sources:
  • CMIP6 ensemble for multi-model comparison (Steinert et al., 2023a).
  • European Space Agency Climate Change Initiative (ESApCCIv3) for observation-constrained reference (Obu et al., 2021).

Main Results

• Deepening the LSM (18L vs. 5L/11L) reduces near-surface soil warming by approximately 0.1 °C per decade at 5 m depth in high radiative forcing scenarios (SSP2-4.5 and SSP5-8.5).
• LSM deepening reduces permafrost extent (PE) retreat by 1.9–3.1 × 10⁶ km² by the end of the 21st century under SSP5-8.5.
• Dry and wet hydrological configurations significantly influence permafrost:
  • Initial permafrost temperatures differ by approximately 3 °C between WET and DRY.
  • Active layer thickness (ALT) shows differences of 1–2 m.
  • Permafrost extent (PE) varies by 5 × 10⁶ km².
• WET simulations exhibit colder mean annual air temperatures (MAATs) than REF (median -9.3 °C vs -8.0 °C), but warmer mean annual ground surface temperatures (MAGST) and mean annual subsurface temperatures at 5 m (MAST 5m) by 1.4 °C and 0.3 °C, respectively, due to enhanced snow insulation.
• The DRY configuration consistently shows the warmest permafrost state.
• Winter offset (ground surface temperature minus surface air temperature in December-January-February) is 10–10.5 °C for WET and DRY, but 8–9 °C lower for REF, attributed to the multi-layer snow scheme and organic layer insulation in WET/DRY.
• The ratio of ground surface temperature (GST) to surface air temperature (SAT) annual cycle amplitude in 1850–1900 is 62–66% for WET/DRY, but 95% for REF.
• DRY simulations show greater ALT deepening during the 21st century, resulting in approximately 20% fewer grid cells retaining permafrost than REF by 2081–2100 under SSP5-8.5.
• Under SSP5-8.5, deep permafrost (ZAA definition) is intensely degraded for 5L and 11L simulations (PE around 2.1–6.9 × 10⁶ km² in 2081–2100), while 18L experiments show less degradation (PE 4.0–9.8 × 10⁶ km²).
• Differences between WET and DRY near-surface PE values correspond to roughly 250–340 petagrams of carbon (Pg C).
• CMIP6 models with LSMs deeper than 40 m exhibit the smallest differences between ZAA and TTOP PE estimates, suggesting more accurate deep permafrost projections.

Contributions

• First comprehensive assessment of the combined impacts of soil thermodynamics and hydrology on Arctic permafrost dynamics within a fully-coupled Earth System Model framework.
• Quantifies the significant role of soil hydro-thermodynamic processes in determining permafrost temperature, active layer thickness, and permafrost extent.
• Demonstrates that the representation of soil hydrology accounts for a substantial portion of inter-model variability in permafrost projections (up to 76% of CMIP6 deep PE interquartile range, 54% for near-surface PE, and approximately 25% of the interdecile range).
• Highlights the critical importance of sufficient Land Surface Model depth for reducing deep permafrost degradation and providing unbiased permafrost extent estimates under global warming, particularly for deep carbon reservoirs like Yedoma.
• Illustrates how model assumptions in permafrost hydrology contribute to uncertainties in current permafrost state knowledge, even in observation-constrained products.

Funding

• GreatModelS (project no. RTI2018-102305-B-C21) from the Spanish Ministry of Science and Innovation (MICINN).
• SMILEME (project no. PID2021-126696OB-C21) from the Spanish Ministry of Science and Innovation (MICINN).
• Scientific Network PolarCSIC, funded by the Spanish Consejo Superior de Investigaciones Científicas (CSIC).
• Deutsches Klimarechenzentrum (DKRZ) for resources under project ID bm1026.
• Félix García-Pereira: contract no. PRE2019-090694 of the MICINN; Ministry for the Ecological Transition and the Demographic Challenge (MITECO); European Commission NextGenerationEU (Regulation EU 2020/2094), through CSIC's Interdisciplinary Thematic Platform Clima (PTI-Clima).
• Philipp de Vrese: European Research Council (ERC) under the European Union's Horizon 2020 research and innovation program (Q-Arctic project, grant agreement no. 951288).
• Article processing charges: CSIC Open Access Publication Support Initiative through its Unit of Information Resources for Research (URICI).

Citation

@article{GarcíaPereira2025Permafrost,
  author = {García‐Pereira, Félix and González‐Rouco, J. Fidel and Meabe-Yanguas, Nagore and Vrese, Philipp de and Steinert, Norman Julius and Jungclaus, Johann and Lorenz, Stephan},
  title = {Permafrost sensitivity to soil hydro-thermodynamics in historical and scenario simulations with the MPI-ESM},
  journal = {˜The œcryosphere},
  year = {2025},
  doi = {10.5194/tc-19-5959-2025},
  url = {https://doi.org/10.5194/tc-19-5959-2025}
}