Scholarlink

<p>Permafrost, which underlies roughly one-fifth of global land area, is highly vulnerable to ongoing climate warming and its degradation has important implications for hydrological processes, carbon cycling, ecosystems, and infrastructure stability. It remains challenging to accurately project its future distribution due to various factors. The Surface Frost Index (SFI) threshold is a critical parameter that captures first-order thermal forcing for projecting future permafrost distribution. Here, the future global annual frozen ground distribution datasets are produced using an optimized frost number threshold (<i>F</i><sub><i>at</i></sub>) calibrated via kappa coefficient-based accuracy assessment. By comparing simulated outputs with benchmark maps, we identify the optimal <i>F</i><sub><i>at</i></sub>. Applying this threshold to downscaled CMIP6 data (0.25° resolution), we generate annual frozen ground distributions (2020–2099) under four SSP scenarios (SSP126, SSP245, SSP370, SSP585). The datasets reveal accelerated permafrost degradation, with mid-century (2040–2060) losses of 19 ± 3% to 28 ± 3% and late-century (2080–2099) losses escalating to 21 ± 3% to 61 ± 5%, peaking under SSP585. These high-resolution projections provide critical insights for assessing climate impacts and guiding cryospheric adaptation strategies.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Future global annual frozen ground distribution datasets based on Frost Number Model with Kappa coefficient

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年2月28日

查看全文
  • Xiaoduo Pan,
  • Hu Li,
  • Xiaowei Nie

摘要

Permafrost, which underlies roughly one-fifth of global land area, is highly vulnerable to ongoing climate warming and its degradation has important implications for hydrological processes, carbon cycling, ecosystems, and infrastructure stability. It remains challenging to accurately project its future distribution due to various factors. The Surface Frost Index (SFI) threshold is a critical parameter that captures first-order thermal forcing for projecting future permafrost distribution. Here, the future global annual frozen ground distribution datasets are produced using an optimized frost number threshold (Fat) calibrated via kappa coefficient-based accuracy assessment. By comparing simulated outputs with benchmark maps, we identify the optimal Fat. Applying this threshold to downscaled CMIP6 data (0.25° resolution), we generate annual frozen ground distributions (2020–2099) under four SSP scenarios (SSP126, SSP245, SSP370, SSP585). The datasets reveal accelerated permafrost degradation, with mid-century (2040–2060) losses of 19 ± 3% to 28 ± 3% and late-century (2080–2099) losses escalating to 21 ± 3% to 61 ± 5%, peaking under SSP585. These high-resolution projections provide critical insights for assessing climate impacts and guiding cryospheric adaptation strategies.