Scholarlink

<p>Urban heat island (UHI) intensity may interact with extreme precipitation, but its directional and seasonal linkages remain insufficiently understood. This study examines the causal relationship between near-surface UHI intensity and extreme precipitation events (EPEs) in Beijing from 2000 to 2025. Hourly 2-m air temperature and precipitation were obtained from ERA5-Land, and urban and rural grid cells were defined using a GAIA-based built-up fraction mask. UHI was calculated as the hourly urban–rural 2-m air temperature difference, while EPEs were identified as hours when urban mean precipitation exceeded the full-period 99th percentile threshold. Monthly UHI and EPE series were analyzed using Convergent Cross Mapping (CCM) with Fourier-based surrogate testing. The full-period results show a significant nonlinear causal signal from EPE to UHI (ρ = 0.694, surrogate mean = 0.424, <i>p</i> = 0.0099), whereas the UHI to EPE direction was not significant (ρ = 0.435, surrogate mean = 0.458, <i>p</i> = 0.57). Seasonal analysis reveals asymmetric coupling: the warm season shows high terminal cross-map skill in both directions, with a slightly stronger UHI-to-EPE tendency, while the cold season and full-period surrogate results are more consistent with a precipitation-feedback pathway from EPE to UHI. These findings indicate that the UHI–EPE relationship is nonlinear and seasonally dependent.</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

Seasonally asymmetric causal linkages between urban heat island intensity and extreme precipitation in Beijing: evidence from convergent cross mapping

  • 期刊论文

  • 发表日期：

    2026年7月24日

查看全文
  • Yi Tang,
  • Ming Li,
  • Longsheng Huang

摘要

Urban heat island (UHI) intensity may interact with extreme precipitation, but its directional and seasonal linkages remain insufficiently understood. This study examines the causal relationship between near-surface UHI intensity and extreme precipitation events (EPEs) in Beijing from 2000 to 2025. Hourly 2-m air temperature and precipitation were obtained from ERA5-Land, and urban and rural grid cells were defined using a GAIA-based built-up fraction mask. UHI was calculated as the hourly urban–rural 2-m air temperature difference, while EPEs were identified as hours when urban mean precipitation exceeded the full-period 99th percentile threshold. Monthly UHI and EPE series were analyzed using Convergent Cross Mapping (CCM) with Fourier-based surrogate testing. The full-period results show a significant nonlinear causal signal from EPE to UHI (ρ = 0.694, surrogate mean = 0.424, p = 0.0099), whereas the UHI to EPE direction was not significant (ρ = 0.435, surrogate mean = 0.458, p = 0.57). Seasonal analysis reveals asymmetric coupling: the warm season shows high terminal cross-map skill in both directions, with a slightly stronger UHI-to-EPE tendency, while the cold season and full-period surrogate results are more consistent with a precipitation-feedback pathway from EPE to UHI. These findings indicate that the UHI–EPE relationship is nonlinear and seasonally dependent.