Scholarlink

<p>Near-surface atmospheric moisture is a key component of the climate and environment systems, exerting significant influences on both nature and human beings. However, existing moisture data are often limited by sparse observations and low spatial/temporal resolution, which restricts their applicability at fine scales, particularly in populated and urbanized regions with strong moisture variability, such as the North China Plain (NCP). Here, we construct a high-resolution (daily and 1 km) near-surface atmospheric moisture index collection comprising six different indicators over the NCP during 2003–2020 (HiMIC-NCP). HiMIC-NCP is generated by the Light Gradient Boosting Machine&#xa0;(LightGBM) algorithm by integrating meteorological observations and multiple covariates, including 2-meter air temperature, land surface temperature, water vapor, topography, and population density. The dataset exhibits a high accuracy with <i>R</i>² values ranging from 0.879 to 0.988, and mean absolute error and root mean square error remaining within reasonable ranges. The dataset also exhibits high consistency with ground observations across spatial and temporal regimes, demonstrating its robustness and reliability, and thereby provides a high-quality foundation for fine-scale climate change assessment, agricultural management, and public health studies.</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

A high-resolution (daily and 1 km) atmospheric moisture collection over the North China Plain during 2003–2020

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年4月27日

查看全文
  • Zhiying Su,
  • Hui Zhang,
  • Xiang Li,
  • Sijia Wu,
  • Manqing Shi,
  • Ming Luo

摘要

Near-surface atmospheric moisture is a key component of the climate and environment systems, exerting significant influences on both nature and human beings. However, existing moisture data are often limited by sparse observations and low spatial/temporal resolution, which restricts their applicability at fine scales, particularly in populated and urbanized regions with strong moisture variability, such as the North China Plain (NCP). Here, we construct a high-resolution (daily and 1 km) near-surface atmospheric moisture index collection comprising six different indicators over the NCP during 2003–2020 (HiMIC-NCP). HiMIC-NCP is generated by the Light Gradient Boosting Machine (LightGBM) algorithm by integrating meteorological observations and multiple covariates, including 2-meter air temperature, land surface temperature, water vapor, topography, and population density. The dataset exhibits a high accuracy with R² values ranging from 0.879 to 0.988, and mean absolute error and root mean square error remaining within reasonable ranges. The dataset also exhibits high consistency with ground observations across spatial and temporal regimes, demonstrating its robustness and reliability, and thereby provides a high-quality foundation for fine-scale climate change assessment, agricultural management, and public health studies.