Scholarlink

<p>Accurate precipitation estimation is critical for water resource management, yet it remains a challenge in data-scarce regions. This study develops and evaluates two novel “bottom-up” hydrological models for daily watershed-scale precipitation estimation: an inverted Probability Distributed Model (PDM) and a hybrid Soil Moisture to Rain (SM2RAIN)-Kirchner model. These new structures were systematically compared against multiple SM2RAIN configurations and a suite of benchmark Global Gridded Precipitation Products (GGPPs) over the Walnut Gulch Experimental Watershed, Arizona, USA. The results demonstrate that locally calibrated backward models significantly outperform established GGPPs, with the Kirchner model driven by Soil Moisture Merged via Modified Collocation (SMMC) achieving the highest performance (KGE = 0.62). A key finding was the successful validation of the novel model structures, with the inverted PDM proving to be a robust new approach (KGE = 0.55). Furthermore, the study revealed a critical insight regarding input data: the spatially integrated SMMC product led to a more robust and generalizable model than one driven by in-situ observations, which caused overfitting in some structures. While the backward models excelled in quantitative accuracy, they were less skillful at event detection than the GGPPs, highlighting an important trade-off for application-specific model selection. This work introduces viable new model structures to the field and confirms that high-accuracy precipitation estimates are achievable in data-scarce regions using merged, globally available datasets.</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

Comparing novel backward hydrological models for watershed-scale precipitation estimation: an evaluation of inverted PDM and Kirchner-hybrid structures

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年3月20日

查看全文
  • Pouria Asgari Dastjerdi,
  • Mohsen Nasseri

摘要

Accurate precipitation estimation is critical for water resource management, yet it remains a challenge in data-scarce regions. This study develops and evaluates two novel “bottom-up” hydrological models for daily watershed-scale precipitation estimation: an inverted Probability Distributed Model (PDM) and a hybrid Soil Moisture to Rain (SM2RAIN)-Kirchner model. These new structures were systematically compared against multiple SM2RAIN configurations and a suite of benchmark Global Gridded Precipitation Products (GGPPs) over the Walnut Gulch Experimental Watershed, Arizona, USA. The results demonstrate that locally calibrated backward models significantly outperform established GGPPs, with the Kirchner model driven by Soil Moisture Merged via Modified Collocation (SMMC) achieving the highest performance (KGE = 0.62). A key finding was the successful validation of the novel model structures, with the inverted PDM proving to be a robust new approach (KGE = 0.55). Furthermore, the study revealed a critical insight regarding input data: the spatially integrated SMMC product led to a more robust and generalizable model than one driven by in-situ observations, which caused overfitting in some structures. While the backward models excelled in quantitative accuracy, they were less skillful at event detection than the GGPPs, highlighting an important trade-off for application-specific model selection. This work introduces viable new model structures to the field and confirms that high-accuracy precipitation estimates are achievable in data-scarce regions using merged, globally available datasets.