Scholarlink

<p>To reliably assess future hydrological risks and support various water resource management applications, hydrological models require long-term, fine-scale precipitation time series that incorporate climate change. This study proposes a novel stochastic framework for generating 5-minute future precipitation time series by strategically minimizing reliance on regional climate models (RCMs). The methodology extracts only the change factor for mean precipitation from RCM outputs and integrates it with robust statistical relationships derived from high-resolution observed data across multiple neighboring stations. These future statistics are then used to calibrate the Shufflex-2 model, an advanced Poisson cluster model, to synthesize the final time series. The framework was applied to 14 sites across Germany and South Korea to investigate future changes in precipitation characteristics. The results reveal stark regional contrasts. Germany is projected to experience a statistically significant drying trend, particularly in July, with a robust decrease in both mean precipitation and the proportion of wet periods at fine timescales. Conversely, in South Korea, projected changes do not emerge as statistically significant trends, indicating they remain within the bounds of the region’s strong natural long-term variability. Furthermore, the analysis of annual maximum precipitation yields complex signals; some regions in Germany show a significant decreasing trend in extremes, a counter-intuitive but physically consistent result of a more intermittent future precipitation regime. In South Korea, no statistically significant trend for extremes was found. This study highlights the critical importance of employing rigorous statistical tests that account for long-range dependence to distinguish robust climate change signals from the inherent variability of precipitation processes.</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

Stochastic poisson cluster simulation of sub-hourly precipitation under climate change in Germany and South Korea

  • 期刊论文

  • 发表日期：

    2026年2月25日

查看全文
  • Jeongha Park,
  • Ivan Vorobevskii,
  • Dongkyun Kim,
  • Rico Kronenberg,
  • Christian Onof,
  • Davide Luciano De Luca

摘要

To reliably assess future hydrological risks and support various water resource management applications, hydrological models require long-term, fine-scale precipitation time series that incorporate climate change. This study proposes a novel stochastic framework for generating 5-minute future precipitation time series by strategically minimizing reliance on regional climate models (RCMs). The methodology extracts only the change factor for mean precipitation from RCM outputs and integrates it with robust statistical relationships derived from high-resolution observed data across multiple neighboring stations. These future statistics are then used to calibrate the Shufflex-2 model, an advanced Poisson cluster model, to synthesize the final time series. The framework was applied to 14 sites across Germany and South Korea to investigate future changes in precipitation characteristics. The results reveal stark regional contrasts. Germany is projected to experience a statistically significant drying trend, particularly in July, with a robust decrease in both mean precipitation and the proportion of wet periods at fine timescales. Conversely, in South Korea, projected changes do not emerge as statistically significant trends, indicating they remain within the bounds of the region’s strong natural long-term variability. Furthermore, the analysis of annual maximum precipitation yields complex signals; some regions in Germany show a significant decreasing trend in extremes, a counter-intuitive but physically consistent result of a more intermittent future precipitation regime. In South Korea, no statistically significant trend for extremes was found. This study highlights the critical importance of employing rigorous statistical tests that account for long-range dependence to distinguish robust climate change signals from the inherent variability of precipitation processes.