Scholarlink

<p>External forcings have left discernible fingerprints on precipitation patterns and intensity. Traditionally, detection of these signals has relied on averaging precipitation across time (e.g., annual or seasonal means) and space (such as across longitudes) to enhance the signal and reduce the noise. Here, we investigate whether externally-forced signals can be detected in precipitation at the finer monthly timescale, while preserving full spatial resolution. We systematically evaluate multiple pattern-based fingerprint methods across diverse observational datasets, considering different processing specifications, such as pooling monthly data versus analyzing months separately, and using either raw or standardized anomalies. Our results show that robust externally-forced signals (e.g., with a signal-to-noise ratio &#xa0;&gt;&#xa0;2) have already emerged across diverse monthly precipitation datasets and that normalizing anomalies by local variability facilitates signal detection without spatial aggregation. By isolating externally-forced changes in precipitation from internal variability at finer spatial and temporal scales, our fingerprints reveal seasonally-varying forced patterns consistent with, yet more nuanced than, the “wet-get-wetter and dry-get-drier” paradigm, with most tropical regions becoming wetter and many subtropics becoming drier; however, these patterns vary across regions, seasons, and models, highlighting the need of region-specific assessments for effective water resource planning.</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

Enhanced detectability of forced signal in monthly precipitation changes

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月31日

查看全文
  • Shiheng Duan,
  • Céline Bonfils,
  • Jia-Rui Shi

摘要

External forcings have left discernible fingerprints on precipitation patterns and intensity. Traditionally, detection of these signals has relied on averaging precipitation across time (e.g., annual or seasonal means) and space (such as across longitudes) to enhance the signal and reduce the noise. Here, we investigate whether externally-forced signals can be detected in precipitation at the finer monthly timescale, while preserving full spatial resolution. We systematically evaluate multiple pattern-based fingerprint methods across diverse observational datasets, considering different processing specifications, such as pooling monthly data versus analyzing months separately, and using either raw or standardized anomalies. Our results show that robust externally-forced signals (e.g., with a signal-to-noise ratio  > 2) have already emerged across diverse monthly precipitation datasets and that normalizing anomalies by local variability facilitates signal detection without spatial aggregation. By isolating externally-forced changes in precipitation from internal variability at finer spatial and temporal scales, our fingerprints reveal seasonally-varying forced patterns consistent with, yet more nuanced than, the “wet-get-wetter and dry-get-drier” paradigm, with most tropical regions becoming wetter and many subtropics becoming drier; however, these patterns vary across regions, seasons, and models, highlighting the need of region-specific assessments for effective water resource planning.