Spatiotemporal assessment and climate teleconnections of drought in Northeast China (2001–2023) using a machine-learning-based meteorological composite index

IF 5 2区地球科学 Q1 WATER RESOURCES

Journal of Hydrology-Regional Studies Pub Date : 2025-08-13 DOI:10.1016/j.ejrh.2025.102693

Yanran Qin , Zhijie Zhang , Guihong Wang , Jintong Ren , Wanchang Zhang

{"title":"Spatiotemporal assessment and climate teleconnections of drought in Northeast China (2001–2023) using a machine-learning-based meteorological composite index","authors":"Yanran Qin , Zhijie Zhang , Guihong Wang , Jintong Ren , Wanchang Zhang","doi":"10.1016/j.ejrh.2025.102693","DOIUrl":null,"url":null,"abstract":"<div><h3>Study region</h3><div>Northeast China (NEC)</div></div><div><h3>Study focus</h3><div>This study proposes a machine learning framework for spatiotemporal drought assessment in NEC (2001–2023), integrating the Light Gradient Boosting Machine (LightGBM) and the Meteorological Composite Index (MCI). Using 14 drought-related predictors at 0.0059° spatial resolution, the model achieved strong predictive performance (R² = 0.8942; drought level consistency = 91.26 %). Drought characteristics were analyzed through frequency, intensity, and centroid metrics. Drought-climate teleconnections were explored using lagged correlation and wavelet methods.</div></div><div><h3>New hydrological insights for the region</h3><div>Mild droughts predominated (78.47 %), with clear seasonality peaking in March–April. A significant wetting trend (p < 0.05) was observed across 84.11 % of the region, though persistent hotspots remained in the southwest, likely due to rain shadow effects (R² = 0.67 with elevation). The drought centroid showed a northward shift before 2010 and a southward shift afterward, suggesting decadal-scale reversals. The North Pacific Index (NPI) was the leading climate driver, with strong coherence at 10–15-month scales and a 5-month lead (r = -0.64, p < 0.01), supporting early warning potential. Multivariate climate indices outperformed single indices in explaining drought variability, with the trivariate combination of the Atlantic Multidecadal Oscillation (AMO), Pacific–North America pattern (PNA), and NPI providing the highest explanatory power. These results provide new insights into drought evolution and climate-drought interactions in NEC.</div></div>","PeriodicalId":48620,"journal":{"name":"Journal of Hydrology-Regional Studies","volume":"61 ","pages":"Article 102693"},"PeriodicalIF":5.0000,"publicationDate":"2025-08-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Hydrology-Regional Studies","FirstCategoryId":"89","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2214581825005221","RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"WATER RESOURCES","Score":null,"Total":0}

引用次数: 0

Abstract

Study region

Northeast China (NEC)

Study focus

This study proposes a machine learning framework for spatiotemporal drought assessment in NEC (2001–2023), integrating the Light Gradient Boosting Machine (LightGBM) and the Meteorological Composite Index (MCI). Using 14 drought-related predictors at 0.0059° spatial resolution, the model achieved strong predictive performance (R² = 0.8942; drought level consistency = 91.26 %). Drought characteristics were analyzed through frequency, intensity, and centroid metrics. Drought-climate teleconnections were explored using lagged correlation and wavelet methods.

New hydrological insights for the region

Mild droughts predominated (78.47 %), with clear seasonality peaking in March–April. A significant wetting trend (p < 0.05) was observed across 84.11 % of the region, though persistent hotspots remained in the southwest, likely due to rain shadow effects (R² = 0.67 with elevation). The drought centroid showed a northward shift before 2010 and a southward shift afterward, suggesting decadal-scale reversals. The North Pacific Index (NPI) was the leading climate driver, with strong coherence at 10–15-month scales and a 5-month lead (r = -0.64, p < 0.01), supporting early warning potential. Multivariate climate indices outperformed single indices in explaining drought variability, with the trivariate combination of the Atlantic Multidecadal Oscillation (AMO), Pacific–North America pattern (PNA), and NPI providing the highest explanatory power. These results provide new insights into drought evolution and climate-drought interactions in NEC.

查看原文本刊更多论文

基于机器学习的气象综合指数2001-2023年东北干旱时空评价与气候遥相关

基于光梯度增强机（Light Gradient Boosting machine, LightGBM）和气象综合指数（Meteorological Composite Index， MCI），提出了东北地区2001-2023年干旱时空评价的机器学习框架。使用0.0059°空间分辨率下的14个干旱相关预测因子，模型获得了较强的预测性能(R²= 0.8942；干旱水平一致性= 91.26 %)。通过频率、强度和质心指标分析干旱特征。利用滞后相关和小波分析方法探讨了干旱与气候的遥相关关系。轻度干旱占主导地位（78.47 %），3 - 4月为明显的季节性高峰。在84.11% %的区域观测到显著的湿润趋势（p <； 0.05），尽管西南地区仍然存在持续的热点，可能是由于雨影效应（R²= 0.67随海拔升高）。干旱质心在2010年之前向北移动，之后向南移动，表明十年尺度的逆转。北太平洋指数（NPI）是主要的气候驱动因素，在10 - 15个月尺度上具有很强的一致性和5个月的领先性（r = -0.64,p <； 0.01），支持早期预警潜力。多变量气候指数在解释干旱变率方面优于单一指数，其中大西洋多年代际振荡（AMO）、太平洋-北美模式（PNA）和NPI的三变量组合解释能力最强。这些结果为东北地区干旱演变和气候-干旱相互作用提供了新的见解。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Journal of Hydrology-Regional Studies Earth and Planetary Sciences-Earth and Planetary Sciences (miscellaneous)

CiteScore

6.70

自引率

8.50%

发文量

284

审稿时长

60 days

期刊介绍： Journal of Hydrology: Regional Studies publishes original research papers enhancing the science of hydrology and aiming at region-specific problems, past and future conditions, analysis, review and solutions. The journal particularly welcomes research papers that deliver new insights into region-specific hydrological processes and responses to changing conditions, as well as contributions that incorporate interdisciplinarity and translational science.