塔里木盆地暖季极端降水事件多维度特征分析OA
Multi-dimensional characteristic analysis of extreme precipitation events during the warm season in the Tarim Basin
研究极端降水事件演变特征,不仅有助于深入了解气候变化,还可为当地防灾减灾、水利工程及城市建设规划等领域提供借鉴.以塔里木盆地为研究区,基于2005-2025年44个国家气象站暖季(5-10月)逐时降水量数据,采用极值分布拟合及Copula联合函数等方法,研究分析了该区域小时降水及过程降水等事件的时空分布特征、极端事件识别参数及重现期规律.结果表明:(1)基于Hill图法结合超限期望图法给出的阈值具有一定稳健统一性,同时受地形对极端降水事件特征影响,海拔自高至低的站点分别以GPD和LN的极值分布最适宜极端小时降水事件研究.(2)盆地过程降水事件的降水时长与累计降水量存在显著依存关系,采用LN和GEV分布能较好拟合持续时间和累计降水量,Gumbel Copula适合刻画降水持续时间在5h以下的短历时事件,而中长历时事件适宜选择Clayton Copula.(3)盆地极端小时降水事件以超过10 mm的事件重现期较短,普遍在2~5 a,而在50 a一遇事件中,西部高值区的降水量极值放大效应远超东部,空间梯度加剧,潜在风险极高.(4)对于累计降水量≥24 mm的极端过程降水事件,盆地西北部在各重现期下所需持续时间最短,易发短时强降水;西南部风险在10 a重现期内凸显,出现此类事件所需持续时间自5~10 a重现期普遍由10~14 h缩短至6~8 h,致灾性加剧,需重点关注防洪规划.
Investigating the evolution characteristics of extreme precipitation events not only improves our under-standing of climate change but also provides references for local disaster prevention and mitigation,water conser-vancy projects,urban construction planning,and other related fields.This study analyzes the spatiotemporal distri-bution characteristics of hourly and process-based precipitation events,extreme event identification parameters,and recurrence period patterns in the Tarim Basin as the research area.The analysis was performed using methods such as extreme value distribution fitting and Copula joint function based on hourly precipitation data from 44 na-tional meteorological stations during the warm season(May to October)of 2005-2025.The results demonstrate that(1)the thresholds derived by combining the Hill and over-limit expectation graph methods exhibit some de-gree of robust uniformity.Due to the influence of terrain on extreme precipitation characteristics,the generalized Pareto and log-normal(LN)distributions are the most suitable for modeling extreme hourly precipitation events across high-to low-altitude stations.(2)A strong dependence exists between the precipitation duration of process-based precipitation events and their cumulative precipitation.The LN and generalized extreme value distributions provide a better fit for event duration and cumulative precipitation.Gumbel Copula is suitable for characterizing short-duration events,whereas Clayton Copula is suitable for characterizing medium-and long-duration events.(3)Extreme hourly precipitation events in the Tarim Basin have a short recurrence period of over 10 mm,general-ly ranging from 2 to 5 years.However,in 50-year events,the amplification effect of extreme precipitation in the high-value areas in the west significantly exceed those in the east,with an intensified spatial gradient and ex-tremely high potential risks.(4)For extreme process-based precipitation events with cumulative precipitation of≥24 mm,the northwest region requires the shortest duration in each recurrence period and is prone to short-term heavy precipitation.Within a 10-year recurrence period,disaster risks in the southwest have become prominent.The duration required for such events to occur has generally been reduced from 10-14 h under a 5-year recur-rence period to 6-8 h under a 10-year recurrence period,indicating an intensification of disaster severity.There-fore,special attention should be paid to flood control planning.
刁鹏;周雪英;刘长墉;杨柳
巴音郭楞蒙古自治州气象局,新疆 库尔勒 841000||新疆人工影响天气工程技术研究中心,新疆 乌鲁木齐 830002巴音郭楞蒙古自治州气象局,新疆 库尔勒 841000巴音郭楞蒙古自治州气象局,新疆 库尔勒 841000巴音郭楞蒙古自治州气象局,新疆 库尔勒 841000
极端降水Copula联合分布重现期塔里木盆地
extreme precipitationCopula joint distributionreturn periodTarim Basin
《干旱区研究》 2026 (10)
2027-2040,14
新疆维吾尔自治区科技厅基金项目(2023D01F29)新疆维吾尔自治区科技厅天山英才项目(2023TSYCJC0064,2024TSYCJC0046)新疆维吾尔自治区气象局引导性计划项目(YD2024018)
评论