孙启发^1,,
梁家昌¹,
蒙红卫¹,
王敏²,
黄林培¹,
沈才明^1,,
1. 云南省高原地理过程与环境变化重点实验室, 云南师范大学地理学部, 云南 昆明 650500
2. 云南大学地球科学学院, 云南 昆明 650500

基金项目: 国家自然科学基金项目（批准号：41372191）、云南省高端科技人才引进项目（批准号：2013HA024）、云南省科技计划重点项目（批准号：2014FA005）和云南师范大学研究生科研创新项目（批准号：ysolyjs2019165）共同资助


详细信息

作者简介: 孙启发, 男, 28岁, 博士研究生, 自然地理学专业, E-mail: sunqifa19920708@foxmail.com

通讯作者: 沈才明, E-mail: cmshen@hotmail.com

中图分类号: P467;P426.6

收稿日期:2020-10-02
修回日期:2021-01-04
刊出日期:2021-03-30

Temporal changes of coherent droughts and floods over Yunnan in rainy season during the past five and a half centuries

SUN Qifa^1,,
LIANG Jiachang¹,
MENG Hongwei¹,
WANG Min²,
HUANG Linpei¹,
SHEN Caiming^1,,
1. Yunnan Key Laboratory of Plateau Geographical Processes and Environmental Changes, Faculty of Geography, Yunnan Normal University, Kunming 650500, Yunnan
2. School of Earth Science, Yunnan University, Kunming 650500, Yunnan


More Information

Corresponding author: SHEN Caiming,E-mail:cmshen@hotmail.com

MSC: P467;P426.6

--> Received Date: 02 October 2020
Revised Date: 04 January 2021
Publish Date: 30 March 2021


摘要
摘要:旱涝灾害是云南的主要气象灾害，严重制约了云南社会经济的发展。文章根据云南21个地区来自历史文献资料的旱涝指数、树木年轮重建的帕默尔干旱指数及观测降水数据（1958~2012 A.D.），采用五级旱涝指数分类系统将其归一化并应用反距离加权插值补充历史文献缺失的记录，最终获得21个地区近5个半世纪（1470~2012 A.D.）完整的雨季旱涝指数序列。应用主成分分析对该旱涝指数数据集作分析以揭示云南旱涝的主要空间模态，并应用系统突变检测、小波分析等方法对主要空间模态的得分作分析以探讨其时间变化特征和周期性。结果显示，主成分分析得出的第一主成分（解释总方差的50.9%）揭示了空间一致性模态，说明一致性旱涝（同旱同涝）是云南过去5个半世纪旱涝的主要空间模态。第一主成分得分显示了3个阶段（1470~1680 A.D.、1681~1910 A.D.和1911~2012 A.D.）的旱涝状况，活跃的1470~1680 A.D.出现较多极端、严重旱涝以及持续数年的云南大旱和大涝，平静的1681~1910 A.D.少见旱涝灾害，又复活跃的最近一个世纪（1911~2012 A.D.）见到较多的干旱，在2009~2012出现持续4年的云南大旱。小波分析结果表明，云南旱涝存在3~7 a、10~15 a和30~50 a周期变化，3~7 a周期贯穿近5个半世纪，显著的10~15 a和30~50 a涛动出现在1580~1750 A.D.和1470~1700 A.D.。将其与一系列气候代用指标和大型气候型的比较表明，云南雨季旱涝主要受控于南亚夏季风，而过去5个半世纪的南亚夏季风主要受ENSO和IOD的年际变化及PDO和AMO的年代际变化影响。
关键词: 云南/
5个半世纪(1470~2012 A.D.)/
旱涝指数/
空间模态/
主成分分析/
小波分析

Abstract:Droughts and floods are major meteorological disasters, severely restricting social and economic development in Yunnan. Based on proxy data of rainy season from historical documents, reconstructed June-August Palmer drought severity index(PDSI) from tree-ring, and observational rainy season rainfall (1958~2012 A.D.) in 21 regions of Yunnan (21°08'32″~29°15'08″N, 97°31'39″~106°11'47″E), five-grade dryness/wetness index(DWI) classification system was used to normalize these data and an inverse distance weighted interpolation was used to fill those missing data in historical document records, thus a dataset of complete rainy season DWI sequences covering nearly five and a half centuries(1470~2012 A.D.) were finally obtained for 21 regions of Yunnan. This DWI dataset was analyzed using principal component analysis(PCA) to reveal principal components(PCs), i.e. major spatial modes of drought and flood in Yunnan, and analyses on the score time series of PCs using regime shift detection and wavelet analysis were conducted to determine their variation characteristics and periodicities. The results showed that the first principal component(PC1) explaining 50.9% of total variance revealed a coherent spatial mode, indicating that the coherent drought or flood was major spatial mode over Yunnan during the last five and a half centuries. The PC1 scores showed a temporal pattern of three stages of drought and flood, i.e. active 1470~1680 A.D.(the 1^st stage), when Yunnan saw many exceptional and severe droughts and floods as well as mega-droughts and mega-floods persisting for several years; inactive 1681 A.D. to 1910 A.D.(the 2^nd stage), when few droughts and floods occurred in Yunnan; and active last century(1911~2012 A.D., the 3^rd stage), when Yunnan saw more droughts, and even experienced a mega-drought lasting four years from 2009 A.D. to 2012 A.D. The result of wavelet analysis showed three periodic bands at 3~7 a, 10~15 a and 30~50 a significant at 95% confidence level. The 3~7 a cycle occurred through nearly five and a half centuries, whereas 10~15 a and 30~50 a oscillations appeared significantly in 1580~1750 A.D. and 1470~1700 A.D., respectively. A comparison of PC1 scores with a series of climatic proxies and large-scale climatic patterns indicated that coherent droughts and floods in rainy season over Yunnan were mainly controlled by the South Asian summer monsoon, which was affected by the interannual variation of ENSO and IOD as well as the interdecadal variation of PDO and AMO.
Key words:Yunnan/
the past five and a half centuries(1470~2012 A.D.)/
dryness/wetness index/
spatial mode/
principal component analysis/
wavelet analysis



PDF全文下载地址:

http://www.dsjyj.com.cn/data/article/export-pdf?id=60768b47c23e6710c26d5b60