地理学报 ›› 2008, Vol. 63 ›› Issue (7): 769-779.doi: 10.11821/xb200807011

• 气候变化 • 上一篇    下一篇

新疆不同区域洪灾受灾面积变化趋势及多尺度分析

王秋香1, 崔彩霞2, 姚艳丽3   

  1. 1. 新疆气象信息中心, 乌鲁木齐830002;
    2. 乌鲁木齐沙漠气象研究所, 乌鲁木齐830002;
    3. 新疆气候中心, 乌鲁木齐830002
  • 收稿日期:2008-01-15 修回日期:2008-03-06 出版日期:2008-07-25 发布日期:2008-07-25
  • 作者简介:王秋香(1961-), 女, 新疆乌鲁木齐人, 高级工程师, 1983 年毕业于南京气象学院获学士学位, 现主要从事气候 变化、气象灾害及气象资源研究。E-mail: wqx188@126.com
  • 基金资助:

    中国沙漠气象科学研究基金(sqj2005004); 国家自然科学基金项目(40475041) 资助

Var iation Tr ends and Multi- scale Analysis of Flood Affected Area in Var ious Regions of Xinjiang

WANG Qiuxiang1, CUI Caixia2, YAO Yanli3   

  1. 1. Xinjiang Meteorological Information Centre, Urumqi 830002, China;
    2. Institute of Desert Meteorology, CMA, Urumqi 830002, China;
    3. Xinjiang Climate Center, Urumqi 830002, China
  • Received:2008-01-15 Revised:2008-03-06 Online:2008-07-25 Published:2008-07-25
  • Supported by:

    Foundation of Desert Meteorology, CMA, No.sqj2005004; National Natural Science Foundation of China, No.40475041

摘要:

选用新疆1950-2006 年84 个县2038 份洪水灾害资料, 采用灰色关联评估模型以县为单位将新疆分为重洪灾区、次重洪灾区、一般洪灾区、小洪灾区、微洪灾区。采用自回归滑 动平均模型ARMA (p, q) 与二阶主值函数分析受灾面积的长期变化趋势, 并用morlet 小波 分析其振荡周期发现, 全疆及各区洪灾受灾面积基本上存在12~15a 的周期变化, 1980 年代之后处于气候背景上的年代际与年际尺度的洪灾频发振荡期, 7~8a、2~3a 的小周期振荡叠加比较频繁。从新疆洪灾变化趋势上说, 前景不容乐观, 因为总体上洪灾受灾面积呈现显著的增加趋势, 目前各洪灾区仍处于频繁发生期的气候变化背景中。天山南北麓洪灾受灾面积的 分布趋势和周期振荡对新疆洪灾受灾面积的年际变化趋势和周期振荡起了决定性作用, 是增加防洪设施和防洪力度的重点地区。另外, 一般洪灾区2000 年以来对全疆洪水受灾面的影响也较大, 而且未来无论从短中长周期变化都有增加趋势。

关键词: 新疆, 洪灾, 受灾面积, 小波分析

Abstract:

In this paper, we selected and used 2038 flood disaster data from 84 counties of Xinjiang in 1950-2006, and adopted the gray relational evaluation model, taking county as a unit, to divide Xinjiang into heavy flood disaster subregion, secondary heavy flood disaster subregion, general flood disaster subregion, light disaster flood subregion and tiny flood subregion. We also adopted autoregression moving average model ARMA(p, q) and second order principal value function to analyze long-term trend of disaster affected area, as well as its oscillation cycle with Morlet wavelet. Results show that the flood affected area throughout the region and every subregion had a changing cycle of 12-15a, the flood frequent oscillation cycle was on interdecadal and interannual scale of climatic background after the 1980s, and the superposition of small oscillation cycle as 7-8a and 2-3a was comparatively frequent. In terms of the flood disaster variation trend of Xinjiang, the outlook is not optimistic for the general flood affected area presented a distinct increasing trend and each flood subregion was in the climatic background with frequent period of flood at present. The distributing trend of flood disaster affected area and the oscillation cycle in the north and south piedmonts of the Tianshan Mountains played a decisive role relating to the interannual long-term trend and its oscillation cycle in the flood affected area in Xinjiang. These regions were the keys in reinforcing facilities and strength for flood prevention and control. Otherwise, the general flood affected area would have a greater impact on the flooded area of entire Xinjiang, leading to an increasing trend in the future regardless of the short, medium or long period changes, hence great attention should be aroused.

Key words: Xinjiang, flood, flood affected area, wavelet analysis