欢迎访问《中国农学通报》,

中国农学通报 ›› 2017, Vol. 33 ›› Issue (9): 94-100.doi: 10.11924/j.issn.1000-6850.casb16080024

所属专题: 农业气象

• 资源 环境 生态 土壤 气象 • 上一篇    下一篇

石羊河流域暴洪灾害特征及预测方法探讨

罗晓玲,王荣喆,杨梅   

  1. 1.甘肃省武威市气象局,1.甘肃省武威市气象局,1.甘肃省武威市气象局
  • 收稿日期:2016-08-04 修回日期:2016-11-21 接受日期:2016-11-23 出版日期:2017-03-28 发布日期:2017-03-28
  • 通讯作者: 罗晓玲
  • 基金资助:
    2016 年甘肃省气象局科技项目“基于物联网的智能温室控制系统推广应用”(GSMACg2016-01)。

Flash Flood in Shiyang River Basin: Disaster Characteristics and Forecasting Method

  • Received:2016-08-04 Revised:2016-11-21 Accepted:2016-11-23 Online:2017-03-28 Published:2017-03-28

摘要: 为了揭示干旱区局地暴洪灾害的分布特征并对其预测预报方法进行探讨,利用石羊河流域的降水、暴洪灾害和灾情等资料,分析了该流域暴洪灾害分布特征,构建了灾害风险评估模型。结果表明:该流域暴洪灾害呈逐年上升之趋势;灾害高风险区主要分布在流域上游沿祁连山一线高海拔地带以及以乌鞘岭为界的南北两侧,从上游天祝县到下游民勤县风险逐渐降低;暴洪主要集中在4-9月,其中7月最多,占总灾害次数的42.6%;暴洪出现的主要时间与强降水出现的时间基本吻合;生态脆弱的凉州区灾情最严重。尝试利用NCEP/NCAR再分析资料、常规要素资料和气候系统监测指数资料,通过逐步消空法和最优子集回归法从短期、中长期降水趋势及大降水场次预测入手,对暴洪趋势进行潜势预报,通过评分检验,效果良好,基本满足业务应用。

关键词: 水稻主茎叶色, 水稻主茎叶色, SPAD值, RGB组分, 模拟模型

Abstract: To reveal the distribution of local flash flood disasters in arid areas, and discuss its forecasting methods, the authors analyzed the characteristics of flash flood disaster and constructed disaster risk evaluation model of the Shiyang River Basin by using its precipitation, flash flood disasters and disaster situation data. The results showed that: the flash flood disaster presented an increasing trend, and its high risk areas were mainly distributed in the Qilian Mountains line and the Wushaoling’s north and south sides in the upstream, the risks from Tianzhu upstream to Minqin downstream gradually reduced; flash flood mainly concentrated in April-September, and that in July accounted for 42.6% of the total number of occurrences; the occurrence time of flash flood and heavy precipitation was consistent; the ecological fragile Liangzhou was the most serious disaster region. Good forecast effect could be achieved by using NCEP/NCAR reanalysis data, conventional elements and the climate monitoring index, with stepwise decreasing FAR and the optimal subset regression method to forecast precipitation and flash flood trend from short-term, medium and long-term precipitation trends and heavy rain date.