| 摘要: |
| 【目的】 采用修正Mann-Kendal法，进行年径流序列的趋势识别，以期为陕北地区的水资源利用与管理提供依据。【方法】 采用Mann-Kendal法，对年径流序列的统计变量和方差、秩次序列的显著自相关系数、年径流序列的方差修正因子和修正方差及修正统计量进行计算。依据陕北地区11个测站的年径流资料，采用传统的参数、非参数检验法和修正Mann-Kendall法,进行了年径流序列趋势识别的实例研究。【结果】 神木站、刘家河站和枣园站的年径流序列存在显著的正自相关系数，绥德站和黄陵站的年径流序列存在显著的负自相关系数；神木站的年径流量呈现明显的下降趋势，绥德站、交口河站、刘家河站、志丹站、张村驿站、枣园站和黄陵站的年径流量略有下降趋势，吴旗站和安塞站的年径流量略有上升趋势，而杏河站的年径流量无明显变化。【结论】 修正Mann-Kendall法充分考虑了序列相关性对趋势检验的影响，是年径流趋势识别的一种新途径。 |
| 关键词: 年径流 趋势识别 自相关系数 传统趋势检验法 修正Mann-Kendall法 |
| DOI： |
| 分类号: |
| 基金项目:国家自然科学基金项目（50579065，50879070）；西北农林科技大学优秀博士论文基金项目 |
|
| Identification of annual runoff series trend |
|
|
| Abstract: |
| 【Objective】 This paper adopts modified Mann Kendall method to identify the annual runoff time series trend,in order to provide the basis for Shaanbei area water resources utilization and management.【Method】 First of all,Mann Kendal method was used to calculate annual runoff series variance and statistical variable;Secondly,the significant autocorrelation coefficient of rank series were calcultated;so were the modified factor variance and modified variance of annual runoff series;Finally,modified statistical variable were measured.Based on 11 stations in Shaanbei area annual runoff data,the traditional parameters,non parametric test and modified Mann Kendall method were used to identify the trend of annual runoff serieses.【Result】 Shenmu station,Liujiahe station and Zaoyuan station annual runoff series have significantly positive autocorrelation coefficient.Suide station and Huangling station annual runoff series have significantly negative autocorrelation coefficient.Shenmu station annual runoff series showed clear downward trend.Suide Station,Jiaokouhe Station,Liujiahe Station,Zhidan station,Zhangcunyi station,Zaoyuan station and Huangling station annual runoff series showed slight downward trend in annual runoff. Wuqi station and Ansai station showed slight upward trend in annual runoff.Xinghe station have no significant change in annual runoff.【Conclusion】 Modified Mann Kendall method which fully takes into account the effect of hydrological time serial autocorrelation is a new method for the time serial trend identification. |
| Key words: annual runoff trend identification autocorrelation coefficient traditional trend test method modified Mann Kendall method |
