摘要
为了研究及预测娘子关泉域降水及径流的变化趋势,采用时间序列分析方法,通过SPSS软件建立ARIMA模型对降水和径流序列进行分析,并预测后续11年的变化趋势,得出预测值的95%置信区间.结果表明,ARIMA(6,1,0)模型是年降水序列的有效拟合模型,ARIMA(10,2,0)模型是年径流序列的有效拟合模型.相比娘子关泉域的年降水序列,年径流序列的下降趋势更为显著,且从2017年开始,年径流序列预测值的95%置信区间开始包含负值,说明可能受人为因素的影响,需要引起人们的重视.
In order to study and predict the trend of precipitation and runoff in Niangziguan Spring area,the time series analysis method was used to establish the ARIMA model by SPSS software to analyze the precipitation and runoff sequences,and predict the trend of the next 11 years.The 95%confidence interval of the predicted value was obtained.The results showed that the ARIMA(6,1,0)model was an effective fitting model for the annual precipitation sequence,and the ARIMA(10,2,0)model was an effective fitting model for the annual runoff sequence.The downward trend of the annual runoff series was more significant compared with the annual precipitation sequence of the Niangziguan Spring area,and since 2017,the 95%confidence interval of the annual runoff series prediction value begins to contain negative values,indicating that it may be affected by human factors and needs to be paid attention to.
作者
王永珂
范永辉
WANG Yong-ke;FAN Yong-hui(School of Mathematical Science,Tianjin Normal University,Tianjin 300387,China)
出处
《哈尔滨商业大学学报（自然科学版）》
CAS
2019年第3期360-366,共7页
Journal of Harbin University of Commerce：Natural Sciences Edition
