1956—2018年青海省黄河流域降水变化特征分析及预测

本研究基于1956—2018年31个气象站月值降水数据,采用M-K检验、Pettitt检验、Morlet小波分析、ARIMA模型等方法,分析青海省黄河流域近63年包括趋势、突变、周期在内的降水量特性以及验证未来5年降水趋势预测变化的合理性。结果表明:(1)近63年青海省黄河流域降水具有集中程度高、年内分配不均、丰枯季明显的特点;不同年代的月均降水量均集中在4—10月,且未来月均降水量有上升的变化趋势。(2)年降水量呈显著增长趋势,增长率为(9.07 mm/10 a),春季和冬季降水增长趋势显著,夏季和秋季降水无显著的增长趋势,2004年为该流域大的降水转折年。(3)年降水时间序列存在23~32 a,15~20 a,9~13 a以及4~6 a的周期变化规律,四级降水主周期分别对应30、15、11和6 a时间尺度。(4)ARIMA(2,1,4)模型能够较好地拟合1956—2018年降水序列并对2019—2023年降水数据合理预测;线性回归及M-K检验分析结果显示1956—2023年降水序列呈显著增长趋势,增长率为(9.22 mm/10 a),与趋势预测结果相一致;ARIMA(2,1,4)模型可以对青海省黄河流域进行短期年降水量预测,为当地水资源合理规划和管理提供参考。

Analysis of Precipitation Change Characteristics and Precipitation Forecast in the Yellow River Basin of Qinghai Province from 1956 to 2018

Based on the monthly precipitation data of 31 weather stations from 1956 to 2018,this study uses Mann-Kendall test,Pettitt test,wavelet analysis,ARIMA model to analyze the trend,sudden change,and periodic precipitation characteristics of the Yellow River Basin in Qinghai Province from 1956 to 2018 and verifies the rationality of precipitation trend forecast from 2019 to 2023.The result shows(1)Precipitation in the Yellow River Basin of Qinghai Province has the characteristics of high concentration,uneven distribution during the year,and obvious wet and dry seasons.The monthly average precipitation in different years is concentrated on April to October,and the average monthly precipitation will increase in the future.(2)From 1956 to 2018,the annual precipitation show a significant growth trend,with a growth rate of(9.07 mm/10 a),spring and winter precipitation increase significantly,summer and autumn precipitation show no significant growth trend;2004 is a major precipitation turning year in the basin.(3)The annual precipitation time series has periodic changes of 23~32 a,15~20 a,9~13 a and 4~6 a.As a result,the four main periods of precipitation correspond to the time scales of 30,15,11 and 6 a.(4)ARIMA(2,1,4)model can better fit the precipitation series from 1956 to 2018,and reasonably predict precipitation from 2019 to 2023;the results of linear regression and Mann-Kendall test analysis show that the precipitation series from 1956 to 2023 has a significant increase trend,the growth rate is(9.22 mm/10 a),which is consistent with the trend prediction result;ARIMA(2,1,4)model can predict the short-term annual precipitation in the Yellow River Basin of Qinghai Province and can provide a reference for the reasonable planning and management of local water resources.