利用松花江流域上的35个台站1956—2015年气温资料,分别采用一元线性回归法、M-K突变检验法和Morlet小波周期法研究了松花江流域近60年的年平均气温、年平均最高气温和年平均最低气温时空变化特征、突变特征和周期特征。结果 表明:近60年来松花江流域气温整体呈现显著上升趋势,且存在明显的增温高值区; 年平均气温增幅为0.322℃/10 a,年平均最高气温和年平均最低气温分别为0.190,0.429℃/10 a; 气温在1987—1992年发生突变,突变后的气温变化趋势呈现明显的上升; 松花江流域3种气温序列均存在3个主要的变化周期,分别为28年、14年和7年。研究成果对于了解松花江流域气候变化趋势、合理开发利用水土资源具有重要意义。
Considering that climate change may bring great changes of regional development, it is of great significance to study the air temperature change over Songhua River basin, which is an important commodity grain base in China. Based on the air temperature data of 35 stations in Songhua River Basin from 1956 to 2015, the trend, variation and periodic characteristics of air temperature are studied through linear trend regression, Mann-Kendall test and Morlet wavelet analysis, respectively. The trend analyses show that during the past 60 years, the air temperature over Songhua River Basin has shown a significant upward trend; the increase of annual mean air temperature is 0.322℃ per decade, while the annual mean maximum and minimum temperatures are 0.190℃ and 0.429℃ per decade, respectively. The spatial variation is analyzed through using GIS tools. The regional air temperature variation chart over Songhua River Basin is drawn by interpolating the increments of each station with inverse distance weighted average method(IDW). The chart shows that the increase ranges of air temperature are not uniform in regional distribution. For areas along the Songhua River and its tributaries, the increase range is much bigger than the other regions. Through the Mann-Kendall test, two sets of order series Sk are constructed by sets of forward and reverse time series, and then two curves are obtained by standard normalizing the order series. The two constructed curves UFk and UBk of these three temperature series intersect at one point which is between 1987 and 1992, and the statistical value of the point is between the upper and lower critical limits of the significance test. So, we can conclude that the air temperature has an abrupt change among this period. With the help of the wavelet analysis tool of MATLAB, the wavelet coefficients are obtained. After calculating the real part and modulus of the wavelet coefficients, the real part contour and modulus of the wavelet coefficients are plotted with Kriging interpolation method. Also, the wavelet variance of air temperature variation is calculated at different time scales to draw the wavelet variance chart. The Morlet wavelet analysis results show that there are three main cycles of temperature change, corresponding cycles are 28-year, 14-year and 7-year, respectively.
