塔里木盆地南缘于田绿洲近50a来气候变化特征分析

基于于田气象站48 a(1960—2007年)逐月平均气温和降水量数据,借助小波分析、人工神经网络等方法研究了塔里木盆地南缘于田绿洲近50 a来在全球气候变暖的背景下气候变化的特征和总体趋势。结果表明,二维小波系数图可以清楚地显示出气候变化在不同时间尺度上的演变特征,并能对未来的气候演变趋势做出定性估计。小波方差图显示于田绿洲逐月平均气温存在4 a为主的短周期和以11~13 a为主的长周期;逐月降水存在以7 a为主的短周期和以准12 a、27 a左右为主的长周期。RBFN(Radial Basis Function Network)模型预测的结果显示,该区2008—2012年降水量的变化处于相对高值期,气温变化处于相对低值期。近50 a于田绿洲气候总体趋于暖湿。

Change of Average Air Temperature and Precipitation at Yutian Oasis in the Southern Part of Tarim Basin During 1960-2007

The characteristics and tendency of climate change at Yutian oasis in the southwestern part of Tarim Basin of China in the recent 48 years (1960—2007) are analyzed using wavelet analysis, artificial neural network in this paper. The results show that the two-dimensional wavelet coefficient chart can manifest clearly the features of climate variation on different time scales and it could also be used to estimate qualitatively the future tendency of climate. The curves of wavelet variance show that prominent periods of annual mean air temperature are 4 years, 11~13 years. And annual precipitation has prominent periods of 7 years, 12 years and 27 years. Finally, climate tendency is predicted with the RBFN (Radial Basis Function Network) model. Results show that there is a increase in precipitation and a decrease in air temperature during 2008—2012. The results also indicate the climate tend to be warm and humid as a whole in Yutian oasis.