气候变化对秦岭北坡径流过程的影响机制研究——以灞河流域为例

Influencing mechanisms of climate change on runoff process in the north slope of Qinling Mountains:A case of the Bahe River Basin

综合运用距平分析、连续小波变换(CWT)、交叉小波变换(XWT)、小波相干(WTC)和时滞相关性分析等方法对秦岭北坡灞河流域径流与六大气象因子进行了多尺度分析。结果表明:水文气象因子在不同时频域中存在1~4个显著性周期,且都能通过95%置信水平检验;太阳黑子数8~11.4 a的显著性周期对同时间尺度年均SOI和年均MEI的连续小波能量谱具有明显的影响;交叉小波变换和小波相干分析相结合的方法在识别水文和气象要素的共振周期、相位角变化、丰枯阶段转换、时滞性、相关性、突变与显著性检验等方面具有独特的优势;同期降水是灞河流域径流形成和变化的控制性因素,年均太阳黑子数的影响作用微乎其微,月均流量滞后月均AOI 0个月或8~9个月时的相关系数绝对值达到最大,滞后月均SOI 2个月或7~8个月达到最大,滞后月均MEI 6~8个月达到最大。

Runoff process is not only directly affected by local rainfall and temperature, but also closely related to southern-northern hemispheric teleconnections, such as Arctic Oscillation, Antarctic Oscillation, Southern Oscillation, E1 Nino, etc. Therefore, there is a great significance to investigate the influencing mechanisms of climate change on runoff process. Qinling Mountains is the dividing line between north and south China, which is also a hot spot of climate change around the world. Therefore, taking the Bahe River Basin in the north of Qinling Mountains as study area, this paper aims to reveal influencing mechanisms of climate change on runoff process, especially changes in climate factors how to lingeringly affect runoff change, and further provide more theoretical foundation of teleconnections for hydrograph forecast. In order to achieve this, a variety of integrated methods in- cluding anomaly analysis, continuous wavelet transform, cross wavelet transform, wavelet coherence and lag cor- relation were applied to complete multi-time scale analysis between runoff and meteorological factors including temperature, precipitation, A01, SO1, sunspot number and ME/. Finally, we conducted the test of delayed correlation analysis. The results show that linear trend estimation demonstrates hydrometeorological factors presented a different trend of increasing or decreasing from 1959 to 2014 in the Bahe River Basin. Over the past 56 years, cli- mate change presented a trend of warm and dry, which had a profound influence on runoff process in this basin. There were 1 to 4 significant （P〈0.05） periods. It is worth noting that the 8-11.4 a significant period of sunspots number distinctly affected the continuous wavelet energy spectrum of annual average S01 and ME1 at the same time scale, because their energy spectrums on the frequency band all showed a characteristic of zonal distribution. The method of combining cross wavelet transform and wavelet coherence analysis has unique advantages in identifying resonant period, phase angle relationship, wet or dry period, lag, relevance, mutability and significance testing between hydrological and meteorological factors. Lag correlation analysis shows that the influence of different meteorological factors on the runoff had some differences in terms of correlation and time lag mechanisms. Precipitation in the same period was the key controlling factor of runoff formation and change, however, annual average sunspot number affected the runoff weakly. When average monthly flow lags monthly AOI for 0 months or 8-9 months, the absolute value of correlation coefficient will reach the maximum; when average monthly flow lags monthly SO1 for 2 months or 7-8 months, or lags monthly ME/for 6-8 months, the maximum absolute value of correlation coefficient will also appear.