黄河流域近50年降水变化趋势分析

Long-term trend of precipitation in the Yellow River basin during the past 50 years

本文简要分析了黄河流域降水空间分布规律，用非参数检验方法（Mann-Kendall法）对流域内77个气象站近50年的降水资料进行了年月降水序列趋势检验，并用线性回归方法与其进行了比较。结果表明，对于年降水序列，有65个气象台站呈现下降趋势，典型月（包括1月、4月、7月、10月）中4月、7月和10月对年降水下降趋势贡献较大，但其趋势空间分布情况存在差异。对于所有月降水序列，全年4～5月以及7～12月降水呈现下降趋势。而其趋势空间分布，黄河上游北纬35。以南地区除7～9月外其余月份降水呈现增长趋势。渭河上游及呼和浩特地区降水趋势随月份时有变化，流域内其他地区降水则呈减少趋势。对比两种方法，Mann—Kendall方法在冬季（12月至次年2月）的检验结果中无变化趋势气象站数明显多于线性回归方法，且后者估计出的趋势变化幅度略大于Mann-Kendall方法所估计出的结果。

Some of previous studies showed that the drying-up of the lower Yellow River was resulted from decreasing precipitation and excessive industrial and agricultural consumption of water from the middle and lower Yellow River. The analysis for the long-term trend of precipitation is helpful for the rational use of water resources in the Yellow River basin. The Mann-Kendall statistic test is a rank-based, non-parametric approach and has been widely used to estimate the significance of long-term trends in the hydro-meteorological time series such as water quality, streamflow, temperature, and precipitation. The long-term trend of precipitation in the Yellow River basin was analyzed in this paper and the Mann-Kendall test was applied to annual and monthly precipitation time series. The linear regression was compared with the results obtained by using Mann-Kendall test. The results show that 65 of 77 stations exhibited downward trend for annual precipitation. April, July and October contributed the most of the decreasing trend of annual precipitation, and the spatial distribution of long-term trend for typical months is different. In all months except January, February, March and June, the values of trend are negative, exhibiting the decreasing tendency. In all seasons except summer, there is a similar increasing trend in the upstream of the Yellow River, south of the northern latitude 35 degrees. Except the regions in the upstream of the Weihe River and Hohhot the precipitation showed a decreasing trend in the Yellow River basin, especially in the downstream of the Yellow River. Comparing the results estimated by two methods, the Mann-Kendall test and linear regression test, the stations in winter （from December to next February）, without trend tested by Mann-Kendall method, are much more than those tested by linear regression method, and the number of stations with trends tested by both methods is similar in other seasons. The magnitude of trend estimated by both methods is also similar in all months except July and December, and the absolute values of trend estimated by linear regression method are greater than those estimated by Mann-Kendall method.