Hydrological Characteristics Analysis In The Upstream Of Heihe River

On account of global warming and the impact of human activities,conditions on water resources of Heihe River basin, which is sensitive to climate change,is of great change.Upstream Yingluoxia,the upper reaches is the main runoff area of Heihe River.As precipitation and temperature are closely related to the runoff, it is necessary to study on the relationship among precipitation, temperature and runoff, in order to provide theoretical basis for development water resources.At first, this study use the curve deduction mass to analysis precipitation、 temperature and runoff data examined by the reliability, consistency, and representative。 Then, the point mutation and changing trends of precipitation、 temperature and runoff data were analyzed by M-K test and the optimal parting of ordered samples. The relationship among precipitation, temperature, and runoff series were studied. The reason which caused the special year was checked by comparing runoff series with precipitation and temperature. Finally, the suitable mathematical model was used to simulate the runoff series. The runoff data was collected from Zhamashike、Qilian and Yingluoxia stations, and the precipitation and temperature datasets were collected from the Yeniugou and Qilian stations.The precipitation measured by Yeniugou and Qilian stations both showed an increasing trend, at rates of0.12mm/a and0.91mm/a. The result of optimal partition of ordered sample showed that precipitation mutation in1971,1975,1998and2003was not obvious. The M-K test was consistent with the optimal partition of ordered sample. Both the two methods found the significant precipitation mutation in2007. The optimal partition of ordered sample method showed that there was no obvious mutation of Qilian except the1974,1988,2003,2008, and the result of M-K test showed the change trend of Qilian was relatively complex, no obvious mutation was detected.The temperatures measured by Yeniugou and Qilian presented obvious rising trend, at a rate of0.03℃/a. The optimal partition of ordered sample and M-K test both displayed a significant mutation from low to high since1997, and the increasing trend would continue.The runoff measured by Zhamashike showed an increasing trend, at a rate of2.2Million Cubic Meters per year. The result of optimal partition of ordered sample showed that runoff mutation in1981,1989and1998was not obvious. Both the optimal partition of ordered sample and M-K test found the significant precipitation mutation in2007, which was similar to precipitation. The runoff measured by Qilian showed an increasing trend, at a rate of3.5Million Cubic Meters per year. The result of optimal partition of ordered sample showed that runoff mutation in1980,1982and1998was not obvious, but the runoff mutation in2003was obvious. The M-K test found a complex change since the1981. After2003, the increasing trend would continue.The relationship between runoff and precipitation showed that precipitation is the mainly supply of runoff during June to September in the upstream of Heihe river. On the contrary, precipitation is not important supply both from January to July and October to December.The relationship between runoff and temperature showed that the as low temperature in January to March or October to December, there was not enough melting water supply the runoff. Although melting water in June to August was enough, these months were rainy season, so, the relationship between runoff and temperature is obvious in April、May and September。The reason in Qilian station was the same.There is a good runoff relationship between Zhamashike、Qilian and Yingluoxia. The maximum precipitation happened in1989all of the three area.1973was the minmun precipitation both happened in Zhamashike and yingluoxia, but Qilian was in1979. It is supposed that1988had more precipitation than1997is the reason why runoff inl989is more than in1998.Results of correlation analysis on temperature precipitation and runoff of Buffalo and Qilian shows good relationship. Results of correlation analysis on averaged precipitation and temperature of Buffalo and Qilian to runoff of Yingluoxia shows good relationship. With precipitation and temperature as input variables, runoff as output variable, a multivariate CAR model is set up, whose simulation result is realistic.