Analysis Of Runoff Characteristics Of Dry Season And Forecasting Of Low Flows In Kuitun River Basin On The North Slope Of Xinjiang Tianshan Mountain

To study low-flow runoff has practical significance for exploitation, utilization and assessment of water resources and for hydrologic analysis of environmental engineering and water conservancy. Kuitun River Basin is one of the largest agricultural irrigation area in Xinjiang and the largest branch in Ebinur lake water system, and researches on its low-flow problems is of great significance.Based on runoff data of seven hydrologic stations in Kuitun River Basin from 1959 to 2006 and according to previous relative researches,this paper analyzes influencing factors and characteristics of low-flow runoff of the three main rivers (Kuitun River, Sikeshu river, Guertu river ) and predicts primarily their minimum monthly flows by use of multiple linear regression analysis. The results showed that:(1) The dry season of Kuitun River Basin is up to 7 months, and its minimum monthly-average flow throughout the year is in April, later than that of other rivers in northern slope of Tianshan Mountains.(2) Drainage area, watershed geology and local forest cover are the mainly impact factors to low flow runoff of Kuitun. Flows during every low-flow periods increase accordingly with drainage area increasing. Geological structures, buried conditions and permeability of different lithologies influence low flows greatly. Local forest has important effect on republishing of groundwater to the rivers in dry season.(3) The optimal linetype of low-flow frequency distribution of the seven stations at different dry periods is different, but the minimum monthly and three-monthly average flow of mountain-pass stations in the three rivers fit Pearson-IIIcurve optimally. The Cs / Cv value is not adjusted by use of taking integral multiples of Cv as Cs customarily such as 2 or 4 times but obtained from minimum difference between theoretical and empirical frequency through least square method with varied multiples from 2 to 9.(4) The monthly-minimum flow of mountain-pass stations in the three rivers is predicted by regression method, and regression equations are established with annual average flow of previous year, flood flow, monthly-maximum flow and 10-month flow as variable factors. The predicted results have certain error, and the error of several yearly-maximum flow is large, while that of annual average flow is small less than 2%. The predicted and measured high and low flows of multi-year are basically in accordance.Based on the conclusions, some recommendations are proposed for actual production activities and further relative studies.