台兰河年平均流量非平稳序列加法模型的建立与预报

摘要：将台兰河作为塔里木河典型源流径流的案列，通过SPSS应用技术，探索适合干旱区河流水文预报方法。首先将台兰河台兰水文站1957-2008年年均流量Q（t）序列分解为趋势、周期、平稳函数项和噪声项，建立了非平稳时间序列加法模型，然后对建模年限内的Q（t）序列作了模拟检验，最后对建模年限外的2009、2010、2011年Q（t）值作了预报，预报结果符合水文情报预报规范要求，可供生产实践参考。

Establishment and Prediction of the Additive Model of nonstationary mean annual discharge series of the Tailan River

This paper introduces the part project achievements of performance prediction of typical origin runoff of Tarim river under climate change. The study area is Tailanhe as the case,by application technology of SPSS. Exploring hydrological forecasting method suitable for the arid area's river. First. the mean annual discharge sequences Q(t)from 1957 to 2008 in hydrological station of Tailanhe was decomposed into trend cycle stationary function item and noise item. Non-stationary time series Additive Model is established. Then. simulation testing Q(t) in modeling. Finally. forecasting Sequences Q(t) except for the year of 2009 2010 and 2011. The forecast result that it is accord with hydrological information and hydrological forecasying. It can be referenced for production.