摘要
基于生产实践中对高精度中长期径流预报的要求,对我国海河流域两大支流之一的滹沱河上小觉水文站(岗南水库入库径流控制站)径流量进行预测研究。采用Mann-kendall检验法对数据进行分析,得到流域径流量的突变年份为1984年。基于此采用1992—2001年的数据(突变年份后的时间序列)训练得到预测模型,采用2002—2004年径流量日值数据就模型对新数据序列的适应性进行验证,建立小觉水文站日径流预测模型。结果表明模型应用在水文时间序列的预测中是合理、可行的,并具有较高的精度。
Based on high precision requirement of medium to long-term runoff forecasts in practice, runoff of Xiaojue hydrological station on the Hutuo River（controlling the runoff of the Gangnan reservoir） was predicted. The sudden change of runoff series appeared in 1984 by using Mann-kendall test, based on which forecasting model was built by using the data of daily runoff series between 1992 to 2001, and test prediction model was based on daily data between 2002 to 2004, then Xiaojue drological station daily runoff forecasting model was established. The result of calculation showed that the model was feasible, reasonable and highly precise.
