论文部分内容阅读
鉴于预报气候变化条件下黑龙江地区的冰情演变趋势对黑龙江地区的防凌减灾工作具有重要意义,通过趋势图法和M-K趋势检验法分析了1957~2012年近60年黑龙江地区漠河站和哈尔滨站的气象要素和冰情要素的变化趋势,并利用多元线性回归法和人工神经网络法对待测站点的冰情资料进行了预报。结果表明,近年来漠河站和哈尔滨站均有封河日期延后、开河日期提前的趋势,且在预测漠河站开河封河日期、哈尔滨站的封河日期时,多元回归法优于人工神经网络法,预测哈尔滨站的开河日期时,两者效果相差不大。
In view of the fact that the trend of icing development in Heilongjiang area under the condition of climatic change is of great significance to the prevention and mitigation work in Heilongjiang region, the trend map method and the MK trend test method were used to analyze the past 60 years of the Hehejiang region Mohe station and Harbin station from 1957 to 2012 Meteorological elements and ice features, and forecast the ice conditions of the stations by multivariate linear regression and artificial neural network. The results show that in recent years, both Mohe Station and Harbin Station have delayed the closure of the river and opened the river ahead of schedule. The multiple regression method is superior to the artificial neural network method in predicting the date of the opening of the Mohe Station and the closure of the He’he River in Harbin. Harbin Railway Station opening date, the effect is not much difference between the two.