中文摘要：

基于1996—2007年逐月时间序列数据,采用季节ARIMA模型对连云港近海表层水温时间序列进行模拟,并依据残差不相关和简洁性原则确定模型的结构,建立最优预测模型ARIMA（1,0,1）（0,1,0）12。运用该模型对2008年逐月表层水温进行预测,预测值与实际值的变动趋势基本一致,且平均相对误差仅为3.5%。在此基础上对2009年连云港近海逐月表层水温进行预报,预报结果符合该海域表层水温的逐月变化趋势,表明模型用于近海表层水温预报是可行的。

英文摘要：

A seasonal ARIMA model was used to simulate the time series of Lianyungang coastal SST based on the monthly SST from 1996 to 2007, and the optimal model ARIMA（1,0,1）（0,1,0）12 was finally established with its structure determined according to the criteria of residual un-correlation and concision. And then the constructed optimal model was applied to predict monthly SST in 2008. The results showed that the predicted values, with the average relative errors of about 3.5 %, almost had the same change trends as the actual values. With the model of ARIMA（1,0,1） （0,1,0） 12, the monthly SST in 2009 were forecasted based on monthly SST from 1996 to 2008, and the forecasting results were consistent with the monthly change tendency, which indicated that the selected model was suitable for forecasting coastal SST in Lianyungang.