目前国内外进行海水温度预报主要是应用数理统计方法，平稳随机序列分析法就是其中之一。本文根据海水温度自身前后期之间的关系，采用平稳随机序列分析法，用小麦岛海洋站３６ａ（１９６０～１９９５年）逐旬表层水温观测资料，建立多元线性回归方程，从而算出未来时刻的水温预报值。 At present, the forecast of seawater temperature at home and abroad mainly uses the mathematical statistics method, and the stationary random sequence analysis method is one of them. Based on the relationship between the seawater temperature and its own period, a series of linear regression equations were established by stationary stochastic sequence analysis method using the surface water temperature observed during the period from 1960 to 1995 at Wheat Island Marine Station to calculate the water temperature in the future Forecast value.