In the previous paper, the authors introduced a method of seasonal adjustment into the Auto-Regression (AR) model prepared for the space-time analyses of seawater qualities measured in broad sea area. They showed that the AR-model with the seasonal adjustment would give estimates for a long term transition of seawater qualities. In the present paper, they discuss about characteristics of long-term trend, seasonal and auto-regressive component using several kinds of data obtained in Osaka bay for a long period of time. They show that the seasonal adjustment is effective, to keep the consistency of measured data at difficult time and position obtained by different organizations. They also show that the AR-components follow the laws of probability with a certain pdf. They finally conclude that the modified AR-model proposed as a suitable tool for the space-time assimilation of seawater quality data.