ABSTRACT
Seawater salinity is an essential part of the characteristics of seawater, and its study can be invoked as a basis for mariculture evaluation, marine environmental research, marine engineering maintenance and analysis of sea ice in winter. In order to accurately predict seawater salinity, this paper summarized and analyzed the winter seawater salinity data in the North Temperature Zone from 1990 to 2019, and established a classification standard suitable for the winter seawater salinity in the North Temperature Zone by means of mean-standard deviation classification method. Based on the above classification standard, the weighted Markov chain was used to predict the winter seawater salinity in 2019 and 2020, and winter seawater salinity in 2019 in the North Temperature Zone was predicted by combining the Grade Characteristic Value (GCV) from the fuzzy-set theory. The results are in agreement with the actual situation. Finally, the stationary distribution is used as a measure of the return periods of each seawater salinity state for further optimizing the prediction system of seawater salinity, which provides an indispensable reference for the prediction of seawater salinity.
