Kim, J. and Park, J., 2013. A statistical model for computing causal relationships to assess changes in a marine environment
In order to manage sustainable coastal development, it is essential to identify the causal relationships among observation parameters that reveal the changes in the marine environment in a quantitative manner. On the Saemangeum coast, a land reclamation project by constructing sea dike has been underway since 1991. To monitor and assess the changes in the marine environment due to the project, the states of the ocean's physics and circulation, water quality, marine geology, and the marine ecosystem have been measured through the integrated ocean observation networks. In this paper, the observed data are statistically investigated with regard to observed and latent variables, in order to identify their causal relationships and compute the degrees of their influence. We performed a multivariate statistical analysis using a structural equation model based on oceanographic theory using the monthly mean observed data. As a result, 15 principal components were extracted and the statistical model was obtained by estimating the standardized regression coefficients of the observed variables and the latent variables, as well as those between the latent variables. The statistical model was applied to both the inner reclaimed area and the outer sea area individually to verify the model quantitatively how well it explains the changes of marine environment due to the blocking of seawater exchange in the inner reclaimed area comparable to the outer sea area separated by the sea dike. Our results show that the proposed statistical approach is quite suitable for understanding phenomena in ocean science that occur causally by identifying theoretically-known causal relationships quantitatively. By predicting the various impacts on environmental changes in advance through the quantitative estimation of a statistic model, we can prepare appropriate countermeasures or alternatives for preserving and protecting the environment, whose vulnerability is an inevitable result of such development. In this respect, our study can be effectively utilized as a tool for coastal management, policy-making, and planning.