An Efficient Method For Simulating Water Exchange Of Large-scale Sea Bay

Water exchange is one of the most important hydrodynamic features of sea bays or coastal waters, which is closely related to the self-purification capacity, environmental capacity, transport of nutrients, and eco-environment of the waters. With the rapid development of technology and social economics, marine exploitation activities are booming. However, these activities will inevitably influence the marine environment. Study of the hydrodynamic characteristics and water exchange as well as their variation under a range of human activities is of great significance for the sustainable development of marine industries and rational utilization of marine resources.Based on the numerical simulation, the water exchange matrix was calculated using the random walk model to study the water exchange in the large-scale bay, and then it was applied to the study of water exchange characteristics among the sub-areas of the bay. Since the hydrodynamic parameters are the preconditions of the random walk model, a parallel computation based stratified 3D tidal model was developed to obtain data with higher accuracy and less time consumption.Some advantages of the random walk particle tracking model could be found through the comparison with other methods used for studying water exchange. Firstly, it remedies the deficiency of traditional Lagrangian particle tracking model in which diffusion effect is neglected. Secondly, the random walk particle tracking model can avoid the numerical oscillation and the resulting incorrect water exchange prediction that might occur when using advection-diffusion model in some circumstance. Moreover, the Lagrangian character of this method makes it more convenient to calculate the time scale of water exchange based on Lagrangian viewpoint, such as age.Being tested and verified with a set of classical examples, the random walk particle tracking model was proved to be valid for simulating advection-diffusion problem. In order to improve the computational efficiency, this model was implemented with parallel computation as well.It was found that if the particles are sufficient enough, the results from random walk model agree well with those of advection-diffusion model when studying the water exchange of a large-scale bay. It was also proved theoretically that the water exchange matrix could be used for medium-long-term quick prediction of the water exchange in a large-scale bay. The influence of particle number, matrix timing and number of subareas in the model on prediction accuracy was analyzed and tested with numerical experiments.The random walk model was applied to study the water exchange in Bohai Sea and Tianjin coastal waters, and the influence of wind on the water exchange in and among the subareas of Bohai Sea was discussed respectively. The concept of reigional average residence time was suggested to study the impact of the land reclamation in Tianjin coastal zone on the water exchange, based on which the variation of water exchange among the subareas was analyzed in comparison. The combination of random walk model and age concept is applied to analyze the impact of the land reclamation on the polltute transportation path of the two discharge points in Tianjin coastal areas.