The Domain Decomposition Of Thermodynamic Sea Ice System And The Control Of Continuous Condition

In view of problems on solving parabolic equation of sea ice thermodynamic system of snow layer, ice layer and water layer with non-differentiability. Optimal control on the problems is discussed in this paper by non-overlapping domain decomposition method. Arctic sea ice plays a key role in the global climate system. When people observe and accumulate information on the arctic sea ice, they also discuss the numerical simulation of arctic sea ice to reveal the mechanisms of arctic sea ice. A sea ice numerical simulation is useful for forecasting the status and variation of sea ice distribution and its effect on the global climate system with the development of numerical simulation. Now, numerical simulation is already being absolutely necessarily mean for studying arctic sea ice. Therefore the researches of this paper have a important value of theory and practice. Main results of this paper are as following:1. When many researchers study the sea ice with snow covered, they do little on the lubricous subdomains. This paper is based on parabolic equation of sea ice thermodynamic system of snow layer, ice layer and water layer with non-differentiability. Non-overlapping domain decomposition method is discussed. Then the method make every subdomains lubricous and conquer non-differentiability of the problems.2. Through introducing the continuous conditions on internal boundary, a distributed optimal control model is established. In the model state variable is temperature in subdomains., and control variable is the continuous conditions. Its objective functional is to minimize the sum of the difference between the continuous conditions on the interfaces among all neighbor subdomains.3. The method is discussed to conquer uncertainty of the control problems, and the exist of optimal control is shown. Numerical simulation is established by actual data, and the results are well.