Numerical Simulation Based On The Field Observation Data Of Sea Ice Temperature Field Of Antarctic

Polar sea ice is an important part of the global climate system. It is known as theindicator of the global climate change. It is an important factor that cannot be ignored ina wide range of environmental discussion and future climate prediction. The optimalcontrol model, which using the least square method, is established to optimize andidentify the relationship of Antarctic sea ice surface temperature and the localatmospheric temperature on the basis of the previous experience. So we can calculatethe ice surface temperature by atmospheric temperature according to the relationship.Then we make the ice surface temperature as the boundary conditions of the Antarcticsea ice temperature field numerical simulation in order to avoid the limitations broughtby making the application of ice temperature field observation data as the boundarycondition. The main findings of the present study are summarized as follows:1、We establish the optimal control model, which using the least square method, tooptimize and identify the relationship of Antarctic sea ice surface temperature and thelocal atmospheric temperature based on the observation data of the21st and22ndAntarctic scientific expedition(2006). The actual measured temperatures data arecollected by the Antarctic sea ice observation point of Zhongshan Railway Station inNella Bay. Then we use the software MATLAB to program and solute the model2、We establish the system equations and the initial boundary value conditions ofthe Antarctic sea ice temperature field numerical simulation. They are on the basis ofthe one dimensional thermodynamic sea ice model and considering the influence of seaice internal heat source on the Antarctic sea ice temperature change. UsingCrank-Nicolson difference scheme and the semi implicit difference format (Saul ’EValgorithm) to solve the initial boundary value problem of the one-dimensional heatconduction equation, respectively. Then we make the numerical simulation of thecorresponding period changes in the Antarctic sea ice temperature field. In this thesis,we make an improvement on the numerical simulation of natural ice temperature fieldin reference [43]. The new method can make the numerical simulation on the icethickness of the Antarctic sea ice temperature field and the ice layer growth temperatureof each point at the same time. 3、We make the numerical simulation on the changes in the Antarctic sea icetemperature field by using C++to solve the initial boundary value problem of the heatconduction equations after scatter. Analyzing comprehensively the result of thenumerical simulation, we consider that the method using Crank-Nicolson differencescheme is better in the accuracy and applicability than the method using the semiimplicit difference format.