Physical Parameter Identification And High Resolution Modelling In Ice And Snow Thermodynamic Processes

Global warming and its resulting events, special for sea level rising, more extreme weather would make some directly or indirectly impacts on human’s production and life. Reasonable forecast of climate change is crucial to put forward corresponding policy. Sea ice and fresh ice in polar or sub-polar regions are primary factors in global climate system, and serve as proxy climate records as their sensitivity to climate change. The optimization of parameterization and the arithmetic are more important than the model strategy to develop the ice thermodynamic model. Based on the data derived from the reservoir-ice measurements in Hongqi-pao reservoir in Heilongjiang province in the winter of2008-2009, lake-ice thermodynamic processes in two Finnish lakes in the winter of2007-2008and the winter of2008-2009, and the field campaigns of landfast sea-ice thermodynamic observations off Zhongshan Station in Prydz Bay, East Antarctica from November2005to December2006, the growth and decay of ice cover have been studied. The sonw or ice surface nonlinear discrete time delay system and the ice bottom energy balance system have been established. The theory of the nonlinear differential equation, the time delay differential equation and bounded variation are use to investigate the existence and uniqueness of the solution of the systems. The optimal identification models have been put forward, and the existence of the optimal solution of the optimal identification models have been proved. The optimal time delay between the snow or ice surface temperature and the air temperature, modelled freezing date and oceanic heat flux are calculated by optimization algorithm. The physical parameters in snow and ice thermodynamic processes which have more influence on the snow and ice growth or decay, have been indicated.The main contributions are as follows:1The inversion of the surface temperature for ice thickness by thermal remote sensing has been proved by the model simulations. The deviations between the observed and calculated snow or ice surface temperature by assumption and iteration methods have been analyzed. The nonlinear discrete time delay system for calculating the time delay between the snow or ice surface temperature and the air temperature has been established. The existence and uniqueness of the solution of the system has been proved by the theory of the time delay differential equation. Taken the time dalay between the snow or ice surface temperature and the air temperature as the identified parameter, the temperature deviation of calculated snow or ice surface temperature and observations is defined as the performance criterion, and then the optimal identification model has been put forward. The existence of the optimal solution of the optimal identification model has been proved, and the optimality conditions of the optimal identification model are provided. Based on the field measurements of Finnish lake ice, the numerical simulations show the optimal time delay between the snow or ice surface temperature and the air temperature was one hour.2The upper and lower envelope of the ice thickness in Hongqi-pao reservoir which can be cite as the reference values for the water project design, have been calculated by some simple analysis models. To notice the difference of the physical parameterization of sea ice and lake ice, snow and ice model has been applied in the numerical simulation of Finnish lake ice. The simulation results show that the sensibility of the snow or ice surface temperature to the variation of the air temperature. With the air temperature increasing1℃, the lake ice season will be shorten by13days, and the annual maximal ice thickness will be6cm(which is about17%of the total ice thickness) more than reference simulation. A high resolution thermodynamic snow and ice model and lake thermodynamic model are applied to investigate the snow and ice thickness in Finnish lakes. The model simulations show that the necessity of the high resolution method in the lake ice thickness calculation and the lake thermodynamic model can give more accurate simulation during the early ice growth period. The modelled freezing date has been definited. The establish of the optimal identification model, which contain the freezing date as identified parameter, the observed air temperature and modelled ice thickness as the constraints, to make the snow and ice model can calculate the freezing date. The freezing date identified by the the optimal identification model more close to the observation.3According to the energy balance equation at the ice bottom, the ice bottom energy balance system has been given, and the existence and uniqueness of the solution of the system lias been proved. The oceanic heat flux is selected as identified parameter and the the ice thickness deviation as the performance criterion, so that the optimal identification model has been presented. The existence of the optimal solution and the optimality conditions of the optimal identification model have been considered by using the theory of bounded variation. This method to estimate the oceanic heat flux is only controlled by observed ice thickness, which can overcome the calculated bias caused by the technique error of observed temperature and the empirical parameters in other methods. Based on the the field campaigns of landfast sea-ice thermodynamic observation off Zhongshan Station in Prydz Bay, East Antarctica in March2006to November2006, the time series of the oceanic heat flux and the oceanic heat flux fitting function of time has been derived. The effectiveness of the ice bottom energy balance system has been provided by the sensitivity study on the measurements in the same observation area from May2005to November2005.