论文部分内容阅读
陆面过程模式中输入参数的不确定性会引起模式模拟偏差.为了改善模式的模拟能力,减小参数的不确定性,通常要进行参数优化过程.利用温江站观测的近地层资料,结合粒子群优化算法(Particle Swarm Optimization,PSO),优化了陆面过程模式SHAW(Simultaneous Heat and Water)中难以直接观测的土壤和植被参数.在此基础上,分别利用优化后的参数和默认参数运行SHAW模式,模拟该地区陆面过程特征,并与观测值进行对比,研究优化参数后对陆面过程模拟的影响.结果表明:利用PSO算法优化SHAW模式后,能提高土壤湿度和潜热通量的模拟性能,模拟的土壤湿度和潜热通量与相应的观测值偏差减小.但与此同时,并没有改进净辐射、土壤温度和感热通量的模拟性能.说明PSO算法可以用于陆面模式参数优化,但仅仅通过参数优化并不能同时提高所有变量的模拟性能.“,”The uncertainty of the input parameters in land surface model can introduce simulation deviation.To improve the capability of the models and reduce the parameter uncertainties,usually the parameter optimization process is necessary.In this study,using the surface layer data observed in Wenjiang station and the particle swarm optimization (PSO) algorithm to optimize soil and vegetation parameters that difficult to obtain by observations in the SHAW (Simultaneous Heat and Water) model.On this basis,the SHAW model was run with the optimized and default parameters.Then the simulations were compared with the corresponding observations to investigate the effect of optimization parameters in land surface process simulation.The following conclusions were drawn:Using the optimized parameters calibrated by PSO algorithm can improve the simulation of the soil moisture and latent heat flux.The biases between simulated soil moisture and latent heat flux with the corresponding observations are decreased,but the net radiation,soil temperature and sensible heat flux simulation are not improved.This study suggests that PSO algorithm can be used for land surface model parameter optimization,but the simulation of all variable cannot be simultaneously improved only by optimization process.