集合变分法在大气边界层湍流粘性系数反演中的应用

利用集合变分法研究了大气边界层EK模式(考虑了水平气压梯度力、科氏力和湍流粘性力3力平衡的大气边界层模式)中湍流粘性系数的反演问题.首先利用集合变分法推导了目标函数关于反演参数的梯度表达式,然后给出了反演计算方案,最后再就反演湍流粘性系数k进行了一系列理想数值试验.数值试验结果表明:利用集合变分法对EK模式中的湍流粘性系数的反演比较成功.通过适当调整集合数,初始扰动均方差及初始值的大小,可以进一步提高集合变分法对湍流粘性系数k的反演精度.将集合变分法应用于大气边界层参数的反演是一种值得研究的计算方案.

Retrieval of ABL Turbulence Viscosity Coefficient Using Ensemble-Based Variational Method

The ensemble based variational method (EnVar) is used to study a retrievable problem about turbulence viscosity coefficient in the atmospheric boundary layer EK model.By using the EnVar, the gradient expression of the cost function in retrieving turbulence viscosity coefficient is deduced, and the retrievable calculation solutions are given. A series of ideal numerical experiments are conducted on the retrieving of turbulence viscosity coefficient. The results of numerical experimentations show that it is successful to use the EnVar to retrieve the turbulence viscosity coefficient in the EK model. In addition, through adjusting ensemble number, initial disturbance mean variance, and initial value appropriately, the inversion precision can be further enhanced. It is feasible to use the EnVar to retrieve the parameter of atmospheric boundary layer.