典型非球形冰晶粒子的凝华增长数值模拟试验

Numerical Simulations of Ice Crystal Growth in the Water Vapor Deposition Process for Typical Ice Habits

在凝华增长过程中,冰晶的形状随着温度和湿度的改变而改变,准确模拟冰晶粒子的演变对于提高云模式的模拟能力起着非常重要的作用。在现有的云模式中,冰晶形状通常假设为球形,而在实际大气中,冰晶形状十分复杂。本研究中,我们根据冰晶凝华增长理论模型建立了一个单个冰晶粒子增长模型,模拟了温度分别为-1°C^-30°C时,单个典型非球形冰晶粒子的凝华增长过程。与风洞观测数据相对比,该模型能够抓住单个冰晶粒子的轴长,质量以及纵横比随温度和湿度的变化过程。我们进一步将该理论增长模型应用到群粒子的凝华增长过程的模拟。我们釆用欧拉二维正定平流输送法(MPDATA)模拟了典型非球形冰晶群粒子的凝华增长,并对比分析了在不同纵横比分辨率下的模拟效果以及温度变化对冰晶形状的影响,结果表明运用该数值方法可以合理地模拟出群粒子在凝华增长过程中纵横比的演变。与目前采用的拉格朗日—欧拉混合平流算法比较,该算法能够耦合到欧拉动力框架下的分档云模式中去,这对我们研究冰晶粒子形状对云微物理过程和动力过程的影响,以及它们对冰粒子凝华增长的反馈作用具有非常重要的科学意义。

A spherical shape is normally assumed for ice particles in most cloud models. However, the shapes of ice crystals in the real atmosphere are much more complicated. Ice crystal habits will change with the variation in ambient temperature and water vapor saturation. Such habits have been proven as critical parameters impacting cloud simulations. In this study, based on the theoretical model of the deposition growth of an ice crystal, we firstly simulated the growth of a single ice crystal by water vapor deposition under temperatures from-1°C to-30°C. The model can capture the evolution of axis length（a for prism face; c for basal face）, mass and aspect ratio in comparison with the data from wind tunnel observations. We further simulated the water vapor deposition growth of non-spherical ice crystals with the twodimensional positive definite advection transport algorithm（MPDATA）. Furthermore, in order to test the feasibility of applying such a treatment into the Eulerian dynamical framework, the mass growth of ice crystals under different bin resolutions for the aspect ratio was simulated. The results showed that the model using the MPDATA method can capture the evolution of ice crystals for both their mass and their aspect ratio. Even though some Lagrangian models with bin microphysics involve microphysical processes for non-spherical ice crystals, their schemes with the hybrid Lagrangian– Eulerian advection transport method cannot be applied into cloud models under the Eulerian dynamical framework, which can simulate more complicated microphysical processes and dynamical processes involved in ice particles.