Effects Of Melting And Fall Speed Parameterizations For Ice-phase Particles On Physical Processes In Idealized Convective Clouds:Numerical Studies

To a great extent,clouds govern the radiation balance and energy fluxes reach Earth and have a strong impact on the hydrological cycle and freshwater distribution over the globe.The formation and the development of clouds are dependent upon the combination effects of dynamics,thermodynamics and microphysics in the atmosphere.Thus,if we parameterized microphysics in a different way in cloud models,we could change the distributions of hydrometeors within simulated clouds.As a consequence,the latent heat release/absorption effects during phase transition and the drag force of hydrometeors would be changed,which would impact the thermodynamical and dynamical structures of clouds and their environment greatly.It is urgently required to characterize the physical processes within clouds appropriately in the cloud models and study the interaction between microphysics,thermodynamics and dynamics in the clouds,which would help us to have a deeper understanding of clouds and evaluate their effects on weather and climate sensitivity using models.How to treat the microphysical processes of ice-phase particles in a proper way is of great significance for the cloud models given the wide variety of ice particle shapes and types and a lack of observation and physical understanding compared to liquid-phase microphysics.In nature,ice particles especially graupel has a large range of densities and fall speeds and their fall speeds are highly related with their densities.However,the traditional bulk and bin microphysics schemes represent ice particles by separating them into predefined categories,such as cloud ice,snow crystal,graupel and/or hail.They treat those particles with a priori specification of key parameters by using fixed bulk densities and corresponding fall speed parameters and their fall speeds are only related to their diameters and independent of their densities.However,fall speeds should depend explicitly on the densities.Recently,some developers are trying to remove the predefined categories by predicting the bulk volume mixing ratio and the bulk density so that important properties of solid hydrometeors could evolve realistically and smoothly in time and space.Despite the conceptual improvement over traditional fixed-category schemes for ice-phase representation,there are still many limitations in the newly developed schemes,such as the parameterizations of the fall speed and the melting process.This Ph.D.thesis aims at filling the gaps and major focus in this thesis is put on the physical processes in the simulated clouds influenced by the choices of different parameterizations of fall speeds and melting process by using the NSSL(National Severe Storms Laboratory)scheme implemented into the WRF(Weather Research and Forecasting)model.Two thermodynamical profiles are used to represent the shallow and the deep convective clouds and force WRF in an idealized mode.Firstly,we modified the melting process of the NSSL scheme by predicting the liquid fraction of ice-phase particles and allowing the representation of the mixed-phase particles,which is one of the highlights of this thesis,and compared it with the traditional scheme of“instantaneous melting".Simulation shows that the modified scheme is likely to produce higher precipitation in the convective region due to lower evaporation rates compared with the traditional scheme.The combination effects of stronger melting and sublimation cooling simulated by the modified scheme would generate stronger downdraft and cold pools.Secondly,we fixed the densites of graupel with low and high values respectively and gave them corresponding fall speed parameters as the traditional schemes did.The low-density assumption would produce much more extensive precipitation due to the longer residence time in the atmospehere.The high-density assumption would generate stronger rainfall rates in the convective region because of the higher riming and melting rates.Larger latent heat release and absorption accompanied by riming and melting in the high-density cases would result in stronger updraft and downdraft.Finally,we tested the sensitivity of the idealized clouds to the fall speed parameterizations by using the density-predicted NSSL scheme.One parameterization would generate slower fall speed.Such slower fall speed would allow the updraft to bring more water vapor to condensate at higher altitude and positive feedback between latent heating and updraft possibly helps to support the growth of graupel mass.Our work and reults suggest that the parameterizations of ice-phase processes are critical to better simulating and latent effects during phase-transition have a strong feedback on the dynamical structures of the simulated clouds.Therefore,special attention should be paid to further observations and theoretical calculations of the dynamics and microphyiscs so that models can be effectively evaluated and improved.