A Lagrangian stochastic model for the trajectories of particle pairs and its application to the prediction of concentration variance within plant canopies

A simple Lagrangian stochastic model for the trajectories of particle pairs in high Reynolds-number turbulent flows is presented. In this model, the velocities of particle pairs are initially correlated but subsequently each particle moves independently. The independent single-particle trajectories are simulated using Thomson's model [J. Fluid Mech. 180, 529–556, 1987]. This two-particle model exactly satisfies the well-mixed condition for Gaussian turbulence when length scales, characterizing the two-point Eulerian velocity correlation function, vanish. Temperature variances, due to heat released as a passive scalar from an elevated plane source, within a model plant canopy (Coppin et al. Boundary Layer Meteorol. 35, 167–191, 1986) are shown to be well predicted by the model. It is suggested that for strongly inhomogeneous flows, the two-point Eulerian velocity function is of secondary importance in determining the simulated trajectories of particle pairs compared to the importance of ensuring satisfaction of the two-to-one constraint (Borgas and Sawford. J. Fluid Mech. 279, 69–99, 1994); i.e ensuring that one-particle statistics obtained from the two-particle model are the same as those obtained from the corresponding one-particle model. Limitations of this modelling approach are discussed.     

一种“准拉格朗日法”和“欧拉法”统一算法时间积分方案

从欧拉算符出发,用泰勒级数展开,给出二阶时空微商余项预报方程,进而讨论一种数值分析新算法———双三次曲面拟合(插值)的准拉格朗日时间积分方案与数值模式。它是将大气运动描述成为非线性的三次运动,即是通过对原始大气运动方程中包括标量、矢量的压、温、湿、风、以及旋转地球上广义牛顿力加速度场和散度场等,做双三次曲面拟合,实现对各个大气运动变量场的二阶可导,即限定气块上游点在各个不同双三次曲面(片)、具有斜率、曲率和挠率的非线性三次变量场上活动,从而可对各个大气运动方程做时间离散积分,即为双三次曲面拟合—时间步积分—双三次曲面拟合……,实现成为一种新动力框架数值模式。由于双三次曲面具有数学定律收敛性和二阶可导最优性,故选用双三次曲面插值求算二阶余差上游点,具有充分必要的数学理由:它包含了大气运动变量场之斜率、曲率和挠率。因此,埃尔米特双三次曲面片具有对网格变量场二阶可导运算等价性、及其数学收敛性与最佳曲率最优性,并且将准拉格朗日法与欧拉法,以及柯朗-弗里德里希斯-列维判据统一起来。容易实现全球网格变量场的双三次曲面拟合,和可按双三次曲面变量场的斜率、曲率或挠率判断,作变量场局域或单点平滑,以此保持三次模式的时间积分稳定...     

半拉格朗日、半隐式欧拉方程组大气数值模式研究

介绍了一个具有较高时间积分效率的三维弹性大气数值模式。其中,控制方程为能够描述大气非静力及可压缩性的欧拉方程组,时间积分采用了高效率的“半拉格朗日、半隐式”方案。通过引用一个简化的“云物理过程”参数化方案,针对大气中深厚湿对流过程的若干问题进行了数值试验,结果表明,模式能够对该过程中的一些基本和复杂现象进行有效地模拟。试验结果还表明,动力学框架具有很好的稳定性,能够实现高效率的时间积分,它还具有较好的频散特性,保证了空间计算的精度,从而完好地描述了风暴的形态。     

Instability in Lagrangian stochastic trajectory models,and a method for its cure

A number of authors have reported the problem of unrealistic velocities (“rogue trajectories”) when computing the paths of particles in a turbulent flow using modern Lagrangian stochastic (LS) models, and have resorted to ad hoc interventions. We suggest that this problem stems from two causes: (1) unstable modes that are intrinsic to the dynamical system constituted by the generalized Langevin equations, and whose actual triggering (expression) is conditional on the fields of the mean velocity and Reynolds stress tensor and is liable to occur in complex, disturbed flows (which, if computational, will also be imperfect and discontinuous); and, (2) the “stiffness” of the generalized Langevin equations, which implies that the simple stochastic generalization of the Euler scheme usually used to integrate these equations is not sufficient to keep round-off errors under control. These two causes are connected, with the first cause (dynamical instability) exacerbating the second (numerical instability); removing the first cause does not necessarily correct the second, and vice versa. To overcome this problem, we introduce a fractional-step integration scheme that splits the velocity increment into contributions that are linear (U _i ) and nonlinear (U _i U _j ) in the Lagrangian velocity fluctuation vector U, the nonlinear contribution being further split into its diagonal and off-diagonal parts. The linear contribution and the diagonal part of the nonlinear contribution to the solution are computed exactly (analytically) over a finite timestep Δt, allowing any dynamical instabilities in the system to be diagnosed and removed, and circumventing the numerical instability that can potentially result in integrating stiff equations using the commonly applied explicit Euler scheme. We contrast results using this and the primitive Euler integration scheme for computed trajectories in a drastically inhomogeneous urban canopy flow.     

Computing Streamfunction and Velocity Potential in a Limited Domain of Arbitrary Shape. Part II: Numerical Methods and Test Experiments

Built on the integral formulas in Part I,numerical methods are developed for computing velocity potential and streamfunction in a limited domain.When there is no inner boundary(around a data hole) inside the domain,the total solution is the sum of the internally and externally induced parts.For the internally induced part,three numerical schemes(grid-staggering,local-nesting and piecewise continuous integration) are designed to deal with the singularity of the Green’s function encountered in numerical calculations.For the externally induced part,by setting the velocity potential(or streamfunction) component to zero,the other component of the solution can be computed in two ways:(1) Solve for the density function from its boundary integral equation and then construct the solution from the boundary integral of the density function.(2) Use the Cauchy integral to construct the solution directly.The boundary integral can be discretized on a uniform grid along the boundary.By using local-nesting(or piecewise continuous integration),the scheme is refined to enhance the discretization accuracy of the boundary integral around each corner point(or along the entire boundary).When the domain is not free of data holes,the total solution contains a data-hole-induced part,and the Cauchy integral method is extended to construct the externally induced solution with irregular external and internal boundaries.An automated algorithm is designed to facilitate the integrations along the irregular external and internal boundaries.Numerical experiments are performed to evaluate the accuracy and efficiency of each scheme relative to others.     

A closure to derive a three-dimensional well-mixed trajectory-model for non-Gaussian, inhomogeneous turbulence

A three-dimensional Lagrangian stochastic (LS) model to evaluate pollutant dispersion in the atmospheric boundary layer has been developed. The model satisfies the well-mixed criterion of Thomson and allows for inhomogeneous, skew turbulence. Making use of the spherical reference frame, one of the possible solutions has been obtained. A skewed joint probability density function (PDF), which reproduces the given velocity moments (means, variances, skewness and covariances), has been built-up by a linear combination of eight Gaussian PDFs. In order to verify consistency with the well-mixed criterion, the long term results have been compared with the theoretical behaviour. A comparison between our model and Thomson's published algorithms was also carried out. By comparing wind-tunnel data and numerical predictions, a further validation of our LS model has been obtained. From an analysis of the numerical results, we can state that our model is able to evaluate dispersion in the case of complex flows where the application of previous models is unsuccessful.     

污染源、干沉降和气相化学影响硫沉降的数值研究

利用欧拉型区域硫沉降模式对污染源、干沉降和气相化学等物理化学过程在酸沉降中的作用进行了数值分析。结果表明:排放源的强度和高度影响着污染物的分布和输送,其影响是非线性的。在区域酸沉降模式中,干沉降速度和气相化学转化率简单地取为常数对SO#-[2]及SO#+[2-]#-[4]浓度分布引起的误差是不可忽视的。     

Runge-Kutta算法与Li差分法不同阶数配合对计算精度影响研究

为了充分发挥高阶Li空间微分方案（Li, 2005）的优点,实现了时间积分为2~6阶Runge-Kutta（简称RK）格式的偏微分方程求解算法（简称RKL算法）。然后通过多组数值试验,研究了时间积分阶数对计算误差的影响。线性平流方程的试验结果表明对于方波函数型初值,2、4、5和6阶RK算法能获得和3阶精度差不多的结果,而对于高斯函数型的初值,高阶RKL算法可以取得较好的计算效果。RK为5（6）阶时,对应的Li微分阶数可达9（10）阶,总误差控制在10-7（10-8）以内。随RK阶数增加Li微分有效阶数有增加的趋势,而总误差在逐渐减小。计算非线性无粘Burgers方程时,RKL算法能否获得好的计算结果,除了受初始场形式的影响,还与计算的目标时刻有关。当目标时刻解的各阶导数连续（且未出现无穷大数值时）,高阶（RK为4~6阶）算法是有效的;若出现了导数间断、或导数为无穷大,就会碰到冲击波解类型的问题,此时高阶RK算法也无法获得很高精度的数值解。此非线性的算例中,Li微分阶数仍然随RK阶数增加而增加,但增加的趋势不是线性的,具体变化关系可以通过实验结果拟合而获得。研究发现时间积分方案阶数大于3之后,对应的最优空间差分精度阶数可以比6阶提高很多,这再次证明了以前研究中6阶以上空间差分格式对结果无改进的现象,是由于没有使用足够高精度的时间积分方案引起的。相比于Taylor-Li（Wang,2017）算法,5~6阶的RK方法编程和实现简单,计算结果的精度比3阶算法要提高很多,因此,它是一种能够对复杂方程适用的简易高阶算法方案,具有一定的实用价值。     

A Hybrid Spectral/Finite-Volume Algorithm for Large-Eddy Simulation of Scalars in the Atmospheric Boundary Layer

Pseudospectral methods are frequently used in the horizontal directions in large-eddy simulation of atmospheric flows. However, the same approach often creates unphysical oscillations for scalar fields if there are horizontal heterogeneities in the sources and/or sinks, as is usual in air pollution problems. A hybrid approach is developed to combine the use of pseudospectral representation of the velocity field and bounded finite-volumes for the scalar concentration. An interpolation scheme that yields a divergence-free interpolated velocity field is derived and implemented, and its importance is illustrated by two sample applications.     

Lagrangian Stochastic Models For Turbulent Dispersion In The Atmospheric Boundary Layer

A one-particle three-dimensional stochastic Lagrangian model fortransport of particles in a horizontally-homogeneous atmosphericsurface layer with arbitrary one-point probability density functionof Eulerian velocity fluctuations is suggested. A uniquely definedLagrangian stochastic model in the class of well-mixed models isconstructed from physically plausible assumptions. These assumptionsare: (i) in the neutrally stratified horizontally homogeneous surface layer, the vertical motion is mainly controlled by eddies whose size is of order of the current height; and (ii), the streamwise drift term is independent of the crosswind velocity. Numerical simulations for neutral stratification have shown a good agreement of our model with the well-known Thomson's model, with Flesch and Wilson's model, and with experimental measurements as well. However there is a discrepancy of these results with the results obtained by Reynolds' model.     

A study on the motion characteristics and their impact on the wind measurement post-processing of the GPS dropwindsonde. Part II: impacts of post-processing/composition schemes

When the Global Positioning System dropwindsonde wind measurements are used, post-processing and composition approaches are necessary to calculate the vertical profiles of wind statistics. In order to understand the impacts of post-processing/composition schemes on the wind statistics calculated based on dropwindsonde measurements, various post-processing/composition schemes were utilized to post-process and to composite pseudo-dropsonde measurements resulting from numerical simulation of dropsonde motions in a pseudo-stochastic wind field. Since the wind statistics of the pseudo-stochastic wind field were specified, the comparison between wind statistics calculated based on pseudo-dropsonde measurements and the specified values provided some insights into the impacts of different post-processing/composition schemes. The post-processing/composition schemes investigated include the low-pass filter to reduce high-frequency dropsonde measurement noises, the finite difference scheme to calculate dropsonde accelerations, which is needed to dynamically correct dropsonde measurements and composition data size.     

A Parameterization of Dry Thermals and Shallow Cumuli for Mesoscale Numerical Weather Prediction

For numerical weather prediction models and models resolving deep convection, shallow convective ascents are subgrid processes that are not parameterized by classical local turbulent schemes. The mass flux formulation of convective mixing is now largely accepted as an efficient approach for parameterizing the contribution of larger plumes in convective dry and cloudy boundary layers. We propose a new formulation of the EDMF scheme (for Eddy Diffusivity\Mass Flux) based on a single updraft that improves the representation of dry thermals and shallow convective clouds and conserves a correct representation of stratocumulus in mesoscale models. The definition of entrainment and detrainment in the dry part of the updraft is original, and is specified as proportional to the ratio of buoyancy to vertical velocity. In the cloudy part of the updraft, the classical buoyancy sorting approach is chosen. The main closure of the scheme is based on the mass flux near the surface, which is proportional to the sub-cloud layer convective velocity scale w _*. The link with the prognostic grid-scale cloud content and cloud cover and the projection on the non- conservative variables is processed by the cloud scheme. The validation of this new formulation using large-eddy simulations focused on showing the robustness of the scheme to represent three different boundary layer regimes. For dry convective cases, this parameterization enables a correct representation of the countergradient zone where the mass flux part represents the top entrainment (IHOP case). It can also handle the diurnal cycle of boundary-layer cumulus clouds (EUROCS\ARM) and conserve a realistic evolution of stratocumulus (EUROCS\FIRE).     

区域空气质量数值预报模型

给出了区域尺度空气质量数值预报模型的基本物理构架, 重点介绍了欧拉型输送／沉降模式对各种物理 （沉降、输送）、化学 （气相化学与液相化学） 过程、初值与边界条件等问题的处理／简化; 利用本模型作了一周的预报实验, 实验结果与分析说明, 区域空气质量数值预报模型能够较好地反映全国重点城市的空气质量以及污染物随时间的演变态势; 最后, 指出了需要进一步完善的工作。     

Inertial Effects on the Vertical Transport of Suspended Particles in a Turbulent Boundary Layer

In many atmospheric flows, a dispersed phase is actively suspended by turbulence, whose competition with gravitational settling ultimately dictates its vertical distribution. Examples of dispersed phases include snow, sea-spray droplets, dust, or sand, where individual elements of much larger density than the surrounding air are carried by turbulent motions after emission from the surface. In cases where the particle is assumed to deviate from local fluid motions only by its gravitational settling (i.e., they are inertialess), traditional flux balances predict a power-law dependence of particle concentration with height. It is unclear, however, how particle inertia influences this relationship, and this question is the focus of this work. Direct numerical simulations are conducted of turbulent open-channel flow, laden with Lagrangian particles of specified inertia; in this way the study focuses on the turbulent transport which occurs in the lowest few meters of the planetary boundary layer, in regions critical for connecting emission fluxes to the fluxes felt by the full-scale boundary layer. Simulations over a wide range of particle Stokes number, while holding the dimensionless settling velocity constant, are performed to understand the role of particle inertia on vertical dispersion. It is found that particles deviate from their inertialess behaviour in ways that are not easily captured by traditional theory; concentrations are reduced with increasing Stokes number. Furthermore, a similarity-based eddy diffusivity for particle concentration fails as particles experience inertial acceleration, precluding a closed-form solution for particle concentration as in the case of inertialess particles. The primary consequence of this result is that typical flux parametrizations connecting surface emission models (e.g., saltation models or sea-spray generation functions) to elevated boundary conditions may overestimate particle concentrations due to the reduced vertical transport caused by inertia in between; likewise particle emission may be underestimated if inferred from concentration measurements aloft.     

A Homogeneous Langevin Equation Model, Part i: Simulation of Particle Trajectories in Turbulence with a Skewed Velocity Distribution

A Lagrangian stochastic model for the time evolution of the velocity of a fluid particle is presented. This model is based on a one-dimensional generalized Langevin equation, and assumes the velocity probability distribution of the turbulent fluid is skewed and spatially homogeneous. This has been shown to be an effective approach to simulating vertical dispersion in the convective boundary layer. We use a form of the Langevin equation that has a linear (in velocity) deterministic acceleration and a random acceleration that is a non-Gaussian, skewed process. For the case of homogeneous fluid velocity statistics, this 'linear-skewed' Langevin equation can be integrated explicitly, resulting in an efficient numerical simulation method. Model simulations were tested using cases for which exact, analytic statistical properties of particle velocity are known. Results of these tests show that, for homogeneous turbulence, a linear-skewed Langevin equation model can overcome the difficulties encountered in applying a Langevin equation with a skewed random acceleration. The linear-skewed Langevin equation model results are compared to results of a 'nonlinear-Gaussian' Langevin equation model, and show that the linear-skewed model is significantly more efficient.     

Optimizing computations in weather and climate prediction models

Summary To meet the challenge of developing a comprehensive weather and climate prediction model which can give realistic scenarios for many time scales, more computer power than is currently available will be needed. One possibility for alleviating this shortcoming is to increase the integration timestep. We propose and test several methods which may prove useful. One procedure is an expansion of the model dependent variables in a Taylor series. Application of this method to simple models indicates acceptable increases in timestep by a factor of five. A multi-level approach which is less complex to apply gives comparable results and is more successful when high accuracy is desired. To bypass the limiting constraint of the Courant-Friedrichs-Lewy (CFL) condition on gravity waves, an approach is suggested in which the prediction model is represented in its normal modes and the high frequency modes are balanced while the low frequency modes are predicted. Experiments with this procedure are described and in combination with the multi-level integration technique show substantial increases in integration timestep for acceptable integration results, both on the forecast and climate scale. Experiments are now underway applying this process to the NCAR/CCM3, a state-of-the-art model.With 10 Figures     

船舶海平面气压观测资料质量控制

随着数值预报模式不断发展,各类观测资料在数值模式中发挥着越来越重要的作用,船舶观测资料是海上3类观测系统 (卫星、飞机、船舶) 资料之一。为了保障船舶观测资料质量并有效地应用于数值模式,该文依据船舶观测资料的时空分布特征,以及2011年1月和7月T639分析场与观测资料对比分析结果,建立了一套船舶海平面气压资料质量控制方案,包括要素极值范围检查、缺测和冗余资料剔除、背景场一致性检查、测站黑名单建立等,并将该方案应用于2011年2月、6月、8月的观测资料。结果表明:船舶观测资料在时间上不连续且空间分布不均匀,会影响时间一致性检查和空间一致性检查质量控制效果;船舶海平面气压观测资料在所有观测要素中资料量最大,但其缺测和冗余资料量约占50%;黑名单资料的质量控制方案能够有效识别和剔除黑名单资料,且有利于对各测站的检查与维修工作;由于五大湖和大奴湖地区地形高度的影响,在背景场一致性检查过程中需对这些区域的背景场资料进行订正。     

On the Formulation of Lagrangian Stochastic Models of Scalar Dispersion Within Plant Canopies

Lagrangian stochastic models, quadratic in velocity and satisfying the well-mixed condition for two-dimensional Gaussian turbulence, are used to make predictions of scalar dispersion within a model plant canopy. The non-uniqueness associated with satisfaction of the well-mixed condition is shown to be non-trivial (i.e. different models produce different predictions for scalar dispersion). The best agreement between measured and predicted mean concentrations of scalars is shown to be obtained with a small sub-class of optimal models. This sub-class of optimal models includes Thomson's model (J. Fluid Mech. 180, 529–556, 1987), the simplest model that satisfies the well-mixed condition for Gaussian turbulence, but does not include two other models identified recently as being in optimal agreement with the measured spread of tracers in a neutral boundary layer. It is therefore demonstrated that such models are not universal, i.e. applicable to a wide range of flows without readjustment of model parameters. Predictions for scalar dispersion in the model plant canopy are also obtained using the model of Flesch and Wilson (Boundary-Layer Meteorol. 61, 349–374, 1992). It is shown that, when used with a Gaussian velocity distribution or a maximum-missing-information velocity distribution, which accounts for the measured skewness and kurtosis of velocity statistics, the agreement between predictions obtained using the model of Flesch and Wilson and measurements is as good as that obtained using Thomson's model.     

集合平均方法减小混沌系统计算误差的效果研究

研究了Lorenz非线性系统中使用的集合平均方法来减小计算误差的效果,通过检查5组数值试验(每组20个样本)的结果发现:集合平均对计算误差的减小和消除不如高精度算法有效,这主要体现在以下几方面:1)普通的算法和双精度的计算环境中,若截断误差是主导误差(当初值误差很小时),各集合的平均结果并不收敛于真值,而是收敛于含截断误差的数值解;2)若初值误差为主导时,系统受到初值误差增长规律的影响,数值解收敛于由初值误差主导的误差解;3)这两种误差量级接近的时候,两种误差都无法消除掉。对解的统计特征进行研究表明,可信的数值解与含计算误差的数值解有许多相似的地方,但是与集合平均的数值解有很大不同,同样说明了集合平均不适用于减小计算误差这样的问题。此外,试验结果表明即使数值解的概率分布形式基本正确,也不能保证数值解是正确的。