一个改进的诊断层积云方案及其对大气环流模式性能的改进

在气候模式中,传统的层积云参数化方案(简称LTS方案)高估了自由大气湿绝热过程所引起的垂直温度梯度的作用.在Wood和Bretherton提出的"估算逆温"(EIS)基础上,本文基于卫星资料构建了一个改良的诊断层积云方案,旨在改进当前广为使用的LTS方案.通过将EIS层积云方案引入大气环流模式GAMIL2.0中,并与传统方案进行对比试验发现,EIS层积云方案能有效地改善GAMIL2.0对东亚-西北太平洋季风区乃至全球云-辐射气候特征的模拟能力,一定程度地纠正了模式原来存在的中高纬度净(短波)云辐射强迫模拟偏弱的现象,而该偏差在当前许多模式中普遍存在.此改进与EIS层积云方案合理处理自由大气的温度层结有关.EIS层积云方案亦显著改进了GAMIL2.0对东亚-西北太平洋季风区云辐射年际变率的模拟能力.加入EIS层积云方案后,模式较好地模拟出了ENSO年冬春季东亚边缘海域的净云辐射正异常、海表入射短波正异常和层云量负异常.究其原因是EIS层积云方案比传统方案对海表温度异常的扰动更敏感.海表温度异常通过EIS层积云方案在对流层低层强迫出更强的云量距平,并通过改变加热率和垂直速度影响整个对流层,最终改进模式对云辐射强迫和地表入射短波通量年际变率的模拟.EIS层积云方案还增强了模式对El Ni?o年冬春季节西北太平洋异常反气旋以及东亚降水异常的模拟效果.     

LASG耦合气候系统模式FGCM-1.0

本文描述了中国科学院大气物理研究所大气科学和地球流体力学数值模拟国家重点实验室(LASG)最新发展的一个耦合气候系统模式的基本性能. 该模式是在LASG灵活的全球耦合气候系统模式（英文缩写为FGCM）的初始版本FGCM-0的基础上发展而来的,是该系列耦合模式的第二个版本，即FGCM-1.0. FGCM-1.0通过一个通量耦合器将大气、海洋和海冰三个分量模式耦合在一起，其中海洋分量模式是LASG发展的一个涡相容分辨率（eddy-permitting）全球海洋环流模式，大气和海冰分量模式则为美国国家大气研究中心（NCAR）的大气环流模式CAM2和海冰模式CSIM4. 耦合模式完整地考虑了海气界面上的动量、热量和淡水通量交换，尽管在模式中没有使用任何形式的人为的通量调整或者通量距平方案，模式还是比较合理地模拟出基本的气候形态. 通过对该耦合模式长期积分结果的进一步分析发现，模式能够比较好地模拟出厄尔尼诺－南方涛动（ENSO）以及印度洋偶极子事件的基本特征；与FGCM系列耦合模式的最初版本FGCM-0相比，FGCM-1.0模拟的北赤道逆流（NECC）和ENSO循环更加真实.     

边界层参数化方案及海气耦合对WRF模拟东亚夏季风的影响

区域气候模式的边界层参数化方案很大程度上影响着陆地-海洋-大气间水汽、动量及热量的交换,该方案的不确定性会给模式结果带来明显误差.本文基于WRF区域气候模式中四种常用的边界层参数化方案(YSU,ACM2,BouLac和MYJ)分别对东亚夏季风进行模拟研究,分析了不同的边界层方案对东亚夏季风环流及降水模拟的影响.结果表明,局地湍流动能方案BouLac和MYJ对东亚夏季风的模拟结果相对于非局地闭合方案YSU和ACM2更接近于观测,前者能更好的模拟出中国东部中低空西南风气流和西太平洋副热带高压.对于东亚夏季风降水,无论是空间分布还是季节内演变,BouLac和MYJ方案都要明显优于YSU和ACM2.此外,通过对比YSU和BouLac两种方案的模拟结果,发现边界层方案对东亚夏季风的模拟在海洋区域的影响更为显著.造成不同方案模拟差异的主要原因是非局地方案YSU和ACM2的边界层垂直混合偏强,使得海表向上输送的潜热通量明显偏强,对流更活跃,导致降水偏多以及相应季风环流的异常偏差.进一步研究指出缺少海气反馈过程使得WRF模式由边界层方案引起的模拟误差在海洋区域更为突出,引入海气耦合可以减小海表热通量误差并明显改善东亚夏季风的模拟结果.     

积云对流参数化方案对热带大气季节内振荡模拟的影响

本文利用中国科学院大气物理研究所大气科学和地球流体力学数值模拟国家重点实验室（LASG）的大气环流谱模式的最新版本SAMIL2.2.3系统分析了积云对流参数化方案对热带大气季节内振荡（ISO）模拟的影响.结果表明该模式的三种积云对流参数化方案都能模拟出热带地区气候平均季节内振荡（CISO）时空演变的基本特征;其中Tiedtke方案相比于Manabe方案和Zhang-McFarlane方案,模拟出的CISO与观测更为接近.而SAMIL2.2.3模式对ISO年际变化部分（TISO）的模拟结果显示尽管SAMIL2.2.3相对于之前的版本对Tiedtke方案和Zhang-McFarlane方案针对不同问题做了必要调整,但三种积云参数化方案中只有Zhang-McFarlane方案能够模拟出27～50天TISO模态的东传特征.另外,SAMIL2.2.3与其较早版本SAMIL2.08中的Manabe方案完全相同,物理过程的其他部分如辐射方案和气溶胶条件有所调整,但仅后者能模拟出100E以西热带27～50天TISO模态的东传特征.因此,模式发展过程中物理参数化方案的改进并非在每一方面上都产生正面效果.动力框架和物理过程,以及不同物理过程之间的协调发展是改进模式的重要途径.     

基于静止轨道卫星监测中国大气污染物的模拟分析

针对对流层主要污染物(对流层臭氧、NO_2、SO_2及HCHO)的高分辨率、高频次监测需求,开展了基于静止轨道卫星监测中国大气污染物的模拟观测实验.在参考现有极轨卫星载荷的遥感监测能力和仪器参数基础上,根据静止卫星观测模型,利用正演辐射传输模式模拟观测光谱.根据最优估计反演理论,分析光谱仪的主要指标参数对各目标气体的反演敏感性和反演误差.基于各目标气体反演精度需求,提出光谱仪的观测波段、光谱分辨率和信噪比等关键指标参数的建议方案.为评估选定的光谱仪参数能否满足需求,利用大气化学输送模式模拟中国区域大气成分的三维分布,进行中国区域的模拟反演实验.结果表明,针对各目标气体,满足反演需求的实验数目均达到90%左右.臭氧总量和平流层臭氧的反演精度可达到2%,且各气体反演在太阳天顶角高于70°范围内仍可以获得有效数据.因此,我们所提出的静止卫星观测方案和仪器参数方案具有一定可行性,为优化将来中国静止轨道卫星光谱仪仪器指标提供了理论依据和模拟工具.     

FY-2D红外亮温模拟及对WRF模式云预报误差的响应

基于WRF和RTTOV模式实现了FY-2D红外模拟亮温图的制作,侧重分析了WRF模式微、宏观云参量模拟误差对FY-2D红外模拟亮温精度的影响.结果表明,在0~48h模拟时效内,本文模拟亮温的均方根误差基本上控制在10~27K,好于以往20~40K的研究结果;在0~24h,各通道模拟与观测亮温值的相关系数均大于0.5;相比其他3个通道,水汽通道模拟性能比较稳定,效果最好.4种云微物理方案的模拟结果均显示云区亮温模拟值偏大,云系分布不够饱满,尤其是在台风区;其中THOM方案的模拟效果最好;数值范围253~273K是各方案存在偏差最大的区间.微观方面,THOM方案的雪水物质混合比远大于其他方案,提高云微物理方案中云中固态水的生成效率或保有量是有必要的,但对亮温模拟影响不大.宏观方面,本文云量诊断方案获得云量数值偏小,对高云诊断方案的改进十分必要,选取精准的云量诊断方案更有利于提高模拟精度.研究结果可为预报员研判未来天气提供了一种更直观的参考资料,也可为开展FY-4红外模拟亮温图制作奠定了技术积累.     

GNSS掩星中大气水汽的非线性反演

研究利用无线电掩星探测数据反演大气水汽参数,文中使用一维非线性方法,反演对流层水汽压廓线.首先,定义了一维目标函数,并以几何光学假设为前提,反演出电波弯曲角;将电波弯曲角作为掩星探测量,大气模式输出的温、湿参量作为初始场,代入目标函数,对目标函数求最优反演出水汽压廓线.文中给出了部分反演结果,经讨论分析认为:非线性方法反演水汽压对初始场的精度依赖较小,能够同时反演出大气温度\水汽压廓线及相应的误差分析,对数值天气预报的应用及数据同化研究具有重要意义.     

一种基于数据同化优化陆面过程模式参数化方案的方法

基于数据同化优化陆面过程模式参数,从物理机制上改进和完善陆面过程模式参数化方案,提高模式的区域适应性和模拟能力,对改善大气环流模式(GCM,General Circulation Model)和区域气候模式(RCM,Regional Climate Model)模拟和气候预测能力具有重要的现实意义.本文发展了一种基于数据同化优化陆面过程模式参数化方案的方法,并以优化非饱和土壤水模型为例,发展基于非饱和土壤水模型和扩展Kalman滤波(EKF,Extended Kalman Filter)算法并结合可变下渗能力模型(VIC,Variable Infiltration Capacity Model)的土壤湿度同化方案,以月为同化窗,利用复合形混合演化算法(SCE-UA,Shuffled Complex Evolution-University of Arizona)极小化目标函数,使模拟与同化的土壤湿度按给定的目标函数度量方式达到最佳拟合,优化非饱和土壤水模型参数:饱和水力传导率、饱和土壤湿度、饱和土壤水势、Clapp和Hornberger常数,得到同化时段1986年模型参数的最优值序列,然后用其改进非饱和土壤水模型,最后利用改进后模型进行1986~1993年的数值模拟试验,评估改进后模型的模拟能力,探讨基于数据同化优化陆面过程模式土壤水参数化方案的方法与实现途径.试验结果表明:基于数据同化优化模型参数,从物理机制上改进和完善模型,提高了模型的模拟能力;优化的模型参数在时间上具有较好的移植性,适应性更强,达到了改进模型的目的,因此,本方法是合理、可行的.本文为基于数据同化优化陆面过程模式参数化方案研究提供了范例,具有较好的借鉴作用.     

影响气候系统模式温室气体敏感度的反馈过程——基于FGOALS模式的研究

气候系统模式对特定温室气体的敏感度对未来气候变化的预估结果至关重要.为理解影响气候模式对温室气体敏感度的关键反馈过程,本文利用瞬间4倍CO2强迫试验,研究了LASG/IAP两个版本的气候系统模式FGOALS-s2和FGOALS-g2的气候响应.将全球平均地表气温(SAT)的变化作为衡量气候模式响应的主要标准.根据SAT的响应随时间的变化趋势,将150年的试验结果分为两个时段:前20年为快响应阶段,后130年为慢响应阶段.采用国际通用的Gregory方法分别估计4倍CO2引起的辐射强迫和平衡态气候敏感度.结果表明,FGOALS-s2中水汽的响应偏强使估算的CO2辐射强迫比FGOALS-g2强7.1%.平衡态气候敏感度定义为2倍CO2强迫下达到新的平衡态时全球平均SAT的变化,其在FGOALS-s2和FGOALS-g2中分别约为4.5和3.7 K,前者较之后者偏强21.6%.FGOALS-s2敏感度大是由于快响应阶段的水汽和反照率正反馈过程偏强,表现为水汽含量的增加和海冰的减少均太快.在慢响应阶段,尽管FGOALS-s2的水汽及反照率正反馈仍然较强,但FGOALS-s2的总负反馈强于FGOALS-g2,这是因为FGOALS-s2中云短波负反馈更强,补偿了其较弱的晴空负反馈.FGOALS-s2中云短波负反馈偏强主要是由于总云量以及云水路径正响应过强.云短波反馈的不确定性是总反馈不确定性的主要来源.     

南京城市化进程对大气边界层的影响研究

为了研究城市化进程对城市边界层结构的影响，通过高分辨率的卫星遥感资料获得土地利用类型,以及地表反照率、植被指数等地表参数，以南京为例，运用数值模拟手段进行了研究，结果显示这是一种可行的手段.数值模拟结果表明，随着城市的发展，城市反照率减小、植被减少、地表湿度减小，蒸发耗热减少，感热通量增加140 W/m2，城市波恩比增加.地表和大气之间热量交换加强，湍流热量通量增大了一倍，湍流交换发展加剧，14：00混合层高度抬高了500 m.地表湍流水汽通量和空气中水汽含量都减小，这也使更多的热量用于加热地表和大气，使地温、气温的日变化幅度增加.     

An improved diagnostic stratocumulus scheme based on estimated inversion strength and its performance in GAMIL2

Based on satellite data and the estimated inversion strength (EIS) derived by Wood et al. (2006), a feasible and uncomplicated stratocumulus scheme is proposed, referred to as EIS scheme. It improves simulation of cloud radiative forcing (CRF) in the Grid-point Atmospheric Model of IAP/LASG version 2 (GAMIL2.0) model. When compared with the original lower troposphere stability (LTS) scheme, the EIS scheme reproduces more reasonable climatology distributions of clouds and CRF. The parameterization partly corrects CRF underestimation at mid and high latitudes and overestimation in the convective region. Such improvements are achieved by neglecting the effect of free-tropospheric stratification changes that follow a cooler moist adiabat at middle and high latitude, thereby improving simulated cloudiness. The EIS scheme also improves simulation of the CRF interannual variability. The positive net CRF and negative stratiform anomaly in the East Asian and western North Pacific monsoon regions (EAWNPMR) are well simulated. The EIS scheme is more sensitive to sea surface temperature anomalies (SSTA) than the LTS. Therefore, under the effect of a warmer SSTA in the EAWNPMR, the EIS generates a stronger negative stratiform response, which reduces radiative heating in the low and mid troposphere, in turn producing strong subsidence and negative anomalies of both moisture and cloudiness. Consequent decreases in cloud reflection and shading effects ultimately improve simulation of incoming surface shortwave radiative fluxes and CRF. Because of the stronger subsidence, a stronger anomalous anticyclone over the Philippines Sea is simulated by the EIS run, which leads to a better positive precipitation anomaly in eastern China during ENSO winter.     

晴天及阴天条件下WRF模式中几种边界层参数化方案的对比分析研究

利用WRF模式中5种常用边界层参数化方案(ACM2、YSU、BouLac、MYJ和MYNN2.5)及美国能源部大气辐射观测试验(ARM)寿县综合观测数据(2008年8—12月),对比分析了晴天及阴天条件下,不同参数化方案对典型农田下垫面气象要素及边界层结构的模拟效果,结果表明:(1)模式对于云层状况的模拟,非局地方案ACM2和YSU方案优于局地方案.(2)对于近地层气象要素,晴天和阴天条件下均是ACM2方案对于2 m温度和比湿的模拟效果最好,MYJ方案对于风向风速的模拟效果最好.(3)对于位温及比湿垂直廓线的模拟,晴天和阴天条件下均是非局地方案(ACM2和YSU)对白天的模拟效果优于局地方案;ACM2方案对夜间弱稳定层结和逆湿结构的模拟最优;(4)对于风速垂直廓线的模拟,白天不稳定条件下,晴天条件MYJ方案最优,阴天MYNN2.5方案的模拟效果最好;夜间弱稳定条件下,晴天条件ACM2方案与观测值之间的偏差最小,阴天YSU方案模拟效果最好;(5)总体而言,在对典型农田下垫面进行模拟时,晴天和阴天条件下均是ACM2方案更具优势.     

Reducing the biases in shortwave cloud radiative forcing in tropical and subtropical regions from the perspective of boundary layer processes

Biases in shortwave cloud radiative forcing (SWCF), which cause overestimates in tropical regions and underestimates in subtropical marine stratocumulus regions, are common in many climate models. Here, two boundary layer processes are investigated in the atmospheric model GAMIL2, entrainment at the top of the boundary layer and longwave radiative cooling at the top of stratocumulus clouds, in order to reduce biases and reveal the mechanisms underlying these processes. Our results show that including the entrainment process in the model can reduce negative SWCF biases in most tropical regions but increases positive SWCF biases in subtropical marine stratocumulus regions. This occurs because entrainment reduces the low-level cloud fraction and its cloud liquid water content by suppressing the vertical turbulent diffusion in the boundary layer and decreasing the relative humidity when warm and dry free atmosphere is entrained in the boundary layer. Longwave radiative cooling at the top of stratocumulus clouds can enhance turbulent diffusion within the stratocumulus-topped boundary layer. When combined with the entrainment process, longwave radiative cooling reduces the positive SWCF biases in subtropical marine stratocumulus regions that are observed using the entrainment process alone. The incorporation of these two boundary layer processes improves the simulated SWCF in tropical and subtropical regions in GAMIL2.     

Effects of alternative cloud radiation parameterizations in a general circulation model

Using the National Center for Atmospheric Research (NCAR) general circulation model (CCM2), a suite of alternative cloud radiation parameterizations has been tested. Our methodology relies on perpetual July integrations driven by ±2 K sea surface temperature forcing. The tested parameterizations include relative humidity based clouds and versions of schemes involving a prognostic cloud water budget. We are especially interested in testing the effect of cloud optical thickness feedbacks on global climate sensitivity. All schemes exhibit negative cloud radiation feedbacks, i.e., cloud moderates the global warming. However, these negative net cloud radiation feedbacks consist of quite different shortwave and longwave components between a scheme with interactive cloud radiative properties and several schemes with specified cloud water paths. An increase in cloud water content in the warmer climate leads to optically thicker middle- and low-level clouds and in turn negative shortwave feedbacks for the interactive radiative scheme, while a decrease in cloud amount leads to a positive shortwave feedback for the other schemes. For the longwave feedbacks, a decrease in high effective cloudiness for the schemes without interactive radiative properties leads to a negative feedback, while no distinct changes in effective high cloudiness and the resulting feedback are exhibited for the scheme with interactive radiative properties. The resulting magnitude of negative net cloud radiation feed-back is largest for the scheme with interactive radiative properties. Even though the simulated values of cloud radiative forcing for the present climate using this method differ most from the observational data, the approach shows great promise for the future.     

Wind flow and wind loads on the surface of a towershaped building: Numerical simulations and wind tunnel experiment

Flow structure and wind pressure distribution caused by obtuse obstacles are usually the focuses in Computational Wind Engineer researches (CWE). By solving the non-hydrostatical dynamic equations, PUMA model (Peking University Model of Atmospheric Environment) was developed and applied to simulating the flow structure and wind pressure distribution around a tower-shaped building. Evaluation about the wind environment and wind loads around the building was obtained through the analysis of the numerical simulation results and wind tunnel data. Comparisons between the simulation and wind tunnel study indicate that numerical simulation results agree well in the flow field and wind pressure distribution around the tower-shaped building. On the other hand, the horizontal grid interval of 2 m and the vertical grid of 3 m were still too crude to simulate the flow structure and wind pressure distribution on the building surface more exactly in detail; and the absence of suitable pressure perturbation parameterization scheme between the solid and the adjacent space also limits the accuracy of the numerical simulation. The numerical simulation model can be used to evaluate the wind environment and wind load around high buildings.     

CAM-EULAG: A non-hydrostatic atmospheric climate model with grid stretching

This study evaluates the capability of a non-hydrostatic global climate model with grid stretching (CEU) that uses NCAR Community Atmospheric Model (CAM) physics and EULAG dynamics. We compare CEU rainfall with that produced by CAM using finite volume dynamics (CFV). Both models simulated climate from 1996 to 2000, using the same parameterization schemes.     

Sensitivity of Typhoon Vamei (2001) Simulation to Planetary Boundary Layer Parameterization Using PSU/NCAR MM5

This paper investigates the sensitivity of the numerical simulations of a near equatorial Typhoon Vamei (2001) to various planetary boundary layer (PBL) parameterization schemes in the Pennsylvania State University (PSU)/National Centre for Atmospheric Research (NCAR) non-hydrostatic mesoscale model (MM5). The numerical simulations are conducted on two domains at 45 and 15 km grids nested in a one-way fashion. Four different PBL parameterization schemes including the Blackadar (BLK) scheme, the Burk–Thompson (BURKT) scheme, the NCEP Eta model scheme and the NCEP medium range forecast (MRF) model scheme are investigated. Results indicate that the intensity and propagation track of the simulated near equatorial typhoon system is not very sensitive to the different PBL treatments. The simulated minimum central pressures and the maximum surface wind speeds differ by only 5–6 hPa and 6–8 ms⁻¹, respectively. Larger variations between the simulations occur during the weakening phase of the typhoon system. While all schemes simulated the typhoon with reasonable accuracy, the ETA scheme produces the strongest storm intensity with the largest heat exchanges over the marine environment and the highest warm moisture air content in the PBL around the core of the storm.     

Parameterizations of cloud feedback in a radiative-convective model

The effect of cloud feedback on the response of a radiative-convective model to a change in cloud model parameters, atmospheric CO₂ concentration, and solar constant has been studied using two different parameterization schemes. The method for simulating the vertical distribution of both cloud cover and cloud optical thickness, which depends on the relative humidity and on the saturation mixing ratio of water vapor, respectively, is the same in both approaches, but the schemes differ with respect to modeling the water vapor profile. In scheme I atmospheric water vapor is coupled to surface parameters, while in scheme II an explicit balance equation for water vapor in the individual atmospheric layers is used. For both models the combined effect of feedbacks due to variations in lapse rate, cloud cover, and cloud optical thickness results in different relationships between changes in surface temperature, planetary temperature, and cloud cover. Specifically, for a CO₂ doubling and a 2% increase in solar constant, in both models the surface warming is reduced by cloud feedback, in contrast to no feedback, with the greater reduction in scheme I as compared to that of scheme II.     

Sensitivity of Indian Monsoon to Entrainment and Detrainment in Mass Flux Schemes

A new analytical formulation of entrainment and detrainment in the Tiedtke's mass flux cumulus parameterization is presented here in which cloud height is one of the key parameters. The proposed analytical profiles of entrainment and detrainment are tested in GCM for long-term simulation and are evaluated in the light of the results from the original Tiedtke's scheme and another mass flux scheme due to Emanuel. The variations of Indian monsoon rainfall have been examined with these schemes in a general circulation model. Evaluation of the simulated rainfall against observations is done by empirical orthogonal function (EOF) analysis for the Indian Monsoon region. It is noted that the spatial and temporal variations of the all-India monsoon rainfall are sensitive to the formulation of entrainment and detrainment in a mass flux scheme, and that the new formulation can effectively represent the increased dilution with height in deep clouds.