尺度自适应大气边界层参数化改进及其对一次海雾的数值模拟研究

大气边界层湍流运动是地球大气运动最重要的能量输送过程之一.当数值模式分辨率接近活跃含能湍涡长度尺度时,湍流运动被部分解析,被称为"灰色区域",传统的边界层方案不适合此时模式湍流问题的描述.为了提高模式边界层方案在包括"灰色区域"的不同网格尺度上的描述能力,适应不同分辨率模式的需要,在雷诺平均湍流理论基础上,修正Mell...     

不同嵌套方式下的积云对流方案对上海崇明极端 暴雨高分辨率模拟的影响研究

使用1980～2017年共38年崇明站逐日降水资料对崇明站年降水量及暴雨日数的特征进行分析,并使用中尺度数值预报模式WRF3.9.1.1(Weather Research and Forecasting model)针对崇明年降水量及暴雨日数异常年份2015年的最强降水过程进行数值模拟,结合站点降水观测资料使用统计方法...     

基于WRF模式的对流尺度边界层方案参数随机扰动方法研究

边界层参数化方案的准确性会影响模式对近地面变量和大气低层热动力结构的模拟,对雷暴等强对流天气的预报非常重要,但边界层方案内在的不确定性使得单一预报具有局限性。为了提高对流尺度数值模式中边界层方案的预报效果,基于WRF（The Weather Research and Forecasting Model）模式,应用随机参数扰动（SPP）方法对Mellor-Yamada-Nakanishi-Niino（MYNN）边界层方案中重要的3个不确定参数进行扰动,探究了该方法对北京地区一次雷暴过程模拟的影响。同时考虑了对流尺度集合预报系统的特点,调整随机参数扰动方法的3个参量（去相关时间尺度、空间尺度和格点标准差）探究了对流尺度中对MYNN方案参数进行扰动的最优设置。结果显示:随机扰动MYNN边界层方案参数（SPPM）方法可以有效提高近地面变量和700 hPa以下低层变量的离散度,同时提高了短时强降水位置和强度的预报技巧。3个参量的试验说明,去相关时间尺度增大到12 h集合离散度有明显提高;格点标准差增大到0.20,预报技巧也略有提高;去相关空间尺度维持在默认值700 km较好,尺度过小（150 km）预报技巧明显降低。上述结果表明,在对流尺度中SPPM方法可以有效表达边界层参数化方案的不确定性,提高集合预报系统的预报技巧。      

Role of convective parameterization in simulations of heavy precipitation systems at grey-zone resolutions — case studies

We have investigated the role of convective parameterization in simulations of heavy precipitation systems at grey-zone (2–10 km) resolutions using an approach similar to that used in “observing system simulation experiment”. Simulations with a 1-km grid serve as benchmark simulations. The impacts of convective parameterization at greyzone resolutions (i.e., 3, 6, and 9 km) are then investigated. This study considers two heavy precipitation systems including one associated with a mesoscale cyclone generated over the Shandong Peninsula on 24–25 July 1991, and the other associated with a cloud cluster occurred on 15–16 July 2009. The present study indicates that convective parameterization does not affect much the simulations of the two heavy precipitation systems with 3-km grid size. However, it significantly affects simulations for grid sizes of 6 and 9 km. Simulations with the Kain-Fritsch scheme produce deficiencies such as relatively small heavy rainfall area, smaller maximum precipitation rate, wider area of weak precipitation, etc. Simulations without convective parameterization have also some negative effects such as the overprediction of area-averaged precipitation rate and others. A modified trigger function in the Kain-Fritsch scheme is found to improve the simulations of the heavy precipitation systems over the Korean Peninsula by reducing excessive trigger of convection, especially for simulations with 6- and 9- km grids.     

不同降水方案对"03.7"一次暴雨过程模拟的影响

观测和数值模拟研究已经表明,潜热释放对中国东部夏季梅雨锋系统及其锋面降水的维持和发展发挥着非常重要的作用.然而,目前对于梅雨锋降水模拟中各种降水方案的相互协调和系统评估方面的工作仍不多见,为了增进对梅雨锋暴雨模拟中降水过程的认识,作者针对2003年7月4～5日一次梅雨锋暴雨过程,构造了四组试验,利用MM5模式考察了两种分辨率(36 km、12 km),各种隐/显式方案搭配下,对所生成的雨带、雨量和降水类型的配置进行了仔细的研究,得到了一些有意义的结论,为今后更好地使用模式、利用数值模式来认识中尺度降水过程中的气象问题打下基础.主要结论包括:模拟总降水的水平分布和强度,以及显式降水和隐式降水的划分对积云参数化方案的选择非常敏感.但对特定积云参数化方案而言,降水的模拟对36 km、12 km水平分辨率不敏感(除Betts-Miller方案外);在中尺度网格分辨率10～50 km范围内,不同积云参数化方案对梅雨锋降水分布和降水量模拟的影响比不同显式方案带来的变化大得多.     

8.19华北暴雨模拟中微物理方案的对比试验

在中尺度模式多种物理过程中,微物理过程是一个非常关键的环节,其不仅直接影响降水预报,而且也影响模式的动力过程.微物理方案有明确的物理基础,但是在实际暴雨模拟中,究竟采取哪一种方案的结果更理想,需要深入比较,因为不同的微物理方案对降水模拟结果有着很大的差异.本文利用中尺度非静力模式WRF (V3.2.1版本),采用36 km、12 km和4 km的格点分辨率,选用七种微物理方案,对2010年8月18~19 日华北地区的暴雨过程进行了敏感性试验.从降水落区和强度方面对总降水的预报性能进行了对比,模拟结果表明:选用不同的微物理方案,可以不同程度地模拟这场暴雨的范围和强度,且选择合理的微物理方案对细网格（4 km）嵌套的模拟也可以相应的提高,从而提高了暴雨模拟的分辨率,为暴雨中小尺度成因分析提供了参考.其中,水平分辨率为36 km时,Lin方案模拟的雨带范围和降水强度与实况拟合的最好;水平分辨率为12 km时,Thompson方案模拟的强降水位置、强度与实况最为接近;而水平分辨率为4 km时,WSM6方案模拟的强降水位置、强度与实况拟合得较好.再结合垂直速度、涡度、散度和雨水混合比等基本物理量的诊断分析,可以更好地理解各微物理方案对降雨预报的影响,所得的结论对我国华北暴雨强降水预报和中尺度模式微物理过程在业务和研究方面有相当的参考价值.     

MM5 simulations of interannual change and the diurnal cycle of southern African regional climate

Summary Two cumulus convection and two planetary boundary layer schemes are used to investigate the climate of southern Africa using the MM5 regional climate model. Both a wet (1988/89) and a dry (1991/92) summer (December–February, DJF) rainfall season are simulated and the results compared with three different observational sources: Climate Research Unit seasonal data (precipitation, 2 m surface temperature, number of rain days), satellite-derived diurnal precipitation and the Surface Radiation Budget diurnal short-wave fluxes and optical depth. Using the ETA model boundary layer in MM5 simulates too much incident short-wave radiation at the surface at 12 UTC, whereas the medium range forecast model boundary layer yields a diurnal cycle of short-wave radiation closer to the observed. The Betts-Miller convection scheme in MM5 simulates peak rainfall later in the day and less rain days than observed, whereas when using the Kain-Fritsch convection scheme a peak rainfall earlier in the day and more rain days than observed are simulated. The intensity of the hydrological cycle is therefore dependent on the choice of convection scheme, which in turn is further modified by the boundary layer scheme. Precipitation during the wet 1988/89 season is reasonably captured by most simulations, though using the Betts-Miller scheme more accurately simulates rainfall during the dry 1991/92 season. Mean DJF biases in the surface temperature and diurnal temperature range are consistent with biases in the number of rain days and the diurnal cycles of surface moisture and energy.     

Evaluation of Heavy Rainfall Model Forecasts over the Korean Peninsula Using Different Physical Parameterization Schemes and Horizontal Resolution

In this study, the accuracy of a Pennsylvania State University-National Center for Atmospheric Research mesoscale model (PSU/NCAR MM5) for predicting heavy summer precipitation over the Korean Peninsula was investigated. A total of 1800 simulations were performed using this model for 30 heavy rainfall events employing four cumulus parameterization schemes (CPS), two grid-scale resolvable precipitation schemes (GRS), and two planetary boundary layer (PBL) schemes in three model resolutions (90 km, 30 km, and 10 km). The heavy rainfall events were mesoscale convective systems developed under the influence of mid-latitude baroclinic systems with low-level moisture transport from the ocean. The predictive accuracy for maximum rainfall was approximately 80% for 10-km resolution and was 60% for 30-km resolution. The predictive accuracy for rainfall position extended to ～150 km from the observed position for both resolutions. Simulated rainfall was most sensitive to CPS, then to PBL schemes, and then to GRS. In general, the Grell (GR) scheme and the Anthes and Kuo (AK) scheme showed a better prediction capability for heavy rainfall than did the Betts-Miller (BM) scheme and the Kain-Fritsch (KF) scheme. The GR scheme also performed well in the 24-h and 12-h precipitation predictions: the parameterized convective rainfall in GR is directly related to synoptic-scale forcing. The models without CPS performed better for rainfall amounts but worse for rainfall position than those with CPS. The MM5 model demonstrated substantial predictive capacity using synoptic-scale initial conditions and lateral boundary data because heavy summer rainfall in Korea occurs in a strong synoptic-scale environment.     

Extreme precipitation in WRF during the Newcastle East Coast Low of 2007

In the context of regional downscaling, we study the representation of extreme precipitation in the Weather Research and Forecasting (WRF) model, focusing on a major event that occurred on the 8^th of June 2007 along the coast of eastern Australia (abbreviated “Newy”). This was one of the strongest extra-tropical low-pressure systems off eastern Australia in the last 30 years and was one of several storms comprising a test bed for the WRF ensemble that underpins the regional climate change projections for eastern Australia (New South Wales/Australian Capital Territory Regional Climate Modelling Project, NARCliM). Newy provides an informative case study for examining precipitation extremes as simulated by WRF set up for regional downscaling. Here, simulations from the NARCliM physics ensemble of Newy available at ～10 km grid spacing are used. Extremes and spatio-temporal characteristics are examined using land-based daily and hourly precipitation totals, with a particular focus on hourly accumulations. Of the different physics schemes assessed, the cumulus and the boundary layer schemes cause the largest differences. Although the Betts-Miller-Janjic cumulus scheme produces better rainfall totals over the entire storm, the Kain-Fritsch cumulus scheme promotes higher and more realistic hourly extreme precipitation totals. Analysis indicates the Kain-Fritsch runs are correlated with larger resolved grid-scale vertical moisture fluxes, which are produced through the influence of parameterized convection on the larger-scale circulation and the subsequent convergence and ascent of moisture. Results show that WRF qualitatively reproduces spatial precipitation patterns during the storm, albeit with some errors in timing. This case study indicates that whilst regional climate simulations of an extreme event such as Newy in WRF may be well represented at daily scales irrespective of the physics scheme used, the representation at hourly scales is likely to be physics scheme dependent.     

WRF模式对高影响浙江型台风微物理和边界层参数化方案的优化试验

选取近年来在浙江和福建登陆高影响浙江的9个西太平洋台风,利用WRF中尺度数值模式,选取多种微物理和边界层参数化方案,量化评估不同方案下的台风路径、强度和降水的模拟性能。结果表明,Kessler和SBU-YLin微物理方案对台风路径和强度模拟较好;而SBU-YLin在浙江降水上比Kessler评分相对更优、误差更小、相关性更佳,但其漏报率相对较高。MYNN2和BouLac边界层方案的最优台风路径和强度比例较高,而BouLac对浙江降水有更好的表现,其降水误差更小、相关性更好,且有更高的TS评分,特别是大雨以上量级或台风登陆前12 h。因此,SBU-YLin微物理和BouLac边界层方案相对更适合浙江台风模拟。此外,浙江沿海地区模拟雨量比实况偏少,且绝对误差较大,内陆反之;雨量误差与浙江地形有较好的对应关系。      

边界层参数化方案在“灰色区域”尺度下的适用性评估

随着数值预报模式分辨率的提高,当模式网格距与含能湍涡的长度尺度相当时,模式动力过程可解析一部分湍流运动,而剩余的湍流运动仍需参数化,此时便产生了湍流参数化的“灰色区域”问题。对传统的PBL（Planetary Boundary Layer）方案在“灰色区域”下的适用性评估,是改进PBL方案以使其能够适应分辨率变化的前提和基础。本研究基于干对流边界层的大涡模拟试验,比较了WRF（Weather Research and Forecast Model）模式中四种常用的边界层参数化方案[YSU（Yonsei University）、MYJ（Mellor-Yamada-Janjic）、MYNN2.5（Mellor-Yamada-Nakanishi-Niino Level 2.5）、MYNN3）]在“灰色区域”尺度下的表现。研究表明,混合层内总热通量对所使用的参数化方案和水平分辨率均不敏感。不同参数化方案中次网格与网格通量的比例表现出对水平网格距不同的依赖性。局地PBL方案（MYJ、MYNN2.5）在混合层内的平均位温随网格距减小而增大,次网格通量随网格距减小而减小,较参考湍流场对次网格通量有所低估。YSU方案的非局地项几乎不随水平格距改变而变化,对次网格通量的表征并未表现出较强的分辨率依赖性,且过强的非局地次网格输送使混合层内温度层结呈弱稳定,抑制了可分辨湍流输送,不易于激发次级环流。MYNN3方案的非局地次网格通量（负梯度输送项）随网格距减小而减小,使其对次网格通量的表征具有较好的分辨率依赖性。PBL方案在“灰色区域”尺度下的适用性与具体分辨率有关。以分辨率500 m为例,四种PBL方案中不存在一种最佳方案,能对边界层的热力结构和湍流统计特征均有准确的描述。     

物理过程参数化方案对中尺度暴雨数值模拟影响的研究

利用中尺度非静力MM 5模式和中国 2 0 0 1年 8月的 4个暴雨个例 ,研究了非绝热物理过程对中国暴雨动力和热力场预报的影响 ,深入分析了对流参数化方案在中尺度暴雨预报中的作用 ,讨论了利用模式扰动方法开展中国暴雨集合预报的可行性。结果表明 ,在短期数值预报中 ,非绝热物理过程对高度场预报影响较小 ,但边界层方案和对流参数化方案对产生暴雨的 3个基本条件即水汽通量散度、垂直速度、不稳定层结的影响很明显。不同对流参数化方案所预报的中尺度热力、动力场离差的结构特征与所预报降水的离差特征相似 ,且主要是在模式积分初期迅速增加 ,其后即趋于稳定。对中国热力场较均匀的暴雨过程 ,可以通过扰动模式的边界层和对流参数化方案 ,构造集合预报模式     

春季黄海海雾WRF参数化方案敏感性研究

利用2005—2011年10次春季黄海海雾个例开展WRF模式参数化方案敏感性研究。结果表明:边界层方案对WRF模式雾区模拟结果起决定作用,而微物理方案影响较小,它主要影响海雾浓度和高度。边界层与微物理方案的最佳组合为YSU与Lin方案,最差为Mellor-Yamada与WSM5方案;Mellor-Yamada和QNSE方案模拟的近海面湍流过强,导致边界层过高,不利于海雾的发展与维持;而MYNN与YSU方案刻画的湍流强度与边界层高度合适,有利于海雾发展与维持。MYNN方案虽与YSU方案相当,但在大多数海雾个例中,后者明显优于前者,而在有些个例中却刚好相反。因此对于某一具体海雾个例而言,所用边界层方案仍需在它们之中选择最优者。这些信息可为黄海海雾WRF模式边界层与微物理方案的选择与改进提供参考。     

Sensitivity Analysis of Ensemble Simulations on a Torrential Rainfall Case over South China Using Multiple PBL and SL Parameterizations

A good representation of the interaction between the planetary boundary layer(PBL) and the surface layer(SL) in numerical models is of great importance for the prediction of the initiation and development of convection. This study examined an ensemble that consists of the available suites of PBL and SL parameterizations based on a torrential rainfall event over south China. The sensitivity of the simulations was investigated against objective measurements using multiple PBL and SL parameterization schemes. The main causes of the bias from different parameterization schemes were further analysed by comparing the good and bad ensemble members. The results showed that good members tended to underestimate the rainfall amount but presented a decent evolution of mesoscale convective systems that were responsible for the torrential rainfall. Using the total energy mass flux(TEMF) scheme, the bad members overestimated the amount and spatial coverage of rainfall. The failure of the bad member was due to a spurious convection initiation(CI) resulting from the overestimated high-θe elevated air. The spurious CI developed and expanded rapidly, causing intensive and extensive rainfall over south China. Consistent with previous studies, the TEMF scheme tends to produce a warmer and moister PBL environment. The detailed sensitivity analysis of this case may provide reference for the operational forecast of rainfall over south China using multiple PBL and SL parameterizations.     

塔克拉玛干沙漠腹地大气边界层参数化方案的模拟评估

沙尘起沙、沉降、传输均受到沙漠地区大气边界层条件的制约。沙漠地区观测资料匮乏,限制大气边界层模拟效果的检验和评估。利用WRFV3.7.1中尺度数值模式中5种边界层参数化方案(ACM2、BL、MYJ、MYNN2.5、YSU),模拟2014年4月塔克拉玛干沙漠大气边界层特征,并与塔中80 m塔及风廓线雷达晴朗天气下的观测资料对比分析。结果表明:5种方案均能模拟出近地面气温及地表温度,边界层高度,感热、潜热、地表热通量的变化趋势,但未能模拟出边界层风速的日变化趋势,温风湿廓线能较好的反映晴日沙漠地区边界层结构的变化特征,但未模拟出风速随高度变化趋势。沙漠地区下垫面干燥,热容量低,晴天极易形成对流不稳定边界层,非局地湍流参数化方案,ACM2方案是沙漠地区大气边界层模拟较为合理的选择。     

不同边界层参数化方案对一次梅雨锋暴雨过程湍流交换特征模拟的影响

利用WRF模式结合不同的边界层参数化方案,对2007年7月3—5日发生在江淮流域的一次梅雨锋暴雨过程进行多组数值模拟试验。结果发现,边界层方案的选取对于降水的落区和强度模拟会产生较显著的影响;在降水率及地面要素的模拟上,各方案在降水中后期的模拟差异明显大于降水发生阶段;不同边界层方案的选取对于降水时段内的水平风场、垂直运动和假相当位温的垂直分布都产生影响,直接影响降水时空分布的模拟;不同方案都模拟出了在降水发生之后不同于晴空日变化的湍流动能垂直分布,经分析发现与局地较强的垂直风切变和近地面强湍流气团被抬升有关,而浮力项起着耗散作用;各方案的湍流交换特征与湍流动能特征基本吻合,相比于其他方案,MYJ方案在降水区域的湍流动能及湍流交换强度明显偏弱,对热通量的输送也偏弱;GBM方案在边界层内的湍流混合偏弱而在边界层以上湍流混合显著偏强,热通量输送在边界层以上的高度上误差明显,影响了对降水区域气象要素的模拟能力,仍需要进一步改进。      

不同边界层方案对一次华北暴雨数值模拟的敏感性研究

以2010年8月17—19日华北地区的暴雨事件为研究对象,采用中尺度数值模式WRF,通过8个数值实验测试暴雨事件模拟结果对不同边界层方案的敏感性。结果表明,采用不同边界层方案的模拟结果存在明显差异。分辨率为12 km时,7个边界层方案实验均模拟出类似观测的四处较强降水,但模拟得到的降水强度和强降水中心位置与实况有所差异,NOPBL实验模拟的雨带收缩,降水相对其他模拟实验较少且更为集中。分辨率为4 km时,采用边界层方案的7个实验均可见小尺度降水结构,模拟出较多虚假降水中心,而在NOPBL试验中降水的小尺度结构不明显。检验表明:分辨率为12 km时,MYJ试验的TS评分、相关系数和误差分析等整体表现最优,分辨率为4 km时,Bou Lac试验整体表现最优。与NOPBL试验相比,加入MYJ边界层方案后,模拟的水汽输送增加、上升运动及涡散度绝对值增大。     

Impacts of Shallow Convection Processes on a Simulated Boreal Summer Climatology in a Global Atmospheric Model

This study investigates the impacts of shallow convection schemes on a simulated seasonal climatology in the Global and Regional Integrated Model system (GRIMs). The eddy-diffusivity scheme of Tiedtke (TDK) is evaluated, focusing on the dependency upon deep convection schemes. Drying and warming near the top of the planetary boundary layer (PBL) and opposing effects above are observed. The height of PBL is reduced due to the increase of thermal stability near the PBL top. The weakened PBL turbulence is partly compensated with the increased downward solar radiation due to the reduction of low clouds. These effects are pronounced over the oceans, which leads to the modulation of tropical precipitation. It is found that the original TDK scheme shows similar behavior regardless of the choice of deep convection schemes. A revised TDK scheme that explicitly couples the PBL and shallow convection processes is proposed and evaluated. The proposed scheme generally improves the simulated climatology over the results with the original TDK scheme, along with further improvement in the case of the revised deep convection scheme. Our results indicate that the role of the shallow convection scheme needs to be carefully examined to improve the performance of atmospheric models, with a focus on modulated PBL and deep convection processes.     

中国夏季不同强度降水模拟对不同积云对流参数化方案的敏感性研究

利用PσRCM9区域气候模式, 分析了中国夏季不同强度降水模拟对不同积云对流参数化方案的敏感性。结果表明, 采用四种积云对流参数化方案, 模式能够模拟出小雨、 大雨和暴雨的雨量百分比和雨日百分比空间分布的一致性特征, 但不能模拟出中雨雨量百分比和雨日百分比空间分布的相似性, 这是由于模式不能模拟中雨雨量百分比的空间分布形式所致。还发现模拟的我国夏季降水以小雨和中雨为主, 四种积云对流参数化方案均低估了中国夏季大雨和暴雨对总降水的贡献, 尤其是在我国西部、 东北和华北地区更明显。不同积云对流参数化方案下模拟的极端强降水阈值的空间分布形式基本与观测一致, 但强度与观测存在较大差异。相比较而言, Grell方案较Kuo、 Anthes-Kuo和Betts-Mille积云对流参数化方案更适合中国东南部地区夏季极端强降水的模拟。     

基于不用分辨率和参数化方案的新疆区域数值模式性能初步评估

基于中尺度数值模式WRF,选取新疆两次强降水过程,设计三个试验方案,其中试验1为控制试验,试验2提高分辨率,试验3提高分辨率并调整物理参数化方案,初步评估不同分辨率和参数化方案对新疆区域2米温度、10米风速、降水预报的影响。结果表明：（1）提高分辨率对2米温度、10米风速模拟精度均有提高,2米温度预报精度提高约0.5℃,降低了日间温度模拟冷偏差;10米风速预报精度提高约0.5 m/s,降低了风速模拟正偏差;但提高分辨率后,模式出现虚假降水预报的情况。（2）提高分辨率并调整物理参数化方案后,2米温度模拟误差略有减小,模拟偏差减小约0.2℃;10米风速模拟误差增大约0.5 m/s,模拟偏差增大超过0.5 m/s;对降水落区、量级的模拟精度显著提高,减小了降水中心的模拟强度,对虚假降水预报有一定修正。