华中区域LRUC系统的构建与试验

如何将区域观测资料在中尺度模式中快速有效同化是提高区域精细化数值预报准确率和时效性的关键所在。本文简单介绍了基于LAPS和WRF建立的快速循环同化预报系统LRUC,以一次影响长江中游的典型梅雨锋暴雨过程为例,设计几种试验方案,结合实况降水分布,对比分析华中区域多普勒雷达资料循环同化对改善暴雨预报的作用;根据个例试验结果,设计批量后报试验方案,对比分析雷达资料同化对华中区域2007年主汛期(6—8月)多个时效降水预报评分。暴雨个例试验结果表明,华中区域多普勒雷达资料在LAPS中的同化,能为WRF提供更优初值场;同化华中区域地面资料和雷达资料后,比仅同化地面资料更能改进模式初值,改善模式降水预报,且降雨量级越大,24 h降水预报TS评分越高。批量后报试验结果表明,建立的快速循环同化预报系统LRUC,能对区域多源观测资料进行多时次循环同化分析,为模式初值增加观测资料的时间演变信息,提高模式对暴雨量级降水的预报水平;系统具有稳定性和一定的评分水平。     

GRAPES区域集合预报系统应用研究

为发展GRAPES(Global and Regional Assimilation and Prediction System)区域集合预报系统(GRAPES Regional Ensemble Prediction System,GRAPES-REPS),采用集合变换卡尔曼滤波(ETKF)初值扰动方法以及多物理过程组合的模式扰动方法,基于业务区域模式GRAPES_MesoV3.3.2.4构建了区域集合预报系统,进行了连续40 d的批量试验,重点分析了ETKF初值扰动的结构及其演变特征,并通过概率预报检验方法对GRAPES-REPS进行了集合预报系统性能检验和降水预报检验,分析了该系统对强降水个例的预报效果。试验结果表明,GRAPES-REPS能产生较合理的集合预报初值扰动,扰动结构随流型依赖并对观测有较好的响应,且扰动成员相互正交。扰动总能量分析表明集合扰动能够随预报时效保持合理增长状态。集合预报检验表明集合预报结果优于控制预报,集合成员间在72 h预报时效内能保持合理的集合离散度。将该区域集合预报系统与业务上基于WRF模式的区域集合预报系统WRF-REPS进行了降水预报对比,表明GRAPES-REPS的降水预报能力表现要优于业务WRF-REPS。强降水个例分析表明集合预报能较好预报出强降水中心,预报效果明显优于控制预报。     

对流尺度集合预报成员数对降水预报的影响北大核心CSCD

选取2022年川渝地区发生的16个强降水个例开展对流尺度集合预报批量试验,并通过对31组初值采用不同集合成员数时的降水集合预报技巧进行检验评估和综合分析。结果表明:集合成员的降水预报技巧总体上大致相当,因而采用不同成员数时预报技巧差异也不明显;表征降水总体分布特征的Talagrand分布和预报失误概率以及表征降水概率预报技巧的相对作用特征面积随着成员数的增加而逐渐改进,但当成员数达到一定数值后继续增大成员数对预报改进不明显。总体而言,对流尺度集合预报成员数设置为16~18最适宜。     

NUDGING方法同化雷达反射率因子个例试验研究

利用ECMEF全球高分辨率模式资料,采用松弛逼近方法(Nudging)同化多普勒雷达反射率因子,通过在华南区域GRAPES_3km模式中对海南岛降水过程个例进行数值模拟,分析不同Nudging方法对模式预报效果的影响。结果表明:(1)采用Nudging同化方法可使背景场和观测反演的资料相协调,缩短模式热启动时间,可增加开始时段的降水量;(2)同化雷达反射率因子能够提高海南岛0～6h逐时降水评分,模拟值更接近实况观测;(3)同化后,华南区域GRAPES_3km模式能够捕捉到更多的中小尺度信息,可改善海南岛24h强降水的量级;(4)在初始场Nudging订正云水、雨水、水汽场等,其水凝物分布合理,降水预报效果最佳。     

潜热加热纳近方法在地面降水资料同化中的应用

基于国家气象中心GRAPES_Meso高分辨率区域模式,针对中尺度数值预报模式中预报雨带形成滞后问题,研究了潜热加热纳近方法在地面降水资料同化中的应用,以期提高短时数值天气预报的水平。2013年6月20日—7月20日的初步试验结果表明:通过调整模式潜热加热廓线,可以改进初始场中温、湿、风等要素的合理分布,增加降水区的对流不稳定性;潜热加热纳近方法可以缩短模式的调整适应 (spin-up) 时间,改进短时降水预报的落区和强度,提高3 h,6 h,12 h的降水预报TS,ETS评分;与传统的冷潜热加热纳近的试验结果相比,改进的暖潜热加热纳近试验对降水落区和强度的预报更接近观测,但强降水中心范围略大。     

基于T639集合预报的持续性强降水中期客观预报技术研究

针对持续性强降水预报困难的问题,根据Anderson-Darling检验原理,构建基于中国气象局T639集合预报系统的持续性强降水中期客观预报方法。对比分析2010-2015年5-9月T639集合预报降水与实况降水的累积概率分布函数差异,在此基础上采用扩展时间序列和空间范围的方法构建3种模式气候累积概率方案,通过批量预报试验和检验,选取最优概率方案纳入预报模型,考察持续性强降水个例的最长预报时间。结果表明,随着预报时效的延长,集合预报模式的降水逐渐集中于小和中雨量级,无降水和暴雨以上量级的降水概率低于观测,168 h以后模式降水概率趋于稳定。通过扩展时间序列和空间范围能弥补模式气候资料年限不足所带来的偏差,根据区域气候特征细分模式气候的方法重点突出了不同区域的降水特征,明显优于简单集合所有区域数据的模式气候方案。基于集合预报的持续性强降水预报模型对持续性强降水个例的预报能力为8-9 d,随着预报时效的延长,降水强度以及雨带位置的预报能力逐渐减弱。      

多普勒雷达风廓线的反演及变分同化试验

为了将雷达风场资料更好地应用到数值预报模式中, 使用VAD方法反演多普勒雷达风廓线并处理成标准的探空资料进行变分同化试验。结果表明: VAD方法反演的风廓线与探空实况对应较好, 验证了用VAD技术反演风廓线的可行性。用GRAPES-Meso模式的三维变分同化系统对雷达风廓线资料进行同化后, 风场的初始场明显改善, 降水强度和落区预报也有不同程度的改善。其中, 对6 h降水预报的改善明显优于对24 h的预报改善。另外, 在短时强降水预报中, 雷达风场资料的同化频率和同化窗口的不同, 对降水预报的改善情况也有所差异。在个例研究中, 同化间隔为1 h的方案6 h降水预报要优于同化间隔为3 h和6 h的方案, 同化窗口为3 h的试验方案6 h降水预报要好于同化窗口为6 h的试验方案。     

风廓线雷达资料质量控制及其同化应用

为更好地同化风廓线雷达观测资料开展了相应的质量控制与同化应用研究。针对2013年5月广东地区13部风廓线雷达的观测数据,采用经验正交函数(EOF) 分析方法对其进行质量控制。相比原始观测,经过质量控制的风场提高(降低)了来自时空大(小)尺度的贡献,较好地滤除了小尺度高频脉动,也较好地保留了大尺度平均状态与局地中小尺度系统的共同影响,并且更加接近ECMWF再分析场。此外,还对质量控制后的数据进行了垂直稀疏化。分别计算了质量控制前、后风廓线雷达观测与NCEP 6 h预报场的差值,对比差值的特征发现,经过质量控制的数据的观测增量更好地满足了高斯分布与无偏假设。针对一个实际天气个例,基于GRAPES 3D-Var同化系统,分析了质量控制后的风廓线雷达资料对模式分析与预报的影响。试验表明,在循环同化过程中加入风廓线雷达资料可以更好地描述模式初始场低层风场的特征,从而对强降水的位置与强度做出更好的预报。针对2013年5月的批量试验表明,同化风廓线雷达资料使短期降水预报有明显的改善。     

C波段雷达资料同化对GRAPES中尺度数值模式短临降水预报的影响

将C波段雷达资料用LAPS模式的云分析系统进行反演,并采用Nudging技术将反演得到的云微物理场引入GRAPES中尺度数值模式,结合1次强降水天气过程的模拟实验,研究了C波段雷达资料同化对GRAPES中尺度数值模式短临降水预报的影响。结果表明：①雷达资料同化能够改进中尺度模式的降水预报,模式前6 h的降水预报相关系数和不同等级降水TS评分都有提高,同时降水峰值提前了1 h,有助于缓解模式spin-up问题。②模式降水预报的改进效果主要由强降水贡献,最大改进效果集中在4~6 h。③同化雷达资料后25 mm以上强降水预报站数比更接近实况,落区偏差幅度更小,降水落区和强度向实况方向得到调整。④10 mm以下弱降水在吸收雷达资料后,站数比相较于控制预报,比实况增加更多,落区偏差幅度增大,存在预报过量的问题。弱降水预报过量主要集中在4~6 h,而前3 h对降水预报的改进有积极作用。     

暴雨数值预报中的初值试验

本文介绍使用五层原始方程有限区域降水预报模式进行汛期暴雨数值预报试验时,引进的新的初值方案。给出了用新初值方案计算的个例和1986年汛期实际预报试验的结果。与其它方案计算结果的比较表明,认真改进初值有助于改进暴雨预报,新的初值方案可以满足暴雨预报业务试验的要求。     

Application of Radar Reflectivity Factor in Initializing Cloud-Resolving Mesoscale Model. Part Ⅱ： Numerical Simulation Experiments

Microphysics elements and vertical velocity retrieved were incorporated using the nudging method into the initial data assimilation of GRAPES （Global/Regional Assimilation and Prediction System） model. Simulation experiments indicated that nudging technique was effective in forcing the model forecast gradually consistent to the observations, yielding the thermodynamically and dynamically balanced analysis field. As viewed from the simulation results, water vapor is vital to precipitation, and it is a governing factor for the amount and duration of precipitation. The initial cloud water, rain water, and vertical velocity determine the strength distribution of convection and precipitation at the beginning time of forecast; the horizontal wind field steers the motion of the mesoscale weather system embedded in and impacts the position of precipitation zone to a large extent. The simulation experiments show that the influence of the initial retrieval data on prediction weakens with the increase of forecast time, and within the first hour of forecast, the retrieval data have an important impact on the evolution of the weather system, but its influence becomes trivial after the first three hours. Changing the nudging coefficient and the integral time-spacing of numerical model will bring some influences to the results. Herein only one radar reflectivity was used, the radar observations did not cover the whole model domain, and some empirical parameters were used in the retrieval method, therefore some differences still lie between simulation and observation to a certain extent, and further studies on several aspects are expected.     

Application of Radar Reflectivity Factor in Initializing Cloud-Resolving Mesoscale Model. Part Ⅱ: Numerical Simulation Experiments

Microphysics elements and vertical velocity retrieved were incorporated using the nudging method into the initial data assimilation of GRAPES (Global/Regional Assimilation and Prediction System) model.Simulation experiments indicated that nudging technique was effective in forcing the model forecast gradually consistent to the observations, yielding the thermodynamically and dynamically balanced analysis field. As viewed from the simulation results, water vapor is vital to precipitation, and it is a governing factor for the amount and duration of precipitation. The initial cloud water, rain water, and vertical velocity determine the strength distribution of convection and precipitation at the beginning time of forecast; the horizontal wind field steers the motion of the mesoscale weather system embedded in and impacts the position of precipitation zone to a large extent. The simulation experiments show that the influence of the initial retrieval data on prediction weakens with the increase of forecast time, and within the first hour of forecast, the retrieval data have an important impact on the evolution of the weather system, but its influence becomes trivial after the first three hours. Changing the nudging coefficient and the integral time-spacing of numerical model will bring some influences to the results. Herein only one radar reflectivity was used, the radar observations did not cover the whole model domain, and some empirical parameters were used in the retrieval method, therefore some differences still lie between simulation and observation to a certain extent, and further studies on several aspects are expected.     

基于贝叶斯方案的雷达反射率反演水汽及其同化试验

雷达资料是目前为数不多有能力为高分辨率预报模式提供高分辨率信息资料的资料之一。为充分利用该资料所包含的中小尺度信息,文中基于雷达反射率,利用贝叶斯方法反演出大气相对湿度;将质控后的资料引入3Dvar系统进行同化分析,为高分辨率模式提供初值场。以台风“妮妲”登陆为例,通过一维反演及三维变分系统分析,有效地订正了实况有回波而模式预报无回波区域的大气湿度趋于合理,增加背景场的湿度,减小模拟回波比观测偏强的区域的大气湿度;同化大气湿度后模式在前6小时报出的台风外围回波分布、演变更合理,改进了降水雨带的分布与强度。1个月的批量试验反映1D+3Dvar同化雷达资料后,大气对流层中低层（850~400 hPa）增湿明显,其增湿影响程度可延续12小时以上。其逐时降水预报在前12小时的TS均比控制试验高,而大于5 mm以上降水预报偏差则与控制试验的大略一致或更接近1。      

Short-range prediction of a heavy precipitation event by assimilating Chinese CINRAD-SA radar reflectivity data using complex cloud analysis

Summary With the ARPS (Advanced Regional Prediction System) Data Analysis System (ADAS) and its complex cloud analysis scheme, the reflectivity data from a Chinese CINRAD-SA Doppler radar are used to analyze 3D cloud and hydrometeor fields and in-cloud temperature and moisture. Forecast experiments starting from such initial conditions are performed for a northern China heavy rainfall event to examine the impact of the reflectivity data and other conventional observations on short-range precipitation forecast. The full 3D cloud analysis mitigates the commonly known spin-up problem with precipitation forecast, resulting a significant improvement in precipitation forecast in the first 4 to 5 hours. In such a case, the position, timing and amount of precipitation are all accurately predicted. When the cloud analysis is used without in-cloud temperature adjustment, only the forecast of light precipitation within the first hour is improved. Additional analysis of surface and upper-air observations on the native ARPS grid, using the 1 degree real-time NCEP AVN analysis as the background, helps improve the location and intensity of rainfall forecasting slightly. Hourly accumulated rainfall estimated from radar reflectivity data is found to be less accurate than the model predicted precipitation when full cloud analysis is used.     

Numerical Study of the Role of Microphysical Latent Heating and Surface Heat Fluxes in a Severe Precipitation Event in the Warm Sector over Southern China

Simulations of the severe precipitation event that occurred in the warm sector over southern China on 08 May 2014 are conducted using the Advanced Weather Research and Forecasting (WRF-ARWv3.5.1) model to investigate the roles of microphysical latent heating and surface heat fluxes during the severe precipitation processes. At first, observations from surface rain gauges and ground-based weather radars are used to evaluate the model outputs. Results show that the spatial distribution of 24-h accumulated precipitation is well reproduced, and the temporal and spatial distributions of the simulated radar reflectivity agree well with the observations. Then, several sensitive simulations are performed with the identical model configurations, except for different options in microphysical latent heating and surface heat fluxes. From the results, one of the significant findings is that the latent heating from warm rain microphysical processes heats the atmosphere in the initial phase of the precipitation and thus convective systems start by self-triggering and self-organizing, despite the fact that the environmental conditions are not favorable to the occurrence of precipitation event at the initial phase. In the case of the severe precipitation event over the warm sector, both warm and ice microphysical processes are active with the ice microphysics processes activated almost two hours later. According to the sensitive results, there is a very weak precipitation without heavy rainfall belt when microphysical latent heating is turned off. In terms of this precipitation event, the warm microphysics processes play significant roles on precipitation intensity, while the ice microphysics processes have effects on the spatial distribution of precipitation. Both surface sensible and latent heating have effects on the precipitation intensity and spatial distribution. By comparison, the surface sensible heating has a strong influence on the spatial distribution of precipitation, and the surface latent heating has only a slight impact on the precipitation intensity. The results indicate that microphysical latent heating might be an important factor for severe precipitation forecast in the warm sector over southern China. Surface sensible heating can have considerable influence on the precipitation spatial distribution and should not be neglected in the case of weak large-scale conditions with abundant water vapor in the warm sector.     

ON USE OF LHN METHOD TO ASSIMILATE THE INTENSIFIED SURFACE PRECIPITATIONS FOR GRAPES_MESO MODEL INITIALIZATION

The quantitative precipitation forecast (QPF) in very-short range (0-12 hours) has been investigated in this paper by using a convective-scale (3km) GRAPES_Meso model. At first, a latent heat nudging (LHN) scheme to assimilate the hourly intensified surface precipitation data was set up to enhance the initialization of GRAPES_Meso integration. And then based on the LHN scheme, a convective-scale prediction system was built up in considering the initial “triggering” uncertainties by means of multi-scale initial analysis (MSIA), such as the three-dimensional variational data assimilation (3DVAR), the traditional LHN method (VAR0LHN3), the cycling LHN method (CYCLING), the spatial filtering (SS) and the temporal filtering (DFI) LHN methods. Furthermore, the probability matching (PM) method was used to generate the QPF in very-short range by combining the precipitation forecasts of the five runs. The experiments for one month were carried out to validate the MSIA and PM method for QPF in very-short range. The numerical simulation results showed that: (1) in comparison with the control run, the CYCLING run could generate the smaller-scale initial moist increments and was better for reducing the spin-up time and triggering the convection in a very-short time; (2) the DFI runs could generate the initial analysis fields with relatively larger-scale initial increments and trigger the weaker convections at the beginning time (0-3h) of integration, but enhance them at latter time (6-12h); (3) by combining the five runs with different convection triggering features, the PM method could significantly improve the QPF in very-short range in comparison to any single run. Therefore, the QPF with a small size of combining members proposed here is quite prospective in operation for its lower computation cost and better performance.     

GRAPES三维云初始场形成及在短临预报中的应用

围绕GRAPES_Meso的云初始场形成,以ARPS模式云分析方案为基础,优化诊断后应用我国风云二号静止气象卫星云产品、多普勒天气雷达三维组网拼图产品等观测资料结合模式背景信息,根据云热力-动力学原理及观测试验经验关系等,对云初始场的信息进行分析并通过松弛逼近同化方法实现对云内微物理信息同化应用。GRAPES_Meso中采用优化后的云初始场方案,水平分辨率为0.03°×0.03°和0.1°×0.1°的1个月（2014年7月15日-8月14日）连续试验和个例分析结果显示:云初始场形成方案能够分析出飑线等天气系统的云系和云内微物理变量特征。从模拟云图看,包含云初始场信息的GRAPES_Meso的云系的形态特征和分布范围短时临近预报结果更为准确。云初始场信息同化应用后,在1 h的时间尺度上,即可预报出与实况更为接近的降水;0~12 h时间范围内对降水均有积极的影响,可满足短时临近预报的需求,降水量级略偏大。批量连续试验（水平分辨率为0.03°×0.03°和0.1°×0.1°）的各个量级降水ETS（equitable threat score）评分都显著提高。     

C波段雷达反射率因子在CMA-GD模式中的应用

使用CMA-GD模式及云分析系统,引入云南C波段多普勒雷达反射率因子资料,对2019年7月9日过程进行模拟试验,分析引入反射率资料对模式初始场和降水过程预报的影响。（1）引入反射率后,云中和底部的云量有所增加。水汽在900~200 hPa有大范围增加,能有效地调整降水区域的水汽分布。对模式顶层温度的调整较大,而对风场的影响较小。（2）引入反射率后,对3 h内降水强度及落区有较大改善,4~6 h的预报有所改善,7 h以后改善不明显。（3）引入反射率资料后,1~4 h大气可降水量增量较明显,5~9 h增量较前4 h明显减小。（4）在河口上空云水和水汽在950~400 hPa增加,霰、云冰和云雪在600~400 hPa增加,雨滴在1 000~500 hPa增加。水凝物增加,有利于河口站降水的发生。      

雷达反射率因子在中尺度云分辨模式初始化中的应用Ⅱ：数值模拟试验

将反演的云微物理量和垂直速度采用"牛顿连续松弛逼近"(Nudging)技术应用到GRAPES模式中,对一次梅雨锋暴雨过程进行了数值模拟。通过设计不同的试验方案,分别对水汽、液态水和垂直速度对Nudging效果以及预报结果的影响进行了考察。研究发现,采用Nudging初始化方法,可使背景场与观测反演资料相协调,实现了模式中对流的热启动,模式预报的开始时刻就产生降水,缩短Spin-up时间。水汽对降水至关重要,对降水的强度和持续时间都有重要影响;云水、雨水和垂直速度决定了初始时刻对流的强弱分布并产生降水;水平风场决定了系统的移动方向,对预报降水的落区有重要影响。模式比较成功地模拟了6 h的降水过程,中尺度天气系统的时空特征比较明显,对流中心上升速度最大值约2.0 m/s,云水含量400 hPa以上较大,最大值约1.5 g/kg,雨水含量500 hPa以下较大,最大约3.0 g/kg,底层辐合高层辐散。反演资料对降水的影响随预报时间的增加而减弱,预报1 h之内反演资料有明显影响,3 h之后的预报则主要依赖模式自身。鉴于仅使用一部雷达资料的反射率因子资料,雷达资料没有覆盖整个模式区域,天气系统被截断,反演和同化过程还采用了一些经验参数方法等原因,数值模拟结果与雷达观测之间还存在一定的差异,有待于更深入的研究。