Impact of ECMWF,NCEP, and NCMRWF global model analysis on the WRF model forecast over Indian Region

The global model analysis has significant impact on the mesoscale model forecast as global model provides initial condition (IC) and lateral boundary conditions (LBC) for the mesoscale model. With this objective, four operational global model analyses prepared from the European Centre for Medium-Range Weather Forecasts (ECMWF), National Centers for Environmental Prediction (NCEP) Global Data Assimilation System (GDAS), NCEP Global Forecasting System (GFS), and National Centre for Medium Range Weather Forecasting (NCMRWF) are used daily to generate IC and LBC of the mesoscale model during 13th December 2012 to 13th January 2013. The Weather Research and Forecasting (WRF) model version 3.4, broadly used for short-range weather forecast, is adopted in this study as mesoscale model. After initial comparison of global model analyses with Atmospheric Infrared Sounder (AIRS) retrieved temperature and moisture profiles, daily WRF model forecasts initialized from global model analyses are compared with in situ observations and AIRS profiles. Results demonstrated that forecasts initialized from the ECMWF analysis are closer to AIRS-retrieved profiles and in situ observations compared to other global model analyses. No major differences are occurred in the WRF model forecasts when initialized from the NCEP GDAS and GFS analyses, whereas these two analyses have different spatial resolutions and observations used for assimilation. Maximum RMSD is seen in the NCMRWF analysis-based experiments when compared with AIRS-retrieved profiles. The rainfall prediction is also improved when WRF model is initialized from the ECMWF analysis compared to the NCEP and NCMRWF analyses.     

SWCWARMS 模式对巴中地区预报能力的检验

应用巴中地区218个自动气象观测站2016年6～8月逐日的24h累积雨量,采取基于客观统计的检验方法,针对24h累积雨量,采取TS评分、BIAS评分和空报率对比分析了 SWCWARMS 模式和 EC 细网格模式在巴中地区的预报能力。研究结果表明:(1)从降水落区来看,两个模式都对雨带的移动都有较好的预报能力;(2)从降水 TS 评分看,SWCWARMS模式在各量级上的 TS 评分都优于 EC 模式;(3)从降水 BIAS 评分看,EC 模式容易漏报中雨及以上量级的降水,而 SWCWARMS 模式则对大暴雨的预报有较大的空报;(4)从逐月评估结果看,6月和7月SWCWARMS 模式较EC模式有更好的预报效果,8月两者差异不大。总体而言,SWCWARMS 模式在巴中地区的预报效果优于 EC 模式。      

一种粒子——烟团扩散模型的应用研究

本文运用随机游动模拟方法改进建立一种粒子—烟团模式，成功地模拟了对流边界层条件下污染物扩散，粒子—烟团模式的模拟结果与水槽试验的结果吻合较好。数值模拟试验结果表明粒子—烟团模式能在施放粒子数较少，从而运行时间很短的情况下，得到优于一般随机游动模式的模拟性能。用KNRC的资料对模式的模拟性能作一验证，表明模式也能很好地模拟实际大气中不同稳定度情况下的扩散。     

复杂陆面模式与中期数值天气预报模式的耦合

将复杂陆面模式(CLM3)耦合到中期数值预报模式YHMPT106中,替换原来的简单3层陆面过程参数化方案.对比分析表明,较为复杂的陆面过程模式CLM3与中期预报业务模式YHMPT106的耦合是成功的,耦合模式可稳定运行,其预报效果也好.这个试验亦为将来气候预测模式中使用精细的大气环流模式提供了有用的经验.     

Simulation of three-dimensional wind flow over complex terrain in the atmospheric boundary layer

A three-dimensional model for wind prediction over rough terrain has been developed for practical use. It is a compromise between hydrodynamic and objective wind models. The proposed model includes: (1) a statistical model to predict the wind velocity and potential temperature at anemometer height at observing stations, (2) the drainage wind model expressed by Prandtl's analytic solution for the slope wind, (3) the Businger-Dyer surface-layer formulation which considers the surface energy budget and (4) the model for three-dimensional boundary-layer solutions to the stationary flow. In this model, mass consistency is guaranteed by using flow fields that satisfy the continuity equation. Model predictions show good agreement with the observations.     

基于MET系统对SWCWARMS及GRAPES模式在西南地区的站点检验分析

应用国家基本观测站资料、自动站逐时降水资料,基于客观统计检验方法,针对降水(12h、24h累积雨量)、近地面要素(2m温度、10m风)和高空要素(风场、温度场、高度场),分别评估SWCWARMS模式和GRAPES模式对2015年西南地区预报能力,得到如下几点结论:(1)SWCWARMS模式降水ETS评分高于GRAPES模式,除24h小雨外SWCWARMS模式偏差值均高于GRAPES模式,两个模式在不同预报时效内对中雨、大雨、暴雨都表现一定程度的空报;(2)12h降水分段评分上,SWCWARMS模式TS评分均高于GRAPES模式,但SWCWARMS模式预报降水范围过大,随着预报时效增长空报多于GRAPES模式;SWCWARMS模式中雨和大雨空报大于其它量级降水,GRAPES模式对大暴雨漏报较多其它量级降水表现为空报;(3)两模式对高度场和温度场预报优于风场,对对流层中层预报优于中低层,SWCWARMS模式对高度场和温度场预报优于GRAPES模式,夏半年SWCWARMS模式均方根误差小于GRAPES模式;(4)两模式都表现出2m温度均方根误差在秋季增加而春季减小这一特征,SWCWARMS模式近地面要素均方根误差均小于GRAPES模式。      

Ekman动量近似下中间边界层模式中的风场结构

发展了一个准三维的、中等复杂的边界层动力学模式，该模式包含了EKman动量近似下的惯性加速度和Blackadar的非线性湍流粘性系数，它进一步改进了Tan和Wu(1993）提出的边界层理论模型。该模型在数值计算复杂性上与经典Ekman模式相类似，但由于包含了Ekman动量近似下的惯性项，使得该模式比传统Ekman模式更近于实际过程。中详细地比较了该模式与其他简化边界层模式在动力学上的差异，结果表明：在经典的Ekman模式中，由于忽略了流动的惯性项作用，导致在气旋性切变气流（反气旋性切变气流）中风速和边界层顶部的垂直速度的高估（低估），而在半地转边界层模式中，由于高估了流动惯性项的作用，结果与经典Ekman模式相反。同样，该模式可以应用于斜压边界层，对于Ekman动量下的斜压边界层风场同时具有经典斜压边界层和Ekman动量近似边界层的特征。     

Long Range Forecasting Indian Summer Monsoon Rainfallby a Hybrid Principal Component Neural Network Model

Summary  The existing methods based on statistical techniques for long range forecasts of Indian monsoon rainfall have shown reasonably accurate performance, for last 11 years. Because of the limitation of such statistical techniques, new techniques may have to be tried to obtain better results. In this paper, we discuss the results of an artificial neural network model by combining two different neural networks, one explaining assumed deterministic dynamics within the time series of Indian monsoon rainfall (Model I) and other using eight regional and global predictors (Model II). The model I has been developed by using the data of past 50 years (1901–50) and the data for recent period (1951–97) has been used for verification. The model II has been developed by using the 30 year (1958–87) data and the verification of this model has been carried out using the independent data of 10 year period (1988–97). In model II, instead of using eight parameters directly as inputs, we have carried out Principal Component Analysis (PCA) of the eight parameters with 30 years of data, 1958–87, and the first five principal components are included as input parameters. By combining model I and model II, a hybrid principal component neural network model (Model III) has been developed by using 30 year (1958–87) data as training period and recent 10 year period (1988–97) as verification period. Performance of the hybrid model (Model III) has been found the best among all three models developed. Rootmean square error (RMSE) of this hybrid model during the independent period (1988–97) is 4.93% as against 6.83%of the operational forecasts of the India Meteorological Department (IMD) using the 16 parameter Power Regression model. As this hybrid model is showing good results, it is now used by the IMD for experimental long-range forecasts of summer monsoon rainfall over India as a whole. Received August 20, 1998/Revised April 20, 1999     

基于DERF2.0的月平均温度概率订正预报

国家气候中心第二代月动力延伸模式回算资料的分析表明,二代模式月平均温度预报与观测实况仍然存在较大偏差,模式预报有较大改进空间。本文采用非参数百分位映射法对模式月平均温度预报进行概率订正,该方法基于模式集合平均给出的确定性预报,结合模式回算资料各集合成员计算得到的模式概率密度分布,给出确定性预报在模式概率密度分布中的百分位值,并将百分位值投影到观测资料的概率密度分布中,得到模式预报的概率订正值。对订正前后模式预报的检验评估显示,该订正方案不仅有效降低了模式预报与实况的均方根误差（RMSE）,对月平均温度距平分布的预报技巧也有所改善,不同超前时间模式预报的预测技巧评分（PS）和距平相关系数（ACC）均有提升,同时模式预报误差的大小对订正效果无明显影响。从分月的订正预报结果来看,对夏季各月的温度预测技巧的提升整体高于冬季各月。     

三种数值模式对四川省汛期降水预报性能的检验

应用国家基本观测站资料,基于MET系统的客观统计检验方法,针对24h降水分别评估SWCWARMS模式、GRAPES模式和ECMWF模式对2017～2019年5～10月四川地区汛期预报能力,得到如下几点结论:(1)SWCWARMS模式小到大暴雨降水范围大于实况,GRAPES模式小到暴雨降水范围大于实况、大暴雨多漏报,ECMWF模式小雨和中雨降水范围大于实况、大到大暴雨多漏报,三个模式无降水或微量降水均少于实况。(2)ECMWF模式对四川雨季小到大雨预报能力优于SWCWARMS和GRAPES模式,SWCWARMS模式在部分时次上暴雨和大暴雨预报优于ECMWF模式,GRAPES模式TS评分略偏低。(3)GRAPES模式在2018年秋季开始中雨及以上量级降水预报上改善大于SWCWARMS和ECMWF模式,SWCWARMS模式2019年空报较2017年和2018年显著降低;3个模式在小雨和中雨预报上不相上下,GRAPES模式优势在2019年大雨和暴雨预报上,ECMWF模式优势在2017年秋季和2018年初夏大雨预报上,SWCWARMS模式大雨和暴雨预报能力介于二者之间。(4)ECMWF和SWCWARMS模式川东预报优于川西,GRAPES模式川西预报优于川东;三个模式存在不同程度空报,川东地区空报略多于川西,其中ECMWF模式空报最多。      

New Models for Long Range Forecasts of Summer Monsoon Rainfall over North West and Peninsular India

Summary New models based on (a) Multivariate Principal Component Regression (PCR) (b) Neural Network (NN) and (c) Linear Discriminant Analysis (LDA) techniques were developed for long-range forecasts of summer monsoon (June–September) rainfall over two homogeneous regions of India, viz., North West India and Peninsular India. The PCR and NN models were developed with two different data sets. One set consisted 42 years (1958–1999) of data with 8 predictors and the other, 49 years (1951–1999) of data with 6 predictors. The predictors were subjected to the Principal Component Analysis (PCA) before model development. Two different neural networks were designed with 2 and 3 hidden neurons. To avoid the nonlinear instability, 20 ensemble runs were made while training the network and the ensemble mean results are discussed. The LDA model was developed with 42 years of data (1958–1999) for classifying three rainfall intervals with equal prior probability of 0.33. Both the PCR and NN models showed useful forecast skill for NW India and Peninsular India. Models with 8 predictors performed better than the models with only 6 predictors. The NN model with 3 hidden neurons performed better than model with 2 hidden neurons. For NW India, the NN model performed better than the PCR model. The RMSE of the NN model and PCR model with 8 predictors for NW India (Peninsular India) during the independent period 1984–99 was 12.5% (12.2%) and 12.6% (11.5%), respectively. Corresponding figures for the models with 6 predictors are 15.0% (13.0%) and 13.9% (11.4%) respectively. During the independent period, model errors were large in 1991, 1994, 1997 and 1999. However all the models showed deteriorating predictive skill after 1988, both for NW India and Peninsular India. The LDA model correctly classified 62% of grouped cases for NW India and Peninsular India. The LDA model showed better skill in classifying deficient rainfall (< − 8%) over NW India and excess rainfall (> 3%) over Peninsular India. Received October 2, 1999 Revised December 28, 1999     

基于Catboost和Stacking融合模型的 长江中下游短时临近降水预报研究

用中国自动站与CMORPH降水产品融合的逐小时降水量网格数据集、全球预报系统(Global Forecasting System, GFS)模式再分析资料,将机器学习特征算法筛选的特征变量作为模型输入数据,运用Catboost模型和以Catboost和随机森林为初级模型、径向基神经网络为次级模型的融合模型预测未来6 h累计降水等级,并应用公平TS评分(Equal Threat Score,ETS)、真实技巧评分(True Skill Statistic,TSS)、混淆矩阵、预报偏差(Bias值)、击中率(Probability of Detection,POD)对预报结果进行检验分析。结果表明:优化变量的输入有利于提高模型的准确率;Catboost模型和融合模型都可以在一定程度上辨别晴雨状况;仅非动力学变量参与的融合模型对雨区预报准确率最高,但容易将暴雨雨区预报得更加广泛。总体而言,融合模型具有更强、更稳定的预报性能,中到暴雨量级预报准确率还待进一步提高。     

输出误差系统的多新息辨识方法

研究了输出误差(OE)系统和输出误差自回归滑动平均(OEARMA)系统(即Box-Jenkins系统)的辅助模型随机梯度算法、辅助模型多新息随机梯度算法、辅助模型递推最小二乘算法、辅助模型多新息最小二乘算法,输出误差系统的修正辅助模型随机梯度算法、遗忘因子辅助模型随机梯度算法、变递推间隔辅助模型随机梯度算法、变递推间隔辅助模型多新息随机梯度算法、变递推间隔辅助模型递推最小二乘算法等,以及输出误差自回归(OEAR)系统的基于滤波的辅助模型多新息广义随机梯度算法和基于滤波的辅助模型多新息广义最小二乘算法.     

A coupled dynamical-radiational model of stratocumulus

A model dealing with interactions between the air and low stratiform clouds is presented based on the mixed-Layer model Lilly (1968) pioneered and on Deardorff’s three dimensional numerical model results. Its main new aspects lie in 1) consideration of the natures of both the atmosphere and cloud; 2) a new entrainment velocity scheme with few arbitrary assumptions; 3) transition from one-mixed layer to two-mixed layer model; and 4) parameterization of radiation and precipitation calculations.The model results for radiation, moisture, and heat turbulent fluxes turn out to be in good agreement with those calculated or observed by Kawa (1988). Nicholls (1984), and Schmets et al. (1981) in California, the North Sea, and the North Atlantic, respectively.Basically, this paper furnishes the theoretical basis for a model to address questions concerning the time-evolu-tion of thermodynamical profiles both in cloud and out of cloud. The applications of this model will be in a separate paper.     

Intercomparison of three urban climate models

An intercomparison of the surface energy budgets from three urban climate models was made to assess the comparability of results, and to evaluate the surface energy fluxes from each model. The three models selected spanned the continuum of approaches currently employed in the treatment of the effects of urban geometry. The first model was an urban canopy-layer model which explicitly examined urban canyon geometry. The second model treated the city as a warm, rough, moist plate but included greatly simplified parameterizations of urban geometry. Neither model included a dynamic link to the urban boundary-layer. The third model was a one-dimensional urban boundary-layer model which utilized a simple warm, rough, moist plate approach but included a dynamic coupling of the urban surface layer to the urban boundary-layer.Results showed considerable disagreement between the three models in regards to the individual energy fluxes. Average rankings of the energy fluxes in terms of comparability from high-to-low similarity were: (1) solar radiation, (2) sensible heat flux, (3) conduction, (4) latent heat flux, (5) longwave re-radiation, and (6) longwave radiation input. In general, the urban canopy-layer model provided more realistic results, although each model demonstrated strong and weak points. Results indicate that current urban boundary-layer models may produce surface energy budgets with lower sensible heat fluxes and substantially higher latent heat fluxes than is supported by field evidence from the literature.     

人为热排放的引入对北京地区精细模拟的改进

将人为热排放纳入到已耦合城市模块Urban Canopy Model(UCM)的中尺度气象模式Weather Research and Forecasting(WRF)中,探讨了人为热排放对于北京地区精细化模拟的重要意义,其影响主要体现在以下几个方面:1)可有效改善气象要素的模拟效果,特别是对于大气边界层高度的显著性改善,该变量是控制空气质量模式中污染物垂直扩散的关键因子;2)可较好地再现城区流场及温度场,使热岛强度和中心配置更接近实况;3)可明显改善数值模式对于污染物垂直分布特征的模拟。     

非均匀地表条件下区域蒸发散通量计算方法的研究

文章改进了计算蒸发散通量的 Kotada-Barton模式（K-B模式)。经长江三角洲地区和淮河流域试验区各种不同陆面条件（水面，森林，麦田和水稻田）下的实际应用并通过和其它方法比较，认为该方法计算精度与涡动相关法及鲍恩比法相当，不论是计算蒸发散通量的瞬时值或日值都具有较高的可信度。该方法的最大优点在于它仅仅依赖于土地资源遥感信息、常规气象资料和基础地理信息如地面高程，而到目前为止上述资料均可以通过卫星等手段及常规观测得到。利用此方法计算了长江三角洲地区各种地表面上1995年各月的蒸发散量及该地区的区域平均蒸发散量。结果证明此方法具有较高的精确度，且基础资料容易得到，是一种估计不同地表覆盖条件下的蒸发散量和具有复杂的地形和土地利用条件下的区域平均蒸发散量的有用工具。     

利用人工神经网络模型预测西北太平洋热带气旋生成频数

通过对60年(1950~2009年)北半球夏、秋季(6~10月)热带气旋(TC)频数与春季(3~5月)大尺度环境变量的相关分析,挑选出8个相关性较高的前期预报因子建立人工神经网络(ANN)模型,对2010~2017年8年夏、秋季TC频数进行回报,并将回报结果与传统多元线性回归(MLR)方法所得结果进行对比分析。结果表明,ANN模型对60年历史数据的拟合精度高,相关系数高达0.99,平均绝对误差低至0.77。在8年回报中,ANN模型相关系数为0.80,平均绝对误差为1.97;而MLR模型相关系数仅为0.46,平均绝对误差为3.30。ANN模型在历史数据拟合和回报中的表现都明显优于MLR模型,未来可考虑应用于实际的业务预测中。     

Probabilistic comparison between turbulence closure model and Bulk Richardson Number approach for ABL height estimation using copula

In this paper, a numerical study has been carried out to assess changes in Atmospheric Boundary Layer (ABL) height under changing hydro-climatic conditions. ABL height is estimated by using the data from experiments conducted at three different sites- LAnd Surface Process EXperiment (LASPEX-97) in Sabarmati river basin, Gujarat, India, during 1997; Boundary Layer Late Afternoon and Sunset Turbulence (BLLAST) in Lannemezan, France, during 2011 and Integrated Global Radiosonde Archive (IGRA) in Tucson at Arizona, USA, during 2011. Here ABL height is estimated, during Intensive Observation Period (IOP), using 1-D turbulence closure model at all three site locations as well as following the Bulk Richardson Number (BRN) Approach. For LASPEX-97 only, the 1-D model with module for sub-surface processes coupled with upper-surface model is used for the estimation and compared with the only upper-surface model. ABL height, estimated by the coupled model, is found to be slightly higher than that by the un-coupled model. A copula-based analysis indicates that ABL height estimated by the 1-D model and same by BRN-based approach are statistically similar within 90% confidence interval (CI) under similar stratification condition. Thus, it validates the 1-D turbulence closure model for ABL height estimation. The sensitivity analysis of height of ABL, derived by 1-D model, against critical parameters reveals that surface condition (soil temperature) is more critical than atmospheric condition (net radiation) in forcing change on height of ABL in summer.