模式输出统计技术在局地中短期天气预报中的研究与应用
|
摘要
局地天气预报一直是天气工作者最为关注的问题,由于全球预报模式的分辨率较低,近地面边界层的物理过程,以及受下垫面影响的动力过程和热力过程在局地尺度的模拟及预报效果并不理想,使得全球预报模式对地面气象要素的预报准确性不能达到令人满意的效果。因此,为了得到更为准确的局地天气预报结果,需要对全球预报模式输出结果进行降尺度。统计-动力相结合的降尺度释用技术(即模式输出统计技术)不仅能弥补动力降尺度法的不足,而且它计算量小、易于操作,很适合基层气象台站的推广使用。它已成为很多发达国家现代天气预报业务的重要工具,特别是在数值模式的后处理中已成为不可缺少的一部分。本文在前人研究工作基础之上,根据实际需要,从科研和实际应用两个方面全面探索和研究模式输出统计中存在的一些问题并针对这些问题给出一套完整的解决方案。
全文以模式输出统计(MOS)技术的理念为核心,紧紧围绕这一中心思想,针对模式输出统计中的预报因子选择问题,预报模型建立问题,进行了深入的研究和试验,取得了很多令人鼓舞的结果。文中试验选取青海省内25个气象站和环渤海地区23个气象站作为研究站点,以2003年1月1日—2009年12月31日的国家气象中心T213L31全球中期数值预报模式逐日输出产品以及上述气象站点的地面观测资料作为试验资料,以最高温度、最低温度和降水三个人们最关心的要素作为试验对象。
本文利用相关系数和逐步回归方法进行统计降尺度的因子选择,因子选择的试验表明：预备因子数虽然从总体水平上对最终的命中率不太敏感,但对于个别站点而言,不同的预备因子数会对预报结果的好坏产生一定的影响;通过对F值的试验可以看出,基本上所有台站预报结果的命中率不受F值的影响,但F值不易过小,因为F值太小会造成预报结果的不太稳定。对于加入历史实况作为因子的试验来说,其冬季的预报效果,特别是短期(24,48小时)的预报效果得到了明显的改善,而对于夏季,其效果就不明显。
在以温度为代表的连续要素建模中,本文重点对神经网络进行了深入的研究,结果表明：神经网络的随机权重初始场的权重值,如果服从高斯分布将能使神经网络能更好的向最优解靠近,采用多次初始化初始权重的方法和动态隐层神经元数的方法能让神经网络更好的逼近真实值。通过对环渤海地区冬季最低温度和夏季最高温度的预报试验发现,神经网络与多元回归方法的预报效果平分秋色；在冬季和夏季的部分预报时效中BP-神经网络比逐步回归的评分有一定的提高,有些时效的预报效果并不如逐步回归好。这说明网络在某些时效中的泛化能力不强,还需要进行更多的研究。
降水预报方面,改进的k-NN方法在一定程度上克服了数值预报空报偏多的现象,改善了降水预报效果。以验证样本和训练样本的TS评分为目标规则的神经网络(神经网络2)训练得到的最优权重模型,比单纯以训练样本的TS评分为目标的神经网络(神经网络3)得到的最优权重模型的泛化能力要强。神经网络2的预报结果整体要比神经网络3预报效果好,神经网络3在部分站点甚至不能很好的做出预报,当然在个别站点的个别时效神经网络3的预报效果也会比神经网络2的预报效果要好,但整体上神经网络3存在不稳定和泛化能力不足的问题。k-NN方法的预报效果却比较稳定,结果也不逊色。虽然神经网络2预报的结果虽比k-NN方法的预报效果好,但是神经网络2的训练时间确是k-NN方法的训练时间的几百倍。这两种方法各有优缺点,所以在业务运行中,应同时使用这两种方法,取长补短。
研究的目的是为了服务于实际业务需要,本文以我们自主研发的青海省中短期数值预报释用系统和葫芦岛中期数值预报释用系统为例,从应用方面全面介绍两个系统的预报系统框架搭建问题和两套系统的优缺点。青海省中短期数值预报释用系统是我们的第一套自主研发的数值预报释用系统。该系统已在青海气象局气象台投入业务使用,其短期预报工作者对该系统24小时的温度预报给出了较高评价,认为有较高的准确率,为贵单位提高温度预报质量做出了贡献。而葫芦岛中期数值预报释用系统可以说是对青海省中短期数值预报释用系统的一次全方位的升级,它吸取了青海省中短期数值预报释用系统中的先进的理念和技术,总结了青海省中短期数值预报释用系统在实际业务中出现的问题,重新设计了整个系统的框架,使得整个系统能更好的适应科研和业务多方面的需求。
Local weather forecast has always always been concerned by weather forecaster. Because of the lower resolution global forecast model, near-surface boundary layer physical processes, as well as dynamic processes and thermal processes by the impact of underlying surface, simulation results are not satisfactory at the local scale. So the global forecasting models can not produce the satisfactory meteorological elements forecasting result. To obtain more accurate forecasting results, the results from global forecast model are downscaled. Because the dynamic downscaling need more resources and better conditions for the computer, the statistical-dynamic downscaling methods (ie, the model output statistical techniques) just can make up for lack of dynamic downscaling, It is less to calculation and easy to operate, very suitable for grass-roots meteorological stations. In many developed countries, it was the backbone of modern weather forecasting, especially in the numerical model of the post-processing, it was an indispensable part. In this paper, based on previous studies and actual needs, some problems about model output statistics are need to research comprehensively from two aspects of research and practical application.Then, a complete solution scheme is presented.
The model output statistics (MOS) technology is core idea of the concept in full-text. According to research needs, the predictor selection and prediction modeling in model output statistics are study depthly. Then many encouraging results have been achieved. Study stations include 25 weather stations in Qinghai Province and the 23 weather stations around the Bohai region. Test data include daily model output products of National Meteorological Center T213L31 the medium-term global numerical prediction as well as the above-mentioned weather site observations data. Datas is from January 1,2003 to December 31,2009. Maximum temperature, minimum temperature and precipitation are as trial weather object.
The stepwise regression methods associated with correlation coefficient are as predictor selection of statistical downscaling. The experiments have shown that: Although the number of preparation factors on the overall level is less sensitive to the forecasting hitrate, but the number of different preparation factors will have an impact for forecast results in a few sites.
From F value test can be seen that the forecasting result is less sensitive to F values, but the F value can not be smaller, because the smaller F value will cause the forecast results unstable. When a historical observation is as a predictor, the forecast results, especially short-term (24,48 hours) forecast TS have been significantly improved in winter. For the summer, the effect is not very obvious.
In temperature modeling represented by continuous element, neural network are in-depth study. Results show that: Random weights which have a Gaussian distribution with zero mean and a standard deviation of one and multiple weights initialization and dynamic number of neurons in neural networks (ANN) can give a better approximation of the true value. Compared with multiple line regression (MLR) method, In some stations, TS from ANN is slightly higher in winter minimum temperatures and summer maximum temperature forecast. But ANN have the same forecasting level with MLR. It shows that the generalization ability of network is not strong, more research is needed.
For precipitation forecast, the phenomenon of false alarm from numerical prediction results is overcome by improved k-NN method to a certain extent. Neural networks 2 (ANN2) use TS of training samples associated with TS of validating samples as the objective rules. Neural networks 3 (ANN3) only use the TS of training samples as the objective rules. Optimal weights model produced by ANN2 has stronger generalization ability than model builded by ANN3. Forecasting results from ANN2 are better than the result predicted by ANN3. In generally, model builded by ANN3 exit instability and generalization capacity problem. The forecasting effect of k-NN method is fairly stable, although the overall TS of k-NN prediction is lower than TS of ANN2, k-NN prediction results still available. Both methods have advantages, so in real operations, you can use both methods, learn from each other.
Purpose of the study is to serve the needs of practical application. In this paper, there are two examples, short-term numerical prediction interpretation system in Qinghai province and the medium-term numerical prediction interpretation system in Huludao set up by us, the framework, advantages and disadvantages of the two systems will be introduced in terms of the application aspects.
Short-term numerical prediction interpretation system in Qinghai province is our first set of self-developed system. The system has been put into operational in Qinghai Bureau of Meteorology.24-hour temperature forecasts are given a higher rating that has a higher accuracy rate. Medium-term numerical prediction interpretation system in Huludao is comprehensively upgrade, which lenan advanced concepts and techniques from short-term numerical prediction interpretation system in Qinghai province. The framework of the whole system are re-designed, So the entire system can better adapt to the various research and business needs.
全文以模式输出统计(MOS)技术的理念为核心,紧紧围绕这一中心思想,针对模式输出统计中的预报因子选择问题,预报模型建立问题,进行了深入的研究和试验,取得了很多令人鼓舞的结果。文中试验选取青海省内25个气象站和环渤海地区23个气象站作为研究站点,以2003年1月1日—2009年12月31日的国家气象中心T213L31全球中期数值预报模式逐日输出产品以及上述气象站点的地面观测资料作为试验资料,以最高温度、最低温度和降水三个人们最关心的要素作为试验对象。
本文利用相关系数和逐步回归方法进行统计降尺度的因子选择,因子选择的试验表明：预备因子数虽然从总体水平上对最终的命中率不太敏感,但对于个别站点而言,不同的预备因子数会对预报结果的好坏产生一定的影响;通过对F值的试验可以看出,基本上所有台站预报结果的命中率不受F值的影响,但F值不易过小,因为F值太小会造成预报结果的不太稳定。对于加入历史实况作为因子的试验来说,其冬季的预报效果,特别是短期(24,48小时)的预报效果得到了明显的改善,而对于夏季,其效果就不明显。
在以温度为代表的连续要素建模中,本文重点对神经网络进行了深入的研究,结果表明：神经网络的随机权重初始场的权重值,如果服从高斯分布将能使神经网络能更好的向最优解靠近,采用多次初始化初始权重的方法和动态隐层神经元数的方法能让神经网络更好的逼近真实值。通过对环渤海地区冬季最低温度和夏季最高温度的预报试验发现,神经网络与多元回归方法的预报效果平分秋色；在冬季和夏季的部分预报时效中BP-神经网络比逐步回归的评分有一定的提高,有些时效的预报效果并不如逐步回归好。这说明网络在某些时效中的泛化能力不强,还需要进行更多的研究。
降水预报方面,改进的k-NN方法在一定程度上克服了数值预报空报偏多的现象,改善了降水预报效果。以验证样本和训练样本的TS评分为目标规则的神经网络(神经网络2)训练得到的最优权重模型,比单纯以训练样本的TS评分为目标的神经网络(神经网络3)得到的最优权重模型的泛化能力要强。神经网络2的预报结果整体要比神经网络3预报效果好,神经网络3在部分站点甚至不能很好的做出预报,当然在个别站点的个别时效神经网络3的预报效果也会比神经网络2的预报效果要好,但整体上神经网络3存在不稳定和泛化能力不足的问题。k-NN方法的预报效果却比较稳定,结果也不逊色。虽然神经网络2预报的结果虽比k-NN方法的预报效果好,但是神经网络2的训练时间确是k-NN方法的训练时间的几百倍。这两种方法各有优缺点,所以在业务运行中,应同时使用这两种方法,取长补短。
研究的目的是为了服务于实际业务需要,本文以我们自主研发的青海省中短期数值预报释用系统和葫芦岛中期数值预报释用系统为例,从应用方面全面介绍两个系统的预报系统框架搭建问题和两套系统的优缺点。青海省中短期数值预报释用系统是我们的第一套自主研发的数值预报释用系统。该系统已在青海气象局气象台投入业务使用,其短期预报工作者对该系统24小时的温度预报给出了较高评价,认为有较高的准确率,为贵单位提高温度预报质量做出了贡献。而葫芦岛中期数值预报释用系统可以说是对青海省中短期数值预报释用系统的一次全方位的升级,它吸取了青海省中短期数值预报释用系统中的先进的理念和技术,总结了青海省中短期数值预报释用系统在实际业务中出现的问题,重新设计了整个系统的框架,使得整个系统能更好的适应科研和业务多方面的需求。
Local weather forecast has always always been concerned by weather forecaster. Because of the lower resolution global forecast model, near-surface boundary layer physical processes, as well as dynamic processes and thermal processes by the impact of underlying surface, simulation results are not satisfactory at the local scale. So the global forecasting models can not produce the satisfactory meteorological elements forecasting result. To obtain more accurate forecasting results, the results from global forecast model are downscaled. Because the dynamic downscaling need more resources and better conditions for the computer, the statistical-dynamic downscaling methods (ie, the model output statistical techniques) just can make up for lack of dynamic downscaling, It is less to calculation and easy to operate, very suitable for grass-roots meteorological stations. In many developed countries, it was the backbone of modern weather forecasting, especially in the numerical model of the post-processing, it was an indispensable part. In this paper, based on previous studies and actual needs, some problems about model output statistics are need to research comprehensively from two aspects of research and practical application.Then, a complete solution scheme is presented.
The model output statistics (MOS) technology is core idea of the concept in full-text. According to research needs, the predictor selection and prediction modeling in model output statistics are study depthly. Then many encouraging results have been achieved. Study stations include 25 weather stations in Qinghai Province and the 23 weather stations around the Bohai region. Test data include daily model output products of National Meteorological Center T213L31 the medium-term global numerical prediction as well as the above-mentioned weather site observations data. Datas is from January 1,2003 to December 31,2009. Maximum temperature, minimum temperature and precipitation are as trial weather object.
The stepwise regression methods associated with correlation coefficient are as predictor selection of statistical downscaling. The experiments have shown that: Although the number of preparation factors on the overall level is less sensitive to the forecasting hitrate, but the number of different preparation factors will have an impact for forecast results in a few sites.
From F value test can be seen that the forecasting result is less sensitive to F values, but the F value can not be smaller, because the smaller F value will cause the forecast results unstable. When a historical observation is as a predictor, the forecast results, especially short-term (24,48 hours) forecast TS have been significantly improved in winter. For the summer, the effect is not very obvious.
In temperature modeling represented by continuous element, neural network are in-depth study. Results show that: Random weights which have a Gaussian distribution with zero mean and a standard deviation of one and multiple weights initialization and dynamic number of neurons in neural networks (ANN) can give a better approximation of the true value. Compared with multiple line regression (MLR) method, In some stations, TS from ANN is slightly higher in winter minimum temperatures and summer maximum temperature forecast. But ANN have the same forecasting level with MLR. It shows that the generalization ability of network is not strong, more research is needed.
For precipitation forecast, the phenomenon of false alarm from numerical prediction results is overcome by improved k-NN method to a certain extent. Neural networks 2 (ANN2) use TS of training samples associated with TS of validating samples as the objective rules. Neural networks 3 (ANN3) only use the TS of training samples as the objective rules. Optimal weights model produced by ANN2 has stronger generalization ability than model builded by ANN3. Forecasting results from ANN2 are better than the result predicted by ANN3. In generally, model builded by ANN3 exit instability and generalization capacity problem. The forecasting effect of k-NN method is fairly stable, although the overall TS of k-NN prediction is lower than TS of ANN2, k-NN prediction results still available. Both methods have advantages, so in real operations, you can use both methods, learn from each other.
Purpose of the study is to serve the needs of practical application. In this paper, there are two examples, short-term numerical prediction interpretation system in Qinghai province and the medium-term numerical prediction interpretation system in Huludao set up by us, the framework, advantages and disadvantages of the two systems will be introduced in terms of the application aspects.
Short-term numerical prediction interpretation system in Qinghai province is our first set of self-developed system. The system has been put into operational in Qinghai Bureau of Meteorology.24-hour temperature forecasts are given a higher rating that has a higher accuracy rate. Medium-term numerical prediction interpretation system in Huludao is comprehensively upgrade, which lenan advanced concepts and techniques from short-term numerical prediction interpretation system in Qinghai province. The framework of the whole system are re-designed, So the entire system can better adapt to the various research and business needs.
引文
[1]. Shapiro M A, Thorpe A J. THORPEX International Science Plan(Version 3). WMO/TD2No.1246, WWRP/THORPEX No.2,2004.
[2]. Richardson L F. Weather Prediction by Numerical Process. Cambridge University, reprinted Dower,1922:236.
[3]. Charney J G, Fjortoft R, Von Neuman J. Numerical integration of the baroscopic equation. Tellus,1950,2:237-244.
[4]. Charney J G. Dynamical forecasting by numerical process. Compendium of meteorology. Amer Meteor Soc, Boston, MA,1951.
[5]. Smagorinsky J. On the inclusion of moist adiabatic processes in numerical prediction models. Ber. D. Deutsche. Wetterd.1956,5:82-90.
[6]. Manabe S J, Smogorinsky, Strickler R F. Simulated climatology of a general circulation model with a hydrological cycle. Mon Wea Rev,1965,93:769-798.
[7]. Kuo H L. On the formation and intensification of tropical cyclones through latent heat release by cumulus convection. J Atmos Sci,1965,22:40-63.
[8]. Smagorinsky J, Manabe S, Holloway J L. Numerical results from a nine-level general circulation model of the atmosphere. Mon Wea Rev,1965,93:727-768.
[9]. Haltiner G J, Williams R T. Numerical Prediction and Dynamic Meteorology. John Witey& Sons Press, reprinted in USA,1980:1-3.
[10].陈德辉,薛纪善.数值天气预报业务模式现状与展望[J].气象学报,2004,62(S)：623-633.
[11].王小萍.对T213L31模式实际可预报性的诊断研究.浙江大学硕士学位论文.2005.
[12].矫梅燕,龚建东,周兵等.天气预报的业务技术进展.应用气象学报,2006,10,17(5)：594-601.
[13].陈东升,沈桐立,马革兰等,气象资料同化的研究进展,南京气象学院学报,2004,8,27(4)：550-564.
[14]. Johan Andreasson; Sten Bergstrom; Bengt Carlsson; L. Phil Graham. The Effect of Downscaling Techniques on Assessing Water Resources Impacts From Climate Change Scenarios. Water Resources Systems-Water Availability and Global Change.2003.160-164.
[15].陈桂英,赵振国.短期气候预报评估方法和业务初估.应用气象学报.1998,9(2)：180-185.
[16].何慧.降尺度方法在广西月降水量预报中的应用研究.南京信息工程大学硕士学位论文.2007.
[17]. Klein W.H., Lewis F., B.M. and Enger I.,1959:Objective prediction of five-day mean temperatures during winter. J.Meteorol.16,672-682.
[18]. Glahn, H.R., D.A.Lowry,1972:The use of Model Output Statistics (MOS) in objective Weather forecasting. J. Appl. Meteor,11,1203-1211.
[19].孙景荣,曹晓钟,张秀年.基于VC++的气象要素客观预报集成系统软件开发.应用气象学报,2006,8,17(增刊)：147-151.
[20].刘还珠,赵声蓉,陆志善等.国家气象中心气象要素的客观预报-MOS系统.应用气象学报[J],2004,15(2)：181-191.
[21]. Glahn,H.R., and J.R.Bocchieri,1976:Testing the limited area fine mesh model for probility of precipitation forecasting, Mon.Wea.Rev.,104,127-132.
[22]. Lowry,D.A.and H.R.Glahn,1976:An operational model for forecasting probability of precipitation-PEATMOS POP, Mon. wea. Rev.,104,221-232.
[23]. BOCCHIERI J.R., GLAHN H.R.,1976:Verification and further development of an operational model for forecasting the probability of frozen precipitation. Mon.Wea.Rev.,104,691-701.
[24]. McCutchan,M.H.,1978:A model for predicting synoptic weather types based on model output statistics,J.Appl.Meteor.,17,1466-1475.
[25]. Agnew,T.,1979:A comparision of two statistical techniques for forecasting probability of precipitation amount, Preprints Sixth Conf. Probability and Statistics, Banff,Alta.Canada,Amer.Meteor.Soc.,87-92.
[26]. Kalnay,M. Kanamitsu, W.E.Baker, Global numerical weather Prediction at the National Meteorological Center. Bull. Amer. Meteor. Soc[J].,1990,71:1410-1428.
[27]. Black,T.L. The new NMC mesoscale Eta model:Description and forecast examples. Weather Forecasting[J],1994,9:265-278.
[28]. Dallavalle,J.P., A Perspective on the use of Model Output Statistics in objective weather forecasting. Preprints,15thConference on Weather Analysis and Forecasting, Norfolk,VA,Amer. Meteor. Soe[J],1996:479-482.
[29]. Dallavalle,J.P., Changes in the National Weather service's statistical forecasting system. Proceedings EPRI 11th Forecasting symposium, Arlington, VA, Electrical Power Research Institute,1997:10pp.
[30]. Long-Jin Baik, Hong-sub Hwang. Tropical Cyclone Intensity Prediction Using Regression Method and Neural Network, Journal of the Meteorology Society of Japan[J],1998,76(5):711-717.
[31]. J. Shao, Improving Nowcasts of Road Surface Temperature by a Back propagation Neural Network, Weather and Forecasting [J],1998,13(1):164-171.
[32]. Sailor D J, Li x. A Semiempirical Downscaling Approach for Predicting Regional Temperature Impacts Associated with Climatic Change. Journal of Climate,1999,12:103-114.
[33]. Ashok Kumar, Parvinder Maini, S. V Singly, An Operational Model for Forecasting Probability of Precipitation and Yes/No Forecast, Weather and Forecasting[J],1999,14(1):38-48.
[34]. Krishnamurti.T.N, Kishtawal.C.M, LaRow.T,etal. Improved weather and seasonal climate forecasts from multi-model super ensemble. Science[J],1999,285:2548-2550.
[35]. Krishnamurti.T.N, Kishtawal.C.M, Zhang.Z,etal. Multi-model Super ensemble Forecasts for weather and seasonal Climate.FSU Report 99-8,1999.
[36]. Mylne.K.R, Clark.R.T, Evans.R.E. Quasi-operational Multi-model Multi-analysis Ensembles on Medium-range Timescales. AMS 13th Conf on Numerical Weather Prediction, Denver, Colorado,1999:204-209.
[37]. Tony Hall, Precipitation Forecasting Using a Neural Network, Weather and Forecasting(J],1999,14(3):338-345.
[38]. David Silverman, John A. Dracup, Artificial Network and Long-range Precipitation in California, Journal of Applied Meteorology [J],2000,39(1):57-66.
[39]. Tim Bellerby, Martin Todd, Dom Kniveton. Chris Kidd, Rainfall Estimation from a Combination of TRMM Precipitation Radar and GEOS Multispectral Satellite Imagery through the Use of an Artificial Neural Network, Journal of Applied Meteorology[J],2000,39 (12):2115-2128.
[40]. James Murphy. Predictions of Climate Change over Europe using Statistical and Dynamical Downscaling Techniques[J].International Journal of Climatology,2000, 20(5):489-501.
[41]. Caren Marzban, Arthur Witt. A Bayesian Neural Network for Severe-Hail Size Prediction, Weather and Forecasting [J],2001,16(5):600-610.
[42]. Andrew R.Dena, Brina H.Fiedler. Forecasting Warm-Season Bum off of Low Clouds at the San Francisco International Airport Using Linear Regression and a Neural Network. Journal of Applied Meteorology[J],2001,41(6):629-639.
[43]. Glahn,H.R.and J.P.Dallavalle, The new NWS MOS development and implementation systems. Preprints,16th Conference on Probability and Statistics in the Atmospheric sciences, Orlando, FL, Amer. Meteor.Soc.[J],2002:78-81.
[44]. Caren Marzban, Caren Marzban, Neural Networks for Postprocessing Model Output:ARPS, Monthly Weather Review[J],2003,131(6):1103-1111
[45]. Zbynksokol, MOS-Based precipitation Forecasts for River Basins, Weather and Forecasting[J],2003,18(5):76-781.
[46]. Lawrence J. Wilson, Marcel Vallee, The Canadian Updateable Model Output Statistics (UMOS) System: Validation against Perfect Prog, Weather and Forecasting[J],2003,18(2):288-302.
[47]. Auroop R. Ganguly, Rafael L. Bras, Distributed Quantitative Precipitation Forecasting Using Information from Radar and Numerical Weather Prediction Models, Journal of Hydrometeorology[J],2003,4(6):1168-1180.
[48]. Ralf Kretzschmar, Pierre Eckert, Daniel Cattani, Neural Network Classifiers for Local Wind Prediction, Journal of Applied Meteorology[J],2004,43(5):727-738.
[49]. Maria Cleofe Valverde Ramirez, Artificial neural network technique for rainfall forecasting applied to the Sao Paulo region, Journal of Hydrology [J],2005, 301(1-4):146-162.
[50].丁士晨.中国MOS预报的进展,气象学报[J],1985,43(3)：332-338.
[51].朱正心.应用欧洲中心中期天气预报产品作西北太平洋海面风的中期解释预报.气象科学[J],1988,10(2)：113-115.
[52].蔡煌东,宫家文,甘骏人等.运用人工神经网络作汛期降水预报,气象科学,1994,14(4)：386-389.
[53]. Jin Long, Lou Ying, Lin Zhenshan, Comparison of Long-term forecasting of June-August rainfall over Changjiang-Huaihe valley. Advances in Atmospheric Sciences,1997,14(1):87-92.
[54].孙永刚,李彰俊,孟雪峰等.天气动力学组合因子在MOS降水预报中的应用, 气象[J],1998,24(2)：27-30.
[55].钟元.数值预报产品统计释用技术在热带气旋路径客观预报中的应用,热带气象学报[J],1998,11,14(4)：314-322.
[56].孙照渤,谭桂容,赵振国,人工神经网络方法在夏季降水预报中的应用,南京气象学院学报[J],1998,21(1)：47-51.
[57].金龙,陈宁,林振山.基于人工神经网络的集成预报方法研究和比较.气象学报[J],1999,57(2)：198-207.
[58].陈宁,金龙,袁成松,最优子集的神经网络预报建模研究,气象,1999,25(1)： 14-19.
[59].魏凤英.全国夏季降水区域动态权重集成预报试验,应用气象学报[J],1999,10(4)：402-409.
[60].王成刚,吴宝俊,朱官忠,BP网络在鲁西南地区西南涡降水量级预报中的应用试验,气象科学[J],1999(1)：158-165.
[61].刘还珠,郝为,林孔元等.基于智能计算的多模型气象综合预报//刘还珠,汤桂生.暴雨落区预报实用方法.北京：气象出版社[M],2000：30-37.
[62].熊秋芬,王丽,郑启松等,三峡区间面雨量预报方法及其试验效果,气象[J],2000,26(11)：19-23.
[63].胡江林,张礼平,宇如聪,神经网络模型预报湖北汛期降水量的应用研究,气象学报[J],2001,59(6)：776-783.
[64].熊秋芬,胡江林,展军.神经网络方法在静止卫星多通道资料估算降水中的应用,气象[J] 2002,28(9)：17-21.
[65].陆虹,金龙,缪启龙等.影响广西热带气旋频数的神经网络预报模型,南京气象学院学报[J],2003,26(1)：56-62.
[66].金龙,罗莹,王业宏等,月降水量的神经网络混合预报模型研究,高原气象[J],2003,22(6)：618-623.
[67].杞明辉,肖子牛,晏红明.一种改进的考虑环流特征的MOS预报方法,高原气象[J],2003,22(4)：405-409.
[68].张华,叶燕华.利用最近资料改进MOS预报的方法,高原气象[J],2003,22(2)：127-131.
[69].邵明轩,吴庆丽,刘凤辉等.分站客观降水预报试验(英文),北京大学学报(自然科学版)[J],2004,40(1)：46-55.
[70].俞善贤,钟元,滕卫平.热带气旋路径人工神经元预报方法对比试验研究,热带气象学报[J],2004,20(5)：523-529.
[71].赵声蓉.多模式温度预报集成.应用气象学报[J],2006,17(1)：52-58.
[72].周兵,赵翠光,赵声蓉.多模式集合预报技术及其分析与检验.应用气象学报[J],2006,17(增刊)：104-109.
[73].朱双,俞剑蔚,胡洛林等.德国在线和江苏MOS的评估分析及应用,气象科学[J],2006,26(3)：346-349.
[74].吴君,裴洪芹,石莹等.基于数值预报产品的地面气温BP-MOS预报方法气象科学[J],2007,27(4)：429-435.
[75].黄嘉佑,MOS方法的进展,气象科技,1983,03：18-22.
[76].陈起英,姚明明,王雨.国家气象中心新一代业务中期预报模式T213L31的主要特点.气象2004.30(10)：16-21.
[77].陈起英,金之雁,伍湘君等.中期数值预报系统T213L31在IBM/SP高性能计算机上的建立.应用气象学报,2004,15(5)：523-533.
[78].冯松,汤憋苍,王冬梅.青藏高原是我国气候变化启动区的新证据[J].科学通报,1998,43(6)：633-636.
[79].张景华,李英年.青海气候变化趋势及对植被生产力影响的研究.干旱区资源与环境,2008,22(2)：97-102.
[80].李秋,仲桂清.环渤海地区旅游气候资源评价.干旱区资源与环 境,2005,19(2)：149-153.
[81].汪青春,秦宁生,唐红玉等.青海高原近44年来气候变化的事实及其特征.干旱区研究,2007,24(2)：234-239.
[82]. Schaefer, J. T.,1990:Critical success index as an indicator of warning skill. Weather and Forecasting,5,570-575.
[83]. Aritolik, M. S.,2000:An overview of the National Weather Service's centralized statistical quantitative precipitation forecasts.J. Hydrol.,239,306-337.
[84].丑纪范,谢志辉,王式功.建立6-15天数值天气预报业务系统的另类途径.军事气象水文,2006,3(专家特稿).
[85].符长锋,黄嘉佑.MOS预报中降水量的正态化处理,气象[J],1992,18(6),25-30.
[86].曾晓梅,国外人工智能技术在天气预报中的应用综述,气象科[J],1994(4)：4-10.
[87].施能,气象科研与预报中的多元分析方法,北京,气象出版社,1995：205-207.
[88].朱正心,杨秋明,徐竞平等.以T42产品为因子用卡尔曼方法作温度预报.气象科学[J],1995,15(1)：85-88.
[89].薛纪善.美国天气预报技术的发展,气象,1998,24(11)：4-6,53.
[90].朱洁华,豫家良.EOF-CCA模型在区域降水预报中的试验研究.大气科学研究及应用,1998, (1)： 115-119.
[91].黄嘉佑.气象统计分析与预报方法.北京：气象出版社[M],2000.
[92].张礼平,杨志勇,陈正洪.典型相关系数及其在短期气候预报中的应用.大气科学,2000,24(3)：427-432.
[93].郭陵之,标准BP网络在降雨预报上的应用研究,计算机应用与软件[J],2000,17(9)：64-68.
[94].龚佃利,王以琳,谢考宪.山东飞机增雨降水区分级预报方法研究.应用气象 学报[J],2001,12(增刊)：139-145.
[95].王迎春,刘凤辉,张晓玲等.北京地区中尺度非静力数值预报产品释用技术研究.应用气象学报[J],2002,13(3)：312-321.
[96].张万诚,解明恩.奇异值分解方法对降水的预报试验.高原气象[J],2002,21(1)：103-107.
[97].宋海鸥,王永红,顾善齐等.应用K指数和TOT指数制作江苏中期降水预报的试验.气象科学[J],2002,22(2)：242-246.
[98].陈力强,韩秀君,张立群.基于MM5模式的站点降水预报释用方法研究.气象科技[J],2003,31(5)：268-272.
[99].金龙,罗莹,王业宏等,月降水量的神经网络混合预报模型研究,高原气象,2003,22(6)：618-623.
[100].俞善贤,钟元,滕卫平.热带气旋路径人工神经元预报方法对比试验研究,热带气象学报[J],2004,20(5)：523-529.
[101].陈红梅,张耀存,何凤翩.上海气温与太平洋海温的相关关系及年代际差异.南京气象学院学报[J],2005,28(6)：775-756.
[102].徐双柱,堪伟,毛以伟.武汉市盛夏气温预报.气象科学[J],2005,25(4)：419-425.
[103].赵声蓉,裴海瑛.客观定量预报中降水的预处理,应用气象学报[J],2007,18(1)：21-28.
[2]. Richardson L F. Weather Prediction by Numerical Process. Cambridge University, reprinted Dower,1922:236.
[3]. Charney J G, Fjortoft R, Von Neuman J. Numerical integration of the baroscopic equation. Tellus,1950,2:237-244.
[4]. Charney J G. Dynamical forecasting by numerical process. Compendium of meteorology. Amer Meteor Soc, Boston, MA,1951.
[5]. Smagorinsky J. On the inclusion of moist adiabatic processes in numerical prediction models. Ber. D. Deutsche. Wetterd.1956,5:82-90.
[6]. Manabe S J, Smogorinsky, Strickler R F. Simulated climatology of a general circulation model with a hydrological cycle. Mon Wea Rev,1965,93:769-798.
[7]. Kuo H L. On the formation and intensification of tropical cyclones through latent heat release by cumulus convection. J Atmos Sci,1965,22:40-63.
[8]. Smagorinsky J, Manabe S, Holloway J L. Numerical results from a nine-level general circulation model of the atmosphere. Mon Wea Rev,1965,93:727-768.
[9]. Haltiner G J, Williams R T. Numerical Prediction and Dynamic Meteorology. John Witey& Sons Press, reprinted in USA,1980:1-3.
[10].陈德辉,薛纪善.数值天气预报业务模式现状与展望[J].气象学报,2004,62(S)：623-633.
[11].王小萍.对T213L31模式实际可预报性的诊断研究.浙江大学硕士学位论文.2005.
[12].矫梅燕,龚建东,周兵等.天气预报的业务技术进展.应用气象学报,2006,10,17(5)：594-601.
[13].陈东升,沈桐立,马革兰等,气象资料同化的研究进展,南京气象学院学报,2004,8,27(4)：550-564.
[14]. Johan Andreasson; Sten Bergstrom; Bengt Carlsson; L. Phil Graham. The Effect of Downscaling Techniques on Assessing Water Resources Impacts From Climate Change Scenarios. Water Resources Systems-Water Availability and Global Change.2003.160-164.
[15].陈桂英,赵振国.短期气候预报评估方法和业务初估.应用气象学报.1998,9(2)：180-185.
[16].何慧.降尺度方法在广西月降水量预报中的应用研究.南京信息工程大学硕士学位论文.2007.
[17]. Klein W.H., Lewis F., B.M. and Enger I.,1959:Objective prediction of five-day mean temperatures during winter. J.Meteorol.16,672-682.
[18]. Glahn, H.R., D.A.Lowry,1972:The use of Model Output Statistics (MOS) in objective Weather forecasting. J. Appl. Meteor,11,1203-1211.
[19].孙景荣,曹晓钟,张秀年.基于VC++的气象要素客观预报集成系统软件开发.应用气象学报,2006,8,17(增刊)：147-151.
[20].刘还珠,赵声蓉,陆志善等.国家气象中心气象要素的客观预报-MOS系统.应用气象学报[J],2004,15(2)：181-191.
[21]. Glahn,H.R., and J.R.Bocchieri,1976:Testing the limited area fine mesh model for probility of precipitation forecasting, Mon.Wea.Rev.,104,127-132.
[22]. Lowry,D.A.and H.R.Glahn,1976:An operational model for forecasting probability of precipitation-PEATMOS POP, Mon. wea. Rev.,104,221-232.
[23]. BOCCHIERI J.R., GLAHN H.R.,1976:Verification and further development of an operational model for forecasting the probability of frozen precipitation. Mon.Wea.Rev.,104,691-701.
[24]. McCutchan,M.H.,1978:A model for predicting synoptic weather types based on model output statistics,J.Appl.Meteor.,17,1466-1475.
[25]. Agnew,T.,1979:A comparision of two statistical techniques for forecasting probability of precipitation amount, Preprints Sixth Conf. Probability and Statistics, Banff,Alta.Canada,Amer.Meteor.Soc.,87-92.
[26]. Kalnay,M. Kanamitsu, W.E.Baker, Global numerical weather Prediction at the National Meteorological Center. Bull. Amer. Meteor. Soc[J].,1990,71:1410-1428.
[27]. Black,T.L. The new NMC mesoscale Eta model:Description and forecast examples. Weather Forecasting[J],1994,9:265-278.
[28]. Dallavalle,J.P., A Perspective on the use of Model Output Statistics in objective weather forecasting. Preprints,15thConference on Weather Analysis and Forecasting, Norfolk,VA,Amer. Meteor. Soe[J],1996:479-482.
[29]. Dallavalle,J.P., Changes in the National Weather service's statistical forecasting system. Proceedings EPRI 11th Forecasting symposium, Arlington, VA, Electrical Power Research Institute,1997:10pp.
[30]. Long-Jin Baik, Hong-sub Hwang. Tropical Cyclone Intensity Prediction Using Regression Method and Neural Network, Journal of the Meteorology Society of Japan[J],1998,76(5):711-717.
[31]. J. Shao, Improving Nowcasts of Road Surface Temperature by a Back propagation Neural Network, Weather and Forecasting [J],1998,13(1):164-171.
[32]. Sailor D J, Li x. A Semiempirical Downscaling Approach for Predicting Regional Temperature Impacts Associated with Climatic Change. Journal of Climate,1999,12:103-114.
[33]. Ashok Kumar, Parvinder Maini, S. V Singly, An Operational Model for Forecasting Probability of Precipitation and Yes/No Forecast, Weather and Forecasting[J],1999,14(1):38-48.
[34]. Krishnamurti.T.N, Kishtawal.C.M, LaRow.T,etal. Improved weather and seasonal climate forecasts from multi-model super ensemble. Science[J],1999,285:2548-2550.
[35]. Krishnamurti.T.N, Kishtawal.C.M, Zhang.Z,etal. Multi-model Super ensemble Forecasts for weather and seasonal Climate.FSU Report 99-8,1999.
[36]. Mylne.K.R, Clark.R.T, Evans.R.E. Quasi-operational Multi-model Multi-analysis Ensembles on Medium-range Timescales. AMS 13th Conf on Numerical Weather Prediction, Denver, Colorado,1999:204-209.
[37]. Tony Hall, Precipitation Forecasting Using a Neural Network, Weather and Forecasting(J],1999,14(3):338-345.
[38]. David Silverman, John A. Dracup, Artificial Network and Long-range Precipitation in California, Journal of Applied Meteorology [J],2000,39(1):57-66.
[39]. Tim Bellerby, Martin Todd, Dom Kniveton. Chris Kidd, Rainfall Estimation from a Combination of TRMM Precipitation Radar and GEOS Multispectral Satellite Imagery through the Use of an Artificial Neural Network, Journal of Applied Meteorology[J],2000,39 (12):2115-2128.
[40]. James Murphy. Predictions of Climate Change over Europe using Statistical and Dynamical Downscaling Techniques[J].International Journal of Climatology,2000, 20(5):489-501.
[41]. Caren Marzban, Arthur Witt. A Bayesian Neural Network for Severe-Hail Size Prediction, Weather and Forecasting [J],2001,16(5):600-610.
[42]. Andrew R.Dena, Brina H.Fiedler. Forecasting Warm-Season Bum off of Low Clouds at the San Francisco International Airport Using Linear Regression and a Neural Network. Journal of Applied Meteorology[J],2001,41(6):629-639.
[43]. Glahn,H.R.and J.P.Dallavalle, The new NWS MOS development and implementation systems. Preprints,16th Conference on Probability and Statistics in the Atmospheric sciences, Orlando, FL, Amer. Meteor.Soc.[J],2002:78-81.
[44]. Caren Marzban, Caren Marzban, Neural Networks for Postprocessing Model Output:ARPS, Monthly Weather Review[J],2003,131(6):1103-1111
[45]. Zbynksokol, MOS-Based precipitation Forecasts for River Basins, Weather and Forecasting[J],2003,18(5):76-781.
[46]. Lawrence J. Wilson, Marcel Vallee, The Canadian Updateable Model Output Statistics (UMOS) System: Validation against Perfect Prog, Weather and Forecasting[J],2003,18(2):288-302.
[47]. Auroop R. Ganguly, Rafael L. Bras, Distributed Quantitative Precipitation Forecasting Using Information from Radar and Numerical Weather Prediction Models, Journal of Hydrometeorology[J],2003,4(6):1168-1180.
[48]. Ralf Kretzschmar, Pierre Eckert, Daniel Cattani, Neural Network Classifiers for Local Wind Prediction, Journal of Applied Meteorology[J],2004,43(5):727-738.
[49]. Maria Cleofe Valverde Ramirez, Artificial neural network technique for rainfall forecasting applied to the Sao Paulo region, Journal of Hydrology [J],2005, 301(1-4):146-162.
[50].丁士晨.中国MOS预报的进展,气象学报[J],1985,43(3)：332-338.
[51].朱正心.应用欧洲中心中期天气预报产品作西北太平洋海面风的中期解释预报.气象科学[J],1988,10(2)：113-115.
[52].蔡煌东,宫家文,甘骏人等.运用人工神经网络作汛期降水预报,气象科学,1994,14(4)：386-389.
[53]. Jin Long, Lou Ying, Lin Zhenshan, Comparison of Long-term forecasting of June-August rainfall over Changjiang-Huaihe valley. Advances in Atmospheric Sciences,1997,14(1):87-92.
[54].孙永刚,李彰俊,孟雪峰等.天气动力学组合因子在MOS降水预报中的应用, 气象[J],1998,24(2)：27-30.
[55].钟元.数值预报产品统计释用技术在热带气旋路径客观预报中的应用,热带气象学报[J],1998,11,14(4)：314-322.
[56].孙照渤,谭桂容,赵振国,人工神经网络方法在夏季降水预报中的应用,南京气象学院学报[J],1998,21(1)：47-51.
[57].金龙,陈宁,林振山.基于人工神经网络的集成预报方法研究和比较.气象学报[J],1999,57(2)：198-207.
[58].陈宁,金龙,袁成松,最优子集的神经网络预报建模研究,气象,1999,25(1)： 14-19.
[59].魏凤英.全国夏季降水区域动态权重集成预报试验,应用气象学报[J],1999,10(4)：402-409.
[60].王成刚,吴宝俊,朱官忠,BP网络在鲁西南地区西南涡降水量级预报中的应用试验,气象科学[J],1999(1)：158-165.
[61].刘还珠,郝为,林孔元等.基于智能计算的多模型气象综合预报//刘还珠,汤桂生.暴雨落区预报实用方法.北京：气象出版社[M],2000：30-37.
[62].熊秋芬,王丽,郑启松等,三峡区间面雨量预报方法及其试验效果,气象[J],2000,26(11)：19-23.
[63].胡江林,张礼平,宇如聪,神经网络模型预报湖北汛期降水量的应用研究,气象学报[J],2001,59(6)：776-783.
[64].熊秋芬,胡江林,展军.神经网络方法在静止卫星多通道资料估算降水中的应用,气象[J] 2002,28(9)：17-21.
[65].陆虹,金龙,缪启龙等.影响广西热带气旋频数的神经网络预报模型,南京气象学院学报[J],2003,26(1)：56-62.
[66].金龙,罗莹,王业宏等,月降水量的神经网络混合预报模型研究,高原气象[J],2003,22(6)：618-623.
[67].杞明辉,肖子牛,晏红明.一种改进的考虑环流特征的MOS预报方法,高原气象[J],2003,22(4)：405-409.
[68].张华,叶燕华.利用最近资料改进MOS预报的方法,高原气象[J],2003,22(2)：127-131.
[69].邵明轩,吴庆丽,刘凤辉等.分站客观降水预报试验(英文),北京大学学报(自然科学版)[J],2004,40(1)：46-55.
[70].俞善贤,钟元,滕卫平.热带气旋路径人工神经元预报方法对比试验研究,热带气象学报[J],2004,20(5)：523-529.
[71].赵声蓉.多模式温度预报集成.应用气象学报[J],2006,17(1)：52-58.
[72].周兵,赵翠光,赵声蓉.多模式集合预报技术及其分析与检验.应用气象学报[J],2006,17(增刊)：104-109.
[73].朱双,俞剑蔚,胡洛林等.德国在线和江苏MOS的评估分析及应用,气象科学[J],2006,26(3)：346-349.
[74].吴君,裴洪芹,石莹等.基于数值预报产品的地面气温BP-MOS预报方法气象科学[J],2007,27(4)：429-435.
[75].黄嘉佑,MOS方法的进展,气象科技,1983,03：18-22.
[76].陈起英,姚明明,王雨.国家气象中心新一代业务中期预报模式T213L31的主要特点.气象2004.30(10)：16-21.
[77].陈起英,金之雁,伍湘君等.中期数值预报系统T213L31在IBM/SP高性能计算机上的建立.应用气象学报,2004,15(5)：523-533.
[78].冯松,汤憋苍,王冬梅.青藏高原是我国气候变化启动区的新证据[J].科学通报,1998,43(6)：633-636.
[79].张景华,李英年.青海气候变化趋势及对植被生产力影响的研究.干旱区资源与环境,2008,22(2)：97-102.
[80].李秋,仲桂清.环渤海地区旅游气候资源评价.干旱区资源与环 境,2005,19(2)：149-153.
[81].汪青春,秦宁生,唐红玉等.青海高原近44年来气候变化的事实及其特征.干旱区研究,2007,24(2)：234-239.
[82]. Schaefer, J. T.,1990:Critical success index as an indicator of warning skill. Weather and Forecasting,5,570-575.
[83]. Aritolik, M. S.,2000:An overview of the National Weather Service's centralized statistical quantitative precipitation forecasts.J. Hydrol.,239,306-337.
[84].丑纪范,谢志辉,王式功.建立6-15天数值天气预报业务系统的另类途径.军事气象水文,2006,3(专家特稿).
[85].符长锋,黄嘉佑.MOS预报中降水量的正态化处理,气象[J],1992,18(6),25-30.
[86].曾晓梅,国外人工智能技术在天气预报中的应用综述,气象科[J],1994(4)：4-10.
[87].施能,气象科研与预报中的多元分析方法,北京,气象出版社,1995：205-207.
[88].朱正心,杨秋明,徐竞平等.以T42产品为因子用卡尔曼方法作温度预报.气象科学[J],1995,15(1)：85-88.
[89].薛纪善.美国天气预报技术的发展,气象,1998,24(11)：4-6,53.
[90].朱洁华,豫家良.EOF-CCA模型在区域降水预报中的试验研究.大气科学研究及应用,1998, (1)： 115-119.
[91].黄嘉佑.气象统计分析与预报方法.北京：气象出版社[M],2000.
[92].张礼平,杨志勇,陈正洪.典型相关系数及其在短期气候预报中的应用.大气科学,2000,24(3)：427-432.
[93].郭陵之,标准BP网络在降雨预报上的应用研究,计算机应用与软件[J],2000,17(9)：64-68.
[94].龚佃利,王以琳,谢考宪.山东飞机增雨降水区分级预报方法研究.应用气象 学报[J],2001,12(增刊)：139-145.
[95].王迎春,刘凤辉,张晓玲等.北京地区中尺度非静力数值预报产品释用技术研究.应用气象学报[J],2002,13(3)：312-321.
[96].张万诚,解明恩.奇异值分解方法对降水的预报试验.高原气象[J],2002,21(1)：103-107.
[97].宋海鸥,王永红,顾善齐等.应用K指数和TOT指数制作江苏中期降水预报的试验.气象科学[J],2002,22(2)：242-246.
[98].陈力强,韩秀君,张立群.基于MM5模式的站点降水预报释用方法研究.气象科技[J],2003,31(5)：268-272.
[99].金龙,罗莹,王业宏等,月降水量的神经网络混合预报模型研究,高原气象,2003,22(6)：618-623.
[100].俞善贤,钟元,滕卫平.热带气旋路径人工神经元预报方法对比试验研究,热带气象学报[J],2004,20(5)：523-529.
[101].陈红梅,张耀存,何凤翩.上海气温与太平洋海温的相关关系及年代际差异.南京气象学院学报[J],2005,28(6)：775-756.
[102].徐双柱,堪伟,毛以伟.武汉市盛夏气温预报.气象科学[J],2005,25(4)：419-425.
[103].赵声蓉,裴海瑛.客观定量预报中降水的预处理,应用气象学报[J],2007,18(1)：21-28.