基于MM5模式的精细化MOS温度预报

利用2002年9月到2003年8月MM5模式每隔1 h的站点基本要素预报场和物理量诊断场资料,以及相应时段内宁夏25个测站的温度自记观测资料,同时采用多元线性和逐步回归2种MOS统计方法,预报宁夏25个测站48 h逐时温度。通过对2004年6月至2005年5月的预报效果检验,结果表明:MOS方法对宁夏48 h逐时温度有较强的预报能力,当天气形势变化较平稳时,MOS预报结果稳定,平均绝对误差控制在2℃之内;当有明显的冷空气活动时,误差变率起伏波动较大,预报结果不稳定,但与原MM5模式直接输出结果相比,MOS预报水平有明显提高,24 h极端温度TS评分个别月接近甚至超过预报员。     

风的精细化MOS预报方法研究

利用MM5模式的站点输出产品,采用多元线性和逐步回归的MOS统计方法,预报宁夏25个测站48 h逐时风速、风向,通过对2004年6月~2005年4月近一年的预报效果检验,结果表明:该方法对逐时风速有较强的预报能力,对逐时风向有一定的预报能力,但尚未达到可用程度,还有待于进一步的改进,但都较原MM5模式预报水平有了显著的提高。     

MOS方法在动力延伸期候平均气温预报中的应用

利用1982年1月-2010年3月动力延伸预报产品及宁夏候平均气温,采用逐步回归的MOS统计方法,预报宁夏24个测站未来40 d逐候平均气温,为了对比模式直接输出结果与资料按月和按季节划分建立的MOS方法预报效果,对2009年和2010年1-3月的预报效果分别进行了对比检验.结果表明:MOS预报效果较模式直接输出有显著...     

基于WRF模式输出统计的逐时太阳总辐射预报初探

基于WRF(weather research and forecasting model)模式逐时输出结果,设计了逐时太阳总辐射的模式输出统计(model output statistics,MOS)预报流程。主要包括:对逐时观测序列进行低通滤波再除以天文辐射,对模式输出因子的筛选和降维,以及建立MOS预报方程,并对2009年1月、4月、8月和10月武汉站逐时太阳总辐射进行预报试验。结果表明,该方案在各月预报相对稳定,拟合和预报效果均较为理想,可使平均绝对百分比误差控制在20%～30%,相对均方根误差控制在30%～40%,相对模式直接预报辐射改进了50%左右。由此可见,通过对模式输出进行解释应用,可以有效提高辐射预报的准确率。此外,客观分析所得的气温、云量、露点、比湿、相对湿度、地面气压等13个模式输出因子可以作为各地区建立MOS辐射预报方程的参考因子。     

支持向量机非线性回归方法的气象要素预报

该文介绍了基于基本的支持向量机非线性回归方法,该方法具有解决非线性问题的能力,在数值预报解释应用技术中,对某些预报量与预报因子之间相关性不显著的要素,如风、比湿等,采用支持向量机非线性回归技术较多元回归的MOS方法更具优势;利用北京市气象局中尺度业务模式 (MM5V3) 的12:00(世界时) 起始数值预报产品和观测资料,制作北京15个奥运场馆站点6~48 h逐3 h的气象要素释用产品。对比MM5V3模式,从均方根误差的平均减小率来看,2 m温度减小12.1%,10 m风u分量减小43.3%,10 m风v分量减小53.4%,2 m比湿减小38.2%。与同期的MOS方法预报结果相比,整体预报效果SVM略优于MOS。由此可见,支持向量机非线性回归方法解决与预报因子之间非线性相关的气象要素较好,具有较高的预报优势。     

BJ-RUC系统模式地面气象要素预报效果评估

利用自动气象站逐小时地面观测资料,采用客观检验方法对北京市气象局快速更新循环预报 (BJ-RUC) 系统在2008—2010年5—9月的预报结果进行检验,初步评估了BJ-RUC系统对地面气象要素的业务预报性能。结果表明:BJ-RUC系统对地面气象要素预报与实况的变化趋势有很好的一致性。其中,2 m温度预报整体偏高,误差范围为-1.5~1.5℃,早上和傍晚偏大,正午偏小;2 m相对湿度的预报整体偏低,误差为-25%~0,白天偏大,夜间偏小;10 m风速预报明显偏大,午后尤为显著,误差为0.6~1.2 m·s-1;6 h累积降水的晴雨预报效果较好,TS评分可达到0.4。系统在初始起报时次的稳定性较差,从第3个起报时次开始逐渐稳定,但预报误差随着预报时效的增长逐渐增大,12 h内的预报误差较小,预报结果较可靠,在短时临近预报中具有参考价值。     

基于OMOS方法的地面气温短时预报——以重庆市为例

基于重庆市气象局业务运行的风暴尺度快速同化和预报系统（Storm-Scale Rapid Assimilation and Forecast System,SSRAFS）、气象信息综合分析处理系统（Meteorological Information Comprehensive Analysis And Process System,MICAPS）地面观测和高空观测资料,进行模式输出统计（Model Output Statistics,MOS）方法和纳入超前实况因子的MOS （MOS with Prior Observation Predictors,OMOS）方法对重庆地区地面气温96 h内逐小时预报试验,并以SSRAFS地面气温预报结果作为参考进行对比分析。结果表明：MOS方法在1~96 h预报时效内的预报技巧高于SSRAFS,气温预报均方根误差（Root Mean Square Error,RMSE）平均减小1.22℃,CC和HR2分别平均增大0.006和20.4%;在1~7 h预报时效,RMSE平均减小1.70℃,CC和HR2分别平均增大0.07和34.5%;且MOS方法在重庆东北部及中南部地区改进效果较为明显。OMOS方法在气温短期预报中表现优于MOS方法,尤其在1~7 h预报时效,比MOS方法RMSE平均减小0.43℃,CC和HR2分别平均增大0.008和8.3%;其在1~4 h预报时效时表现更加优异,与MOS方法相比,RMSE平均减小0.66℃,CC和HR2分别平均增大0.13和12.3%。因此,在MOS的基础上,OMOS能够进一步提升地面气温的预报技巧,且在重庆东北部及中南部地区的预报效果有明显改进。     

ROAD模式渤海10 m风场预报误差订正

使用"递减平均法"和渤海28个石油、平台、浮标站资料分析得到的渤海10m风速逐时格点场,对ROAD模式(Regional Ocean and Atmosphere Model)渤海区域10m风速预报进行误差订正,不同权重系数试验表明:对于渤海10m风速预报场,权重系数取0.18,订正效果最佳,12～72h预报时效内,月逐时均方根误差和平均偏差均有明显改善,分别减小1.0～1.5m/s和2.3～3.0m/s;对比渤海北部、西部和中部代表格点72h预报时效内,逐12h最大风速评分结果发现:60h预报时效内,当风速预报在5～6级时,渤海北部、中部和西部订正后的预报评分大多有所提高;当风速预报在7级时,渤海北部和中部分别在36～72h和24～48h,预报评分有提高;在实况出现最多的风力分布范围内评分提高最多。     

MOS方法在短时要素预报中的应用与检验

通过对MM5数值预报产品的释用,将MOS统计方法应用到短时要素预报中,综合利用MM5数值预报产品、自动站实况数据和雷达数据等资料,建立降水和温度的4 h预报模型;降水作为不连续变量,将其通过建立降水可能函数的方法转化为连续变量,利用统计预报方法,可以达到定量预报的目的。通过对4140个时次的样本进行检验。结果表明：MOS预报结果较MM5直接输出结果整体有所改进,当数值模式误差较大时,统计方法显示出一定的优势;降水预报检验结果显示,TS评分为65%,预报正确率PC为91%。降水明显的样本(3 h雨量>5 mm)平均误差在8 mm以内,弱降水样本(3 h雨量<3 mm)平均误差在1 mm以内,预报方程对非雨日样本的整体预报效果较好,优于MM5模式预报,预报正确率高达98%,但对流性降水仍是预报难点;对于温度预报,20-08时段误差较小、平均在1.0℃以内,而11-17时、误差平均在1.5℃左右,但经过误差的季节订正,可以控制在1.0℃左右。     

MOS方法在机场气象要素客观预报中的应用

利用WRF中尺度数值模式和模式输出统计（MOS）方法,研究建立乌鲁木齐机场逐时温度、相对湿度的回归预报模型,并尝试针对冬季低云、低能见度等天气建立分类预报模型,通过对统计模型的检验可以看到：逐时温度绝对差为1.09耀2.33益;逐时相对湿度绝对差为4.7%耀9.6%;11月至翌年2月低云量跃5分量分类预报准确率76.94%耀83.13%,TS评分54.27%耀66.50%;11月至2月能见度臆800m的分类预报准确率为89.83%耀92.04%,TS评分为29.09%耀46.05%。该方法预测效果较好,因此可以尝试使用本方法为日后航空气象业务提供机场客观预报指导产品。     

宁夏区域精细化温度预报业务平台

介绍以宁夏中尺度数值模式温度预报为基础，以宁夏精细化预报系统温度预报产品为核心，结合自动气象站等多种资料，以图形方式显示、修改和制作宁夏各站逐时温度预报业务平台。该平台以宁夏各地区代表站与该地区其它站之间的回归方程的计算量为依据，在温度预报物理过程不变的情况下，通过修改曲线的方式，完成对大数据量温度预报值的订正。该平台的建成，为制作高时间密度的预报提供了技术支撑。     

基于MOS方法的风向预测方案对比研究

基于MOS方法的数值预报产品释用技术是要素预测较好手段,风向的释用预测方案一般采用标量式的释用预测方案,效果并不理想。本文给出一种基于MOS方法的风向矢量预测方案,利用相关系数和逐步回归方法分别得到u风和v风在不同站点和不同时效的风向模型,根据模型预测u风和v风合成风向。试验选择风向矢量、风向标量和模式的直接输出三种方案进行对比。试验选择T639模式逐日多个物理量观测场,其中建模样本选择2008-2012年每年5月15日至9月15日的数据,验证样本选择2012年6月1-30日的数据,数值试验预测北京地区20个站点未来4个时次的定点定时风向。结果表明：基于MOS方法的风向矢量预测方案12、24、36h和48h的准确率比基于MOS的风向标量观测方案的各时效准确率分别提高了85.0%、26.3%、113.0%和19.9%,比模式直接输出结果方案的各时效准确率分别提高了35.3%、20.2%、43.6%和21.0%。总体来看,基于MOS方法的风向矢量预测方案在风向方面有较好的预测能力。对于个别站点,可能对因子模型的估计不理想或风向变化较为频繁,导致预测难度较大。     

基于特定因子的河南省干热风客观预报方法

利用2000—2014年5月1日到6月10日河南省121个气象观测站点的逐日观测数据、欧洲中心模式预报资料,对河南省干热风天气进行分析,总结了干热风天气形势分类模型,同时利用多元回归法建立了河南省干热风天气的客观预报方法。分析结果得出:河南省干热风天气发生主要形势为西北气流型、高压脊型和纬向环流型3类;通过多元回归分析筛选出日最高温度预报因子为前一日最高气温、当日最低气温、08:00气温、EC850hPa 24h温度预报,相对湿度预报因子为EC850hPa 24h相对湿度预报、前一日14:00相对湿度、当日08:00露点温度,风速预报因子为EC细网格过去3h10m阵风预报,建立温度、湿度和风速3要素的预报方程;利用预报方程对2014年预报时段的天气进行检验,结果表明,对于轻干热风预报的TS评分为62%,重干热风预报的TS评分为64%。     

ECMWF模式地面气温预报的四种误差订正方法的比较研究

采用均方根误差对欧洲中期天气预报中心（ECWMF）确定性预报模式2007年1月至2010年12月的地面气温预报结果进行评估,并分别利用一元线性回归、多元线性回归、单时效消除偏差和多时效消除偏差平均的订正方法,对ECMWF模式地面气温预报结果进行订正。结果表明,4种订正方法都能有效地减小地面气温多个时效预报的误差,改进幅度约为1℃。在短期预报中仅考虑最新预报结果的一元线性回归订正方法要优于考虑多个预报结果的多元集成预报订正方法。在中期预报中考虑多个预报结果的多元集成预报订正方法更优,更稳定。在模式预报误差较大的情况下,多时效集成的订正方法能更稳定地减小误差。     

Comparison of spatial interpolation methods for gridded bias removal in surface temperature forecasts

All numerical weather prediction (NWP) models inherently have substantial biases, especially in the forecast of near-surface weather variables. Statistical methods can be used to remove the systematic error based on historical bias data at observation stations. However, many end users of weather forecasts need bias corrected forecasts at locations that scarcely have any historical bias data. To circumvent this limitation, the bias of surface temperature forecasts on a regular grid covering Iran is removed, by using the information available at observation stations in the vicinity of any given grid point. To this end, the running mean error method is first used to correct the forecasts at observation stations, then four interpolation methods including inverse distance squared weighting with constant lapse rate (IDSW-CLR), Kriging with constant lapse rate (Kriging-CLR), gradient inverse distance squared with linear lapse rate (GIDS-LR), and gradient inverse distance squared with lapse rate determined by classification and regression tree (GIDS-CART), are employed to interpolate the bias corrected forecasts at neighboring observation stations to any given location. The results show that all four interpolation methods used do reduce the model error significantly, but Kriging-CLR has better performance than the other methods. For Kriging-CLR, root mean square error (RMSE) and mean absolute error (MAE) were decreased by 26% and 29%, respectively, as compared to the raw forecasts. It is found also, that after applying any of the proposed methods, unlike the raw forecasts, the bias corrected forecasts do not show spatial or temporal dependency.     

Impact of surface observations on simulation of rainfall over NCR Delhi using regional background error statistics in WRF-3DVAR model

In this study, efforts are made to improve the simulation of heavy rainfall events over National Capital Region (NCR) Delhi during 2010 summer monsoon, using additional observations from automatic weather stations (AWS). Two case studies have been carried out to simulate the relative humidity, wind speed and precipitation over NCR Delhi in 48-h model integrations; one from 00UTC, August 20, 2010, and the other from 00UTC, September 12, 2010. Several AWS installed over NCR Delhi in the recent past provide valuable surface observations, which are assimilated into state-of-the-art weather research and forecasting (WRF) model using the three-dimensional variational data assimilation (3DVAR). The quality of background error statistics (BES) is a key component in successful 3DVAR data assimilation in a mesoscale model. In this study, the domain-dependent regional background error statistics (RBS) are estimated using National Meteorological Center method in the months of August and September 2010 and then compared with the global background error statistics (GBS) in the WRF model. The model simulations are analyzed and validated against AWS and radiosonde observations to quantify the impact of RBS. The root mean square differences in the spatial distributions of precipitation, relative humidity and wind speed at the surface showed significant differences between both the global and regional BES. Similar differences are also observed in the vertical distributions along the latitudinal cross section at 28.5°N. Model-simulated fields are analyzed at five different surface stations and one upper air station located in NCR Delhi. It is found that in 24-h model simulation, the RBS significantly improves the model simulations in case of precipitation, relative humidity and wind speed as compared to GBS.     

OTS、MOS和OMOS方法及其优化组合应用于72 h内逐3 h降水预报的试验分析研究

开展了夏半年72 h内逐3 h降水预报试验,针对ECMWF模式预报、基于ECMWF的模式输出统计(MOS)预报、纳入超前空间实况信息的OMOS预报,以及三种预报的最优TS评分订正(OTS)预报,对比分析预报效果,探讨一种多方法结合能够提供良好预报性能的3 h定量降水预报技术方案。结果表明:在短期预报中,MOS预报与OTS订正相结合的MOSOTS综合预报方法的预报性能最好,而且MOS-OTS方法的3 h强降水预报与业务运行的城镇指导预报中融合主客观预报的降水预报相比,也具有一定优势;而在临近3 h预报中,则OMOS预报与OTS订正相结合的OMOS-OTS综合预报方法最优,3 h内0.1、3和10 mm以上降水的TS评分最高,比原始模式预报分别提高73%、198%和483%,Bias评分接近于1,在夏半年的逐日晴雨预报中,OMOS-OTS方法在大部分日期都稳定优于MOS-OTS预报和城镇指导预报。     

A Deep Learning Method for Bias Correction of ECMWF 24–240 h Forecasts

Correcting the forecast bias of numerical weather prediction models is important for severe weather warnings. The refined grid forecast requires direct correction on gridded forecast products, as opposed to correcting forecast data only at individual weather stations. In this study, a deep learning method called CU-net is proposed to correct the gridded forecasts of four weather variables from the European Centre for Medium-Range Weather Forecast Integrated Forecasting System global model(ECMWF-IFS): 2-m temperature, 2-m relative humidity, 10-m wind speed, and 10-m wind direction, with a forecast lead time of 24 h to 240 h in North China. First, the forecast correction problem is transformed into an image-toimage translation problem in deep learning under the CU-net architecture, which is based on convolutional neural networks.Second, the ECMWF-IFS forecasts and ECMWF reanalysis data(ERA5) from 2005 to 2018 are used as training,validation, and testing datasets. The predictors and labels(ground truth) of the model are created using the ECMWF-IFS and ERA5, respectively. Finally, the correction performance of CU-net is compared with a conventional method, anomaly numerical correction with observations(ANO). Results show that forecasts from CU-net have lower root mean square error, bias, mean absolute error, and higher correlation coefficient than those from ANO for all forecast lead times from 24 h to 240 h. CU-net improves upon the ECMWF-IFS forecast for all four weather variables in terms of the above evaluation metrics, whereas ANO improves upon ECMWF-IFS performance only for 2-m temperature and relative humidity. For the correction of the 10-m wind direction forecast, which is often difficult to achieve, CU-net also improves the correction performance.     

Analysis of the surface temperature and wind forecast errors of the NCAR-AirDat operational CONUS 4-km WRF forecasting system

Investigating the characteristics of model-forecast errors using various statistical and object-oriented methods is necessary for providing useful guidance to end-users and model developers as well. To this end, the random and systematic errors (i.e., biases) of the 2-m temperature and 10-m wind predictions of the NCAR-AirDat weather research and forecasting (WRF)-based real-time four-dimensional data assimilation (RTFDDA) and forecasting system are analyzed. This system has been running operationally over a contiguous United States (CONUS) domain at a 4-km grid spacing with four forecast cycles daily from June 2009 to September 2010. In the result an exceptionally useful forecast dataset was generated and used for studying the error properties of the model forecasts, in terms of both a longer time period and a broader coverage of geographic regions than previously studied. Spatiotemporal characteristics of the errors are investigated based on the 24-h forecasts between June 2009 and April 2010, and the 72-h forecasts between May and September 2010. It was found that the biases of both wind and temperature forecasts vary greatly seasonally and diurnally, with dependency on the forecast length, station elevation, geographical location, and meteorological conditions. The temperature showed systematic cold biases during the daytime at all station elevations and warm biases during the nighttime above 1,000 m above sea level (ASL), while below 600 m ASL cold biases occurred during the nighttime. The forecasts of surface wind speed exhibited strong positive biases during the nighttime, while the negative biases were observed in the spring and summer afternoons. The surface wind speed was mostly over-predicted except for the stations located between 1,000 and 2,100 m ASL, for which negative biases were identified for most forecast cycles. The highest wind-speed errors were found over the high terrain and near sea-level stations. The wind-direction errors were relatively large at the high-terrain elevation in the Rocky and Appalachian mountain ranges and the western coastal areas and the error structure exhibited notable diurnal variability.