共查询到19条相似文献，搜索用时 234 毫秒
1.
2.
MOS，PP方法在降水及温度预报中的效果对比检验 总被引：1，自引：0，他引：1
利用1995～1997年的北京T106数值预报产品资料,建立了山东省各测站夏季（6～9）1～5天日极端温度、降水MOS预报方程和PP预报方程,并对1999年7～9月的预报进行了对比检验,结果表明,MOS方法预报效果明显优于PP方法。 相似文献
3.
朱国栋 《沙漠与绿洲气象(新疆气象)》2013,7(3):13-16
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。 相似文献
4.
基于MM5模式的精细化MOS温度预报 总被引：5，自引：0，他引：5
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。 相似文献
5.
支持向量机（SVM）方法在气象预报中的个例试验 总被引：4，自引：0，他引：4
SVM（Support Vector Machine）方法的基本思想是：通过非线性映射把样本空间映射到一个高维乃至于无穷维的特征空间（Hilbert空间），在特征空间中寻求最优划分或回归线性超平面，从而解决样本空间中的高度非线性分类和回归等问题。利用1998～2000年9～11月T106—36小时预报的各种输出产品构造因子，以单站有无降水为预报对象，采用MOS法方式构造样本，通过训练建立了单站的晴雨SVM分类预报模型，利用1990～2000年4～9月ECMWF北半球的500hPa高度、850hPa温度、地面气压的0小时分析场资料，确定关键区域，构造预报因子，以PP法方式构造样本，通过训练建立了四川盆地内单站气温的SVM回归预报模型，并进行了模拟试验，结果表明，无论是单站晴雨的SVM分类预报模型还是单站平均气温的SVM回归预报模型都显示出了艮好的预报能力。 相似文献
6.
7.
8.
按我区自然天气季节的划分,初夏为6月14日至7月15日,初秋为8月31日至9月14日。在每个天气阶段内,根据影响我区天气的形势特征划分环流型,按型建立预报方程,这是我们改进MOS预报的一个方面。从投入业务使用的效果来看,按自然天气季 相似文献
9.
利用2016年4月1日—2018年12月31日GRAPES_GFS模式的业务预报数据,将GRAPES_GFS模式在东亚地区144小时预报500 hPa高度场的距平相关系数小于0.4,均方根误差大于60 gpm的事件定义为模式在东亚地区的中期预报技巧极端下降事件,据此标准筛选出29个个例进行了研究。结果表明,GRAPES_GFS在东亚地区中期预报技巧极端下降事件的高发期主要在春秋季,春季和秋季分别占比31.03%、41.38%,预报技巧极端下降事件通常表现在对于东亚大槽、西伯利亚槽脊、副热带高压的预报失败,四个季节合成的模式偏差均与该季节影响东亚地区主要天气系统的预报偏差大有关系。进一步通过典型个例的研究表明,东亚地区中期预报技巧极端下降事件的误差来源在不同季节表现出不同特征。冬夏季的预报偏差来源于高纬极区,多与模式在极区存在较大预报误差关联;春秋季的预报偏差主要来源于上游地区,与模式在东亚上游预报误差向下游的传播有关,未能合理预报台风活动也是预报偏差来源之一。 相似文献
10.
SVM(Support Vector Machine)方法的基本思想是:通过非线性映射把样本空间映射到一个高维乃至于无穷维的特征空间(Hilbert空间),在特征空间中寻求最优划分或回归线性超平面,从而解决样本空间中的高度非线性分类和回归等问题.利用1998～2000年9～11月T106-36小时预报的各种输出产品构造因子,以单站有无降水为预报对象,采用MOS法方式构造样本,通过训练建立了单站的晴雨SVM分类预报模型,利用1990～2000年4～9月ECMWF北半球的500hPa高度、850hPa温度、地面气压的0小时分析场资料,确定关键区域,构造预报因子,以PP法方式构造样本,通过训练建立了四川盆地内单站气温的SVM回归预报模型,并进行了模拟试验,结果表明,无论是单站晴雨的SVM分类预报模型还是单站平均气温的SVM回归预报模型都显示出了良好的预报能力. 相似文献
11.
北半球环状模(NAM)、东北冷涡与前汛期华南旱涝 总被引：11，自引：3，他引：11
利用中国气象局国家气象信息中心提供的1951-2004年全国160站月平均降水资料和欧洲中心提供的ERA-40再分析资料，对近50年华南前汛期降水、东北冷涡、前期北半球环状模和海温的关系进行了统计分析，定义了一个前汛期东北冷涡强度指数（NECVI），结果表明：前汛期东北冷涡强度与华南降水存在显著的正相关，东北冷涡强年，华南降水偏多，前期2-3月份北半球环状模（NAM）偏弱；反之，东北冷涡偏弱年，华南降水偏少，前期2-3月份NAM偏强。此外，前汛期东北冷涡的强度和前期的中国近海海温存在显著的负相关，前期的NAM和中国近海海温的异常可以作为前汛期东北冷涡异常的一个前兆信号，进而为华南地区前汛期降水异常的预测提供参考依据。 相似文献
12.
S. del Río A. Cano-Ortiz L. Herrero A. Penas 《Theoretical and Applied Climatology》2012,109(3-4):605-626
This study analyzes the mean maximum and minimum temperature trends on a monthly, seasonal, and annual timescale by applying various statistical tools to data from 476 Spanish weather stations during the period between 1961 and 2006. The magnitude of the trends was derived from the slopes of the regression lines using the least squares method, and the nonparametric Mann–Kendall test was used to determine the statistical significance of the trends. Temperature significantly increased in over 60% of the country in March, June, spring, and summer in the case of maximum temperatures and in March, May, June, August, spring, and summer for minimum temperatures. At the annual resolution, temperatures significantly increased in over 90% of Spain with a rise of around 0.3°C/decade. The maximum temperature increased at a higher rate than the minimum temperature from midsummer to early winter as well as in winter, spring, and summer and also on an annual basis. 相似文献
13.
D. R. Pattanaik U. C. Mohanty H. R. Hatwar G. Srinivasan Y. V. Ramarao P. Sinha A. Brookshaw 《Theoretical and Applied Climatology》2008,92(3-4):165-179
Summary A comparative study was performed to evaluate the performance of the UK Met Office’s Global Seasonal (GloSea) prediction General
Circulation Model (GCM) for the forecast of maximum surface air temperature (Tmax) over the Indian region using the model generated hindcast of 15-members ensemble for 16 years (1987–2002). Each hindcast
starts from 1st January and extends for a period of six months in each year. The model hindcast Tmax is compared with Tmax obtained from verification analysis during the hot weather season on monthly and seasonal scales from March to June.
The monthly and seasonal model hindcast climatology of Tmax from 240 members during March to June and the corresponding observed climatology show highly significant (above 99.9% level)
correlation coefficients (CC) although the hindcast Tmax is over-estimated (warm bias) over most parts of the Indian region. At the station level over New Delhi, although the forecast
error (forecast-observed) at the monthly scale gradually increases from March to June, the forecast error at the seasonal
scale during March to May (MAM) is found to be just 1.67 °C. The GloSea model also simulates well Tmax anomalies on monthly and seasonal scales during March to June with the lower Root Mean Square Error (RMSE) of bias corrected
forecast (less than 1.2 °C), which is much less than the corresponding RMSE of climatology (reference) forecast. The anomaly
CCs (ACCs) over the station in New Delhi are also highly significant (above 95% level) on monthly to seasonal time scales
from March to June, except for April.
The skill of the GloSea model for the seasonal forecast of Tmax as measured from the ACC map and the bias corrected RMSE map is reasonably good during MAM and April to June (AMJ) with higher
ACC (significant at 95% level) and lower RMSE (less than 1.5 °C) found over many parts of the Indian regions.
Authors’ addresses: D. R. Pattanaik, H. R. Hatwar, G. Srinivasan, Y. V. Ramarao, India Meteorological Department (IMD), New
Delhi, India; U. C. Mohanty, P. Sinha, Centre for Atmospheric Sciences, Indian Institute of Technology, Hauz Khas, New Delhi
110016, India; Anca Brookshaw, UK Met Office, UK. 相似文献
14.
Alice Favre Bruce Hewitson Mark Tadross Christopher Lennard Ruth Cerezo-Mota 《Climate Dynamics》2012,38(7-8):1473-1487
The characteristics of Southern Cut-off Lows (CoLs) are studied for the period 1979–2008. The systematic identification of CoLs is realized by applying an original automated scheme using mean daily geopotential height and air temperature at 500?hPa NCEP-DOE II Reanalysis data. From closed lows’ trajectories established from the Equator to the polar jet stream, extratropical lows are analyzed and the stage of cut-off is defined as a secluded low presenting a cold core. From 4,843 cases the general CoL features are presented and confirm several previous results such as the geographic distribution which shows that they are more frequent in the latitudinal band contained between 20°S and 45°S and in close proximity to the continents. On a seasonal time scale, CoLs are more numerous from late summer to autumn, with a maximum of frequency in March/April. In winter (June–July–August), they are fewer but deeper than during the rest of the year. In the median domain (~32.5°S), the annual cycle of the frequency is bimodal and present two peaks during transitional seasons. In this zone, the seasonal cycle varies in accordance with the Semiannual Oscillation. Thereby, when the meridional gradient of temperature/pressure is reinforced between mid and high latitudes, CoLs are more frequent in the median domain. Over the period 1979–2008, the annual CoLs’ frequency exhibits a positive trend of about 25%. This increase is associated with a widening of the latitudinal domain of occurrence equatorward as well as poleward. The trend is linked with an abrupt positive shift in the number of CoLs’ generation between 1998 and 1999. The geographical distribution of CoLs frequency varies significantly in accordance with El Ni?o Southern Oscillation with more CoL’s at lower (higher) latitudes during La Ni?a (El Ni?o) events, principally in the Southern Pacific. 相似文献
15.
Influence of surface heterogeneity on scalar dissimilarity in the roughness sublayer 总被引：1，自引：8，他引：1
Christopher A. Williams Todd M. Scanlon John D. Albertson 《Boundary-Layer Meteorology》2007,122(1):149-165
While it is generally known that surface heterogeneity weakens the application of Monin–Obukhov similarity (MOS), few studies
have investigated how seasonal changes in the degree of surface heterogeneity at a particular site may influence the validity
of the similarity application. Exploiting seasonal changes in forest function associated with senescence, we conduct a unique
evaluation of the effects of surface heterogeneity on the validity of similarity theory at two sites through time. Using high
frequency (10 Hz) velocity and scalar time series collected within the roughness sublayer over mixed hardwood deciduous and
coniferous forests during both periods of peak leaf area and senescence of deciduous foliage, we examined conformity with
proposed universal flux-variance predictions and agreement amongst normalized standard deviations of different scalars (temperature,
water vapour and carbon dioxide concentrations). Normalized scalar standard deviations were elevated above MOS flux-variance
predictions, with more pronounced deviations observed during and following senescence, particularly in the case of CO2. Power-law scaling of normalized standard deviations as a function of stability was upheld and robust to seasonal changes
in surface heterogeneity. However, dissimilarity of normalized standard deviations for the scalars increased during senescence,
as heterogeneity in the source/sink field increased. Scalewise decomposition of scalar time series using wavelet analysis
indicated that correlations between scalars were conservative through much of the inertial cascade but decayed for eddies
< 10 m. Senescence lowered correlations between scalars over a wide range of eddy sizes. These results demonstrate how seasonal
changes in surface physiology can cause a temporal production of heterogeneity in the source/sink field, thus weakening similarity
applications in the roughness sublayer. 相似文献
16.
利用常规观测资料、NCEP再分析资料和风云2E卫星资料对2006年3月-2016年3月由温带气旋引发黑龙江省暴雪天气过程普查,共筛选出37次区域性暴雪天气过程,再运用统计和诊断分析得出以下结论:(1)近10 a黑龙江省区域性暴雪主要集中在东部山区东南部;区域暴雪主要出现在季节过渡期;黑龙江省区域性暴雪年际变化大;(2)蒙古气旋作为冬季黑龙江省降雪的主要影响系统,其出现的频率和时间跨度远超过其它影响系统,黑龙江省西北部只有在蒙古气旋影响下才可能产生暴雪天气。(3)单独北上温带气旋通常发生在季节转换期间,造成区域暴雪主要集中在黑龙江省东部;(4)北上温带气旋与蒙古气旋合并而形成的合并气旋都发生11-12月和3-4月季节转换期间,造成区域暴雪主要集中在黑龙江省东部。 相似文献
17.
C. Lavaysse C. Flamant S. Janicot D. J. Parker J.-P. Lafore B. Sultan J. Pelon 《Climate Dynamics》2009,33(2-3):313-330
The West African heat low (WAHL), a region of high surface temperatures and low surface pressures, is a key element of the West African monsoon system. In this study, we propose a method to detect the WAHL in order to monitor its climatological seasonal displacement over West Africa during the period 1979–2001, using the European Centre for Medium-range Weather Forecast (ECMWF) ERA-40 reanalyses. The low-level atmospheric thickness (LLAT), a variable defined as the difference of geopotential heights at 700 and 925 hPa, is used to detect the dilatation of these levels generated by an increase of the temperature. We define grid points with 10% highest values of the LLAT as the WAHL. We show that our method reliably positions the WAHL over areas of high surface temperatures and low surface pressures, and that it is effective at detecting heat lows. In the course of the year, the climatological WAHL is shown to migrate north-westward from a position south of the Darfur mountains in the winter (November–March) to a location over the Sahara, between the Hoggar and the Atlas mountains, during the summer (June–September). The temperature tendency equation is used to investigate the processes controlling the displacement of the WAHL, and more particularly the heating at low levels. The specific period of the onset of the WAHL in its summer location over the Sahara (referred to as the Saharan heat low -SHL- onset) is also analysed during the 1984–2001 period, using complementary brightness temperature data from the European Union-funded Cloud Archive User Service (CLAUS). The climatological onset of the SHL occurs around 20 June, i.e. just before the climatological monsoon onset date. The present study suggests that the onset of the WAHL occurs approximately 5 days before the monsoon onset for the 1984–2001 period. This is confirmed independently by comparing the SHL onset date and the monsoon onset date for the 1984–2001 period. The seasonal evolution of the WAHL for the year 2006 (the year of the African Monsoon Multidisciplinary Analysis project Special Observation Period) is analysed and compared with the climatological results. The operational ECMWF analyses were used for that purpose. Except in April, the spatial distribution of the WAHL remains relatively unchanged and agrees with the climatology. The onset of the SHL in 2006 occurs on 18 June, which is close to the climatological date, in spite of the delay in the onset of the rainy season in Sahel. 相似文献
18.
《Atmospheric Research》2005,73(1-2):69-85
This paper reports the results of the analysis of annual mean temperature and precipitation series from 171 meteorological stations distributed over Castile and Leon [Castilla y León in Spanish] in Spain on monthly, seasonal and annual time-scales for a 37-year study period (1961–1997). Various statistical tools were used to detect and characterize significant changes in these series. The magnitude of the trends was derived from the slopes of the regression lines using the least squares method, and the statistical significance was determined by means of nonparametric tests. Positive trends of about 0.33 °C in the annual mean temperature were found for the whole period. Mean temperatures increased in spring and winter, the winter trend being statistically significant. The months of December and March also showed significant trends. Decreases in rainfall were found for three seasons (winter, spring and autumn), with statistically significant trends in March. Summer precipitation showed slight increases over the 37-year period. On this basis, the authors consider that the increase in summer precipitation and the decrease in the range of average temperatures between the warmest and the coldest months of the year (continentality), point towards a trend to a more oceanic climate in Castile and Leon. 相似文献
19.
利用2010年3月至2003年12月镇江市水泥路面、柏油路面两种下垫面的地面气温数据,分析镇江市道路结冰的气候特征,结合降水性质、是否有积雪、日最低气温和日平均气温,定义道路结冰影响指数,并对其分级预报进行检验。结果表明：镇江市道路结冰月季分布不均,主要集中在冬季,11月和3月出现次数很少,其变化趋势与镇江11月至翌年3月常年月平均气温的变化趋势呈反位相;在考虑当日或者前一日出现降雪或者有积雪的情况下,水泥、柏油两种下垫面的地温≤0 ℃并持续16 h以上将会出现对交通有严重影响的道路结冰;道路结冰主要出现在夜间,白天如果出现了道路结冰,当日其它时次也会出现严重的道路结冰,对交通的影响将比仅夜间出现更大。道路结冰影响指数对连续的道路结冰具有一定的预报能力,但是空报率比较高。 相似文献