共查询到20条相似文献，搜索用时 640 毫秒
1.
西北太平洋热带气旋强度统计释用预报方法研究 总被引：4，自引：1，他引：4
为了提高西北太平洋地区热带气旋（TC）强度预报准确率，在气候持续预报方法基础上，考虑气候持续性因子、天气因子、卫星资料因子，以TC强度变化为预报对象，运用逐步回归统计方法，建立西北太平洋地区24、48、72小时TC强度预报方程。通过不同的分海区试验（远海区域、华东近海、华南近海），证明回归结果较好。逐一分析选入因子发现：气候持续性因子在方程中相当重要；同时对远海区域和华东近海而言，海温影响也不容忽视，对华南近海而言，反映动力强迫作用的因素也较为重要。卫星资料的加入，对回归结果略有改进。用“刀切法”作独立样本检验，与气候持续法比较，预报误差明显减小。 相似文献
2.
Application of Equivalent Black Body Temperature in the Forecast of Tropical Cyclone Intensity 
下载免费PDF全文
下载免费PDF全文 Using equivalent black body temperature (TBB) data retrieved from meteorological satellite GMS-5 during 1996-2002,the correlation between the circular symmetric/asymmetric component of TBB and the intensity of tropical cyclone (TC) at various time lags from 0 to 48 h is analyzed for the Northwest Pacific (0°-50°N,120°-155°E),excluding landed and near-coast samples.It is found that the total TBB near southeast of the eyewall,the circular symmetric component,and the sum of the amplitudes of tangential wave numbers 1-10 (SA10) of the TBB between the radii of 0.8°and 1.7°are significantly and negatively correlated with the TC intensity at various time lags from 0 to 48 h.Especially,the maximum 24-h lag correlation coefficients reach -0.52,-0.58,and -0.625,respectively. A statistical prediction scheme for TC intensity is developed based on climatic persistent,synoptic,and TBB factors by stepwise regression technique.It is found that the variance contribution of the averaged TBB over the ring between 1.0°and 1.5°from the TC center ranks the fourth in the equation for 12-h TC intensity prediction,and those of the total TBB near southeast of the eyewall and the difference between maximum and minimum TBB between 1.1°and 1.5°rank the third and fifth respectively in the 24-h forecast equation.It is also shown that,with TBB factors,the following predictions are improved compared to the scheme without TBB factors:48-h prediction for severe tropical storm (STS),12-h prediction for TC with a weakening rate greater than 15 m s~(-1)/12 h,24-h intensity prediction for TC with almost no intensity change,and 48-h prediction for TC intensifying faster than 10 m s~(-1)/48 h. 相似文献
3.
A western North Pacific tropical cyclone (TC) intensity prediction scheme (WIPS) is developed based
on TC samples from 1996 to 2002 using the stepwise regression technique, with the western North Pacific
divided into three sub-regions: the region near the coast of East China (ECR), the South China Sea region
(SCR), and the far oceanic region (FOR). Only the TCs with maximum sustained surface wind speed greater
than 17.2 m s-1 are used in the scheme. Potential predictors include the climatology and persistence factors,
synoptic environmental conditions, potential intensity of a TC and proximity of a TC to land. Variances
explained by the selected predictors suggest that the potential intensity of a TC and the proximity of a TC
to land are significant in almost all the forecast equations. Other important predictors include vertical wind
shear in ECR, 500-hPa geopotential height anomaly at the TC center, zonal component of TC translation
speed in SCR, intensity change of TC 12 or 24 h prior to initial time, and the longitude of TC center in
FOR.
Independent tests are carried out for TCs in 4 yr (2004-2007), with mean absolute errors of the maximum
surface wind being 3.0, 5.0, 6.5, 7.3, 7.6, and 7.9 m s-1 for 12- to 72-h predictions at 12-h intervals,
respectively. Positive skills are obtained at all leading time levels as compared to the climatology and
persistence prediction scheme, and the large skill scores (near or over 20%) after 36 h imply that WIPS
performs especially better at longer leading times. Furthermore, it is found that the amendment in TC
track prediction and real-time model analysis can significantly improve the performance of WIPS in the
SCR and ECR. Future improvements will focus on applying the scheme for weakening TCs and those near
the coastal regions. 相似文献
4.
夏季南海台风移动路径的一种客观预报方法 总被引：1，自引：0，他引：1
以1960—2003年共44a夏季的7月、8月、9月西行进入南海海域的台风样本为基础,综合考虑南海台风移动路径的气候持续因子和数值预报产品物理量因子,运用条件数方法选取因子并建立回归方程,进行台风路径预报模型的预报建模研究。通过对比分析发现,基于条件数方法的南海台风移动路径模型具有较好的预报效果,7月、8月、9月3个月24h台风路径预报的平均距离误差为153．9km,预报能力明显高于目前国内外的其他一些台风路径客观预报方法。该方法的预报精度相对于逐步回归方法有了很大的提高,相对于气候持续法也为正的预报技巧水平。 相似文献
5.
以1960-2007年共48年6月份西行进入南海海域的热带气旋样本为基础, 将热带气旋中心附近最大风速作为台风强度, 以气候持续预报因子作为模型输入, 采用模糊神经网络方法, 进行了热带气旋强度预报模型的预报建模研究。结果表明, 对175个独立预报样本模糊神经网络方法的南海热带气旋强度24 h的预报平均绝对误差为3 m·s-1。另外, 根据相同的热带气旋样本及预报因子, 还进一步将该预报方法与国内外普遍采用的气候持续法热带气旋强度预报方法进行对比分析, 结果表明, 气候持续预报方法的预报误差明显偏大, 独立样本强度预报平均绝对误差为4.54 m·s-1。 相似文献
6.
A western North Pacific tropical cyclone (TC) intensity prediction scheme (WIPS) is developed based on TC samples from 1996
to 2002 using the stepwise regression technique, with the western North Pacific divided into three sub-regions: the region
near the coast of East China (ECR), the South China Sea region (SCR), and the far oceanic region (FOR). Only the TCs with
maximum sustained surface wind speed greater than 17.2 m s−1 are used in the scheme. Potential predictors include the climatology and persistence factors, synoptic environmental conditions,
potential intensity of a TC and proximity of a TC to land. Variances explained by the selected predictors suggest that the
potential intensity of a TC and the proximity of a TC to land are significant in almost all the forecast equations. Other
important predictors include vertical wind shear in ECR, 500-hPa geopotential height anomaly at the TC center, zonal component
of TC translation speed in SCR, intensity change of TC 12 or 24 h prior to initial time, and the longitude of TC center in
FOR. 相似文献
7.
南海热带气旋大风的遗传－神经网络集合预报 总被引：1，自引：0，他引：1
利用1980-2012年的南海热带气旋实况资料和NCEP/NCAR再分析资料,将热带气旋定位中心周边6×6格点上的地面风速作为预报对象,以气候持续预报因子和前期风速预报因子作为模型输入,采用遗传—神经网络集合预报方法,进行热带气旋定位中心周边36个格点上的风速预报模型的预报建模研究.分别对2008-2012年7-9月共368个独立预报样本进行遗传-神经网络集合方法的分月预报结果表明,南海热带气旋中心周边风速24h的预报平均绝对误差为2.35m.s-1.另外,本文还进一步将该预报方法与国内外普遍采用的逐步回归预报模型进行对比分析,在相同的预报量和预报因子的条件下的对比分析表明,新预报模型对≥10m.s-1的强风预报结果较逐步回归方法的优势明显,预报性能较好,可为沿海热带气旋大风预报提供新的参考. 相似文献
8.
论文以逐日气温和降水量数据、NCEP/NCAR再分析资料以及预报场资料为基础,将表征冬季低温冷害的冷湿指数作为预报量,先利用随机森林方法进行冬季逐日冷湿极端天气定性判别预报分析,再进一步以粒子群算法为基础的模糊神经网络集成个体生成技术方法,建立一种新的非线性智能计算定量集成预报模型（PSO-FNN）,进行了广西冷湿极端天气定量预报模型的预报建模研究。结果表明,论文提出的这种以不同的智能计算方法构建的定性、定量综合预报分析方法,比较符合极端天气小概率事件的预报特点,其中随机森林算法构建的定性预报模型,对广西冷湿极端天气事件的预报TS评分（Threat Score）为0.77,空报率为0.23,漏报率为0,ETS评分（Equitable Threat Score）为0.41,TSS评分（True Skill Statistic）为0.53。而采用粒子群—模糊神经网络方法构建的极端冷湿指数定量集成预报模型比其他线性和非线性预报模型具有更好的预报精度。其中PSO-FNN集成预报模型在预报建模样本和独立预报样本个例相同的情况下,比回归方法的预报平均绝对误差下降了25%以上,比一般的普通模糊神经网络预报平均绝对误差下降了14.37%。主要原因是因为PSO-FNN集成预报模型通过改进集成个体的预报能力和增强集成个体的种群差异性,提高了集成预报模型的预报精度。因此,该智能计算集成预报模型的泛化能力显著提高,预报结果稳定可靠,为冷湿极端天气客观预报提供了新的预报工具和预报建模方法。 相似文献
9.
A predictive model for 24-120 h track of the tropical cyclone over the South ChinaSea is set upusing the predictors of CLIPER and with reference to those of EOF and CCA. This EOF-and CCA-basedmodel is compared with dependent and independent samples in veritring forecasts'Analpees appltwg themethods to 500-hPa geopotential heights have shown that the CCA method is able toconcentrate moreoriginal information of variable fleld that has the best global correlation withtropical cyclone track andthus reduces more efficiently error arisen in forecast,despite that can0nicaldistribution of weighted vari-able coefficients derived by it is less well-defined or smoothed of a spacialpattern as is done by eigenvec-tor in the EOF method. The verification indicates a certain degree of skill in both track predictive model,though better results are yielded in the CCA-based one, whose mean vector errorare 159.27, 314.84,524.12, 813.03, and 987.12 km, respectively for periods of 24, 48, 72, 96, and 120 h. The modelmakes an objective approach for prediction of short-and medium-range track of tropical cyclone in theSouth China Sea. 相似文献
10.
Accurate prediction of tropical cyclone (TC) intensity remains a challenge due to the complex physical processes involved in TC intensity changes. A seven-day TC intensity prediction scheme based on the logistic growth equation (LGE) for the western North Pacific (WNP) has been developed using the observed and reanalysis data. In the LGE, TC intensity change is determined by a growth term and a decay term. These two terms are comprised of four free parameters which include a time-dependent growth rate, a maximum potential intensity (MPI), and two constants. Using 33 years of training samples, optimal predictors are selected first, and then the two constants are determined based on the least square method, forcing the regressed growth rate from the optimal predictors to be as close to the observed as possible. The estimation of the growth rate is further refined based on a step-wise regression (SWR) method and a machine learning (ML) method for the period 1982?2014. Using the LGE-based scheme, a total of 80 TCs during 2015?17 are used to make independent forecasts. Results show that the root mean square errors of the LGE-based scheme are much smaller than those of the official intensity forecasts from the China Meteorological Administration (CMA), especially for TCs in the coastal regions of East Asia. Moreover, the scheme based on ML demonstrates better forecast skill than that based on SWR. The new prediction scheme offers strong potential for both improving the forecasts for rapid intensification and weakening of TCs as well as for extending the 5-day forecasts currently issued by the CMA to 7-day forecasts. 相似文献
11.
基于主成分分析的人工智能台风路径预报模型 总被引：1，自引：0，他引：1
利用主成分分析可以从具有随机噪声干扰的气象场提取主要信号特征,排除随机干扰的能力,论文以1980～2010年共31年6～9月西行进入南海海域的台风样本为基础,综合考虑台风移动路径的气候持续因子和数值预报产品动力预报因子,采用主成分分析的特征提取与逐步回归计算相结合的预报因子信息数据挖掘技术,以进化计算的遗传算法,生成期望输出相同的多个神经网络个体,建立了一种新的非线性人工智能集合预报模型,进行了分月台风路径预报模型的预报建模研究。在预报建模样本、独立预报样本相同的情况下,分别采用人工智能集合预报方法和气候持续法进行了预报试验,试验对比结果表明,前者较后者在6、7、8和9月份台风路径预报中,平均绝对误差分别下降了7.4%、4.8%、12.4%、17.0%。另外,论文进一步在初选预报因子和样本个例相同的情况下,通过比较新模型与直接采用主成分分析方法选因子并分别运用逐步回归和遗传—神经网络集合预报模型进行计算的预报精度差异表明,前者具有更高的预报精度,其原因是该方法挖掘利用了全部备选预报因子的有用预报信息,而且遗传—神经网络集合预报模型的是由多个神经网络个体预报结果合成,集合模型的各个神经网络个体的网络结构,是通过遗传算法的优化计算确定的,因此,该集合预报模型的泛化能力显著提高,在实际天气预报中具有较好的实用性和推广价值。 相似文献
12.
After the consideration of the nonlinear nature changes of monsoon index,and the subjective determination of network structure in traditional artificial neural network prediction modeling,monthly and seasonal monsoon intensity index prediction is studied in this paper by using nonlinear genetic neural network ensemble prediction（GNNEP）modeling.It differs from traditional prediction modeling in the following aspects： （1）Input factors of the GNNEP model of monsoon index were selected from a large quantity of preceding period high correlation factors,such as monthly sea temperature fields,monthly 500-hPa air temperature fields,monthly 200-hPa geopotential height fields,etc.,and they were also highly information-condensed and system dimensionality-reduced by using the empirical orthogonal function（EOF）method,which effectively condensed the useful information of predictors and therefore controlled the size of network structure of the GNNEP model.（2）In the input design of the GNNEP model,a mean generating function（MGF）series of predictand（monsoon index）was added as an input factor;the contrast analysis of results of predic- tion experiments by a physical variable predictor-predictand MGF GNNEP model and a physical variable predictor GNNEP model shows that the incorporation of the periodical variation of predictand（monsoon index）is very effective in improving the prediction of monsoon index.（3）Different from the traditional neural network modeling,the GNNEP modeling is able to objectively determine the network structure of the GNNNEP model,and the model constructed has a better generalization capability.In the case of identical predictors,prediction modeling samples,and independent prediction samples,the prediction accuracy of our GNNEP model combined with the system dimensionality reduction technique of predictors is clearly higher than that of the traditional stepwise regression model using the traditional treatment technique of predictors,suggesting that the GNNEP model opens up a vast range of possibilities for operational weather prediction. 相似文献
13.
Prediction of extreme rainfall event using weather pattern recognition and support vector machine classifier 总被引：1，自引：1，他引：0
A major component of flood alert broadcasting is the short-term prediction of extreme rainfall events, which remains a challenging task, even with the improvements of numerical weather prediction models. Such prediction is a high priority research challenge, specifically in highly urbanized areas like Mumbai, India, which is extremely prone to urban flooding. Here, we attempt to develop an algorithm based on a machine learning technique, support vector machine (SVM), to predict extreme rainfall with a lead time of 6–48 h in Mumbai, using mesoscale (20–200 km) and synoptic scale (200–2,000 km) weather patterns. The underlying hypothesis behind this algorithm is that the weather patterns before (6–48 h) extreme events are significantly different from those of normal weather days. The present algorithm attempts to identify those specific patterns for extreme events and applies SVM-based classifiers for extreme rainfall classification and prediction. Here, we develop the anomaly frequency method (AFM), where the predictors (and their patterns) for SVM are identified with the frequency of high anomaly values of weather variables at different pressure levels, which are present before extreme events, but absent for non-extreme conditions. We observe that weather patterns before the extreme rainfall events during nighttime (1800 to 0600Z) is different from those during daytime (0600 to 1800Z) and, accordingly, we develop a two-phase support vector classifier for extreme prediction. Though there are false alarms associated with this prediction method, the model predicts all the extreme events well in advance. The performance is compared with the state-of-the-art statistical technique fingerprinting approach and is observed to be better in terms of false alarm and prediction. 相似文献
14.
为了进一步提高国家气象中心区域模式台风数值预报系统（GRAPES_TYM）的预报能力,2016年对模式参考大气廓线以及涡旋初始化方案进行了改进:由模式初始场水平方向平均的一维参考大气代替原来的等温大气,涡旋初始化方案取消了原涡旋重定位并将涡旋强度调整半径由原来的12°减小到4°。对2014—2016年的生命史超过3 d的所有台风进行了回算,路径及近地面最大风速统计误差分析表明:参考大气的改进可以减小模式对台风预报路径预报的系统北偏和平均路径误差,尤其是140°E以东的转向台风。涡旋初始化方案中强度调整半径的减小会进一步减小模式预报路径的北偏趋势,从而进一步减小平均误差。同业务系统预报结果相比,改进后的GRAPES_TYM（包括参考大气和涡旋初始化）可以使平均路径误差分别减小10%（24 h）,12%（48 h）,16%（72 h）,14%（96 h）以及15%（120 h）。同美国NCEP全球模式路径预报相比,GRAPES_TYM在西行、西北行登陆我国的台风路径预报有一定优势。 相似文献
15.
利用2003-2007年国家气象中心T213L31全球中期数值预报模式逐日输出产品与青海地区25个气象站的观测数据作为试验资料, 利用相关系数和逐步回归进行因子选择, 并以单隐层神经网络和多元回归作为降尺度方法进行对比研究, 用2003-2006年间的11月1日～次年3月1日的资料作为训练样本, 以数值预报产品和前一日观测的最低温度作为因子, 建立青海省25个气候站的冬季最低温度的24, 48, 72 h预报模型, 并且以2006年12月和2007年的1、 2月作为24, 48, 72 h逐日最低温度预报试验时段。试验表明, 对于青海地区来说, 青海北部地区的预报命中率总体好于南部高原地区; 在4种对比方案中, 以选择数值预报资料结合前一日地面观测的最低温度作为主要因子的方法相对较优, 随着预报时效的延长, 24 h历史实况的作用逐渐减弱; 对于所有台站来说, 这4种方案各有优缺点, 没有一种方案可以完全代替其他所有方案; 在实际业务运行中, 对不同的台站应采用不同的预报方案进行实际业务预报。 相似文献
16.
基于遗传算法的神经网络短期气候预测模型 总被引：15，自引：3，他引：12
用遗传算法优化神经网络的连接权和网络结构,并在遗传进化过程中采取保留最佳个体的方法,进行短期气候预测建模研究。该方法克服了由于神经网络初始权值的随机性和网络结构确定过程中所带来的网络振荡,以及网络极易陷入局部解问题。作为应用实例,以广西全区4月份平均降水作为预报量及前期500hPa月平均高度场,海温场高相关区作为预报因子,建立基于遗传算法的神经网络短期气候预测模型。将这种方法与传统的逐步回归方法作对比分析,结果表明,该方法具有预报精度高,稳定性好的特点。 相似文献
17.
The effect of three-dimensional variational data assimilation of QuikSCAT data on the numerical simulation of typhoon track and intensity 总被引：4，自引：0，他引：4
In this paper, the three-dimensional variational data assimilation scheme (3DVAR) in the mesoscalemod el version 5 (MM5) of the US Pennsylvania State University/National Center for Atmospheric Research is used to study the effect of assimilating the sea-wind data from QuikSCAT on the prediction of typhoon track and intensity. The case of Typhoon Dujuan (2003) is first tested and the results show appreciable improvements. Twelve other cases in 2003 are then evaluated. The assimilation of the QuikSCAT data produces significant impacts on the structure of Dujuan in terms of the horizontal and vertical winds, sealevel pressure and temperature at the initial time. With the assimilation, the 24-h (48-h) track prediction of 11 (10) out of the 12 typhoons is improved. The 24-h (48-h) prediction of typhoon intensity is also improved in 10 (9) of the 12 cases. These experiments therefore demonstrate that assimilation of the QuikSCAT sea-wind data can increase the accuracy of typhoon track and intensity predictions through modification of the initial fields associated with the typhoon. 相似文献
18.
19.
Predictor selection is a critical factor affecting the statistical downscaling of daily precipitation. This study provides a general comparison between uncertainties in downscaled results from three commonly used predictor selection methods (correlation analysis, partial correlation analysis, and stepwise regression analysis). Uncertainty is analyzed by comparing statistical indices, including the mean, variance, and the distribution of monthly mean daily precipitation, wet spell length, and the number of wet days. The downscaled results are produced by the artificial neural network (ANN) statistical downscaling model and 50 years (1961–2010) of observed daily precipitation together with reanalysis predictors. Although results show little difference between downscaling methods, stepwise regression analysis is generally the best method for selecting predictors for the ANN statistical downscaling model of daily precipitation, followed by partial correlation analysis and then correlation analysis. 相似文献