Further experiments on cyclone track prediction with a quasi-Lagrangian limited area model

Summary Results of an earlier study of cyclone track prediction using a quasi-Lagrangian model (QLM) to generate track forecasts of up to 36 hours were reported by Prasad and Rama Rao (2003). Further experiments to produce track forecasts of up to 72 hours with an updated version of the same model have been carried out in the present study. In this case, the ability of the model to predict recent historical cyclones in the Bay of Bengal and Arabian Sea has been assessed. Analysis of some of the structural features of analyzed and predicted fields has been carried out. Such fields include wind distribution and vertical motion around the cyclone centre. In addition, the merging of an idealized vortex with the large scale initial fields provided by a global model, has been carried out for a particular case study of a May 1997 storm, which hit the Bangladesh coast. This current study has demonstrated that the model generates a realistic structure of a tropical cyclone with an idealized vortex. Performance evaluation has been carried out by computing the direct position errors (DPE). The results of which show that the mean error for a 24 h forecast is about 122 km, which increases to about 256 km for a 48 h forecast and 286 km for a 72 h forecast. These figures are comparable to similar errors in respect of tropical cyclone forecasts produced by an advanced NWP centre, viz., the UKMO global model during the corresponding period, 1997–2000 (obtained from UKMO web site). The average forecast errors of the UKMO model are 160 km for 24 h, 265 km for 48 h, 415 km for 72 h forecast ranges.     

Evaluation of long-term trends in tropical cyclone intensity forecasts

Summary The National Hurricane Center and Joint Typhoon Warning Center operational tropical cyclone intensity forecasts for the three major northern hemisphere tropical cyclone basins (Atlantic, eastern North Pacific, and western North Pacific) for the past two decades are examined for long-term trends. Results show that there has been some marginal improvement in the mean absolute error at 24 and 48 h for the Atlantic and at 72 h for the east and west Pacific. A new metric that measures the percent variance of the observed intensity changes that is reduced by the forecast (variance reduction, VR) is defined to help account for inter-annual variability in forecast difficulty. Results show that there have been significant improvements in the VR of the official forecasts in the Atlantic, and some marginal improvement in the other two basins. The VR of the intensity guidance models was also examined. The improvement in the VR is due to the implementation of advanced statistical intensity prediction models and the operational version of the GFDL hurricane model in the mid-1990s. The skill of the operational intensity forecasts for the 5-year period ending in 2005 was determined by comparing the errors to those from simple statistical models with input from climatology and persistence. The intensity forecasts had significant skill out to 96 h in the Atlantic and out to 72 h in the east and west Pacific. The intensity forecasts are also compared to the operational track forecasts. The skill was comparable at 12 h, but the track forecasts were 2 to 5 times more skillful by 72 h. The track and intensity forecast error trends for the two-decade period were also compared. Results showed that the percentage track forecast improvement was almost an order of magnitude larger than that for intensity, indicating that intensity forecasting still has much room for improvement.     

Improved tropical cyclone track predictions using error recycling

Summary Errors produced by a nonlinear predictive scheme contain information about both the observations and the prediction system. Therefore, its error history would be expected to contribute to increasing the skill of the predictions if it is included in the forecast. In this study an error recycling procedure is developed for tropical cyclone track prediction. Errors are defined here as differences between the model forecast and the best track position. Error histories are incorporated into a nonlinear analogue, or simplex, forecast scheme and applied to tropical cyclone track prediction, using the archives of observed position data associated with the forecast errors. Various forecast experiments of the cyclone tracks are performed: standard simplex predictions using observed positions only; simplex predictions improved by error forecasts based on libraries of both observations and the recycled forecast errors; and, finally, predictions that include NWP-model forecasts and their errors as predictors. The resulting gains in skill of predictions out to 72 hours ahead are found to be substantial. Received August 12, 1999 Revised November 5, 1999     

Ensemble forecasting of tropical cyclone motion: comparisonbetween regional bred modes and random perturbations

Summary Random perturbations (RPs) and a modified version for breeding of growing modes are used with a regional baroclinic mesoscale model to perform ensemble forecasting of tropical cyclone motion. Based on a sample of six cases, similar conclusions are found as in previous barotropic modeling studies. Even after introducing a larger spatial correlation into the RPs using a multi-quadric analysis scheme, the skill of this ensemble mean track prediction is almost always lower than that of the control forecast in the cases considered. The track prediction performance of the ensemble using regional bred modes (RBMs) as perturbations has a higher average skill. At nearly all forecast intervals except less than 24 h when the initial position error still dominates, the ensemble mean tracks in all six cases are improved over the control forecast. In the 6 h–24 h range, the success rate (ratio of the cases with a forecast improvement to the total number of cases) has a value of 10/24. In the 30 h–48 h range, the success rate increases to 20/24, but drops to 18/24 in the 54 h–72 h range. A relative skill score (RSS) is used to compare the skills of the two perturbation methodologies. It is found that the average RSSs of using RBMs are significantly higher than the corresponding ones of RPs at the 99% confidence level in all three 24-h periods. Note that the above conclusion is only based on ensemble mean forecasts. All of the possibilities from an ensemble-based probabilistic track distribution are not explored in this paper. The ensemble spreads in these RBM ensembles are large enough to include the verifying tracks in all the cases considered. It is also found that the ensemble spread is well correlated with the average error in an ensemble when using RBMs, but not with the ensemble mean forecast error in both methodologies. Received February 7, 2001/Revised April 18, 2001     

High resolution simulations of an intense convective precipitation event

Summary A newly developed non-hydrostatic model (MOLOCH), operating at a resolution of about 2 km, is run for a case of heavy precipitation over southeastern France. The event (8–9 September 2002) was characterized by intense convective activity leading to a severe flash flood in the region of the Gard river, south of the Massif Central. An almost stationary mesoscale convective system (MCS), developing well in advance of an approaching cold front, discharged a huge amount of rainfall over the same area, more than 600 mm in 24 hours. Several simulations are performed in order to test the model set-up, evaluate the sensitivity on different initial conditions, and analyse the case-study. The quantitative precipitation forecasts (QPF) appear to vary widely among the experiments, depending on the initialization time chosen (00, 06, and 12 UTC, September 8). Only the run starting at 06 UTC predicts, with a satisfactory degree of accuracy, the location where the MCS developed and its almost stationary behavior during the first stage (∼12 hours) of the event. In all the simulations, the convective system then propagated northward over the Massif Central. In addition to experiments starting from standard ECMWF analyses, an assimilation procedure, based on Optimal Interpolation, is applied to the initial conditions. Surface observations of temperature, wind and relative humidity have been assimilated. The assimilation produces an improvement in the forecasts of surface fields and leads to a better location of the initial triggering phase. Further experiments, performed by changing the orography in the model, allow the investigation of the role of the Massif Central in triggering the mesoscale convective system and in controlling its evolution.     

Improving Precipitation Forecasts over the Global Tropical Belt

Summary  This study explores the nowcasting and short-range forecasting (up to 3 days) skills of rainfall over the tropics using a high resolution global model. Since the model-predicted rainfall is very sensitive to model parameters, four key model parameters were first selected. They are the Asselin filter coefficient, the fourth order horizontal diffusion coefficient, the surface moisture flux coefficient, and the vertical diffusion coefficient. The optimal values were defined as those which contributed to the best one day rainfall forecasts in the present study. In order to demonstrate and improve the precipitation forecast skill, several numerical experiments were designed using the 14-level Florida State University Global Spectral Model (FSUGSM) at a resolution of T106. Comparisons were also made of the short-range forecasts obtained from a control experiment subjected to normal mode initialization (NMI) versus experiments based on physical initialization (PI). The latter experiments were integrated using the original FSUGSM and a modified version. This modified FSUGSM was developed here by applying a reverse cumulus parameterization alorithm to the regular forecast model, which restructures the vertical humidity distribution and constrains the large-scale model’s moisture error growth during the model integration. An improved short-range rainfall prediction skill was achieved from the modified FSUGSM in this study. The results showed a better agreement between model-based and observed rainfall intensity and pattern. Received January 18, 1999     

A Bogus Typhoon Scheme and Its Application to a Movable Nested Mesh Model

ＡＢｏｇｕｓＴｙｐｈｏｏｎＳｃｈｅｍｅａｎｄＩｔｓＡｐｐｌｉｃａｔｉｏｎｔｏａＭｏｖａｂｌｅＮｅｓｔｅｄＭｅｓｈＭｏｄｅｌＷａｎｇＧｕｏｍｉｎ（王国民）（ＤｅｐａｒｔｍｅｎｔｏｆＡｔｍｏｓｐｈｅｒｉｃＳｃｉｅｎｃｅｓ，ＮａｍｉｎｇＵｎｉｖｅｒｓｉｔｙ...     

Adaptive observations for hurricane prediction

Summary This study proposes a method that can be used to provide guidelines to aircraft reconnaissance for hurricane observations. The method combines numerical weather prediction (NWP) model with a statistical approach to target adaptive observations over areas where the hurricane predictions are very sensitive to the initial analysis for the NWP-model. A single model experiment is performed using regular initial analysis, while 50 other ensemble runs are performed from randomly perturbed initial states. Under the perfect model assumption, the single model experiment serves as a true state. The method first computes the forecast error variances at a certain verification time, e.g. hour 48, and then locates the maximum centers of variances. After the locations of the maximum forecast error variances are known, various correlations of different variables between these maximum variance points and the perturbation fields at the target time, e.g. hour 12, are calculated to identify those locations at the target time, over where the observational errors might be responsible for the growth of forecast error variances at the verification time. Statistically, these correlation fields indicate where the most sensitive areas are at the target time, i.e. where the need for additional observations is suggested. Hurricane Fran of 1996 is used to test the proposed method. The reason for choosing this case is that, during the first 48 hour forecast, the track forecast from NWP-model was very close to the best track. Two additional experiments were designed to examine the method. One experiment updates predicted variables at the target time (12 h) over the areas, to where the proposed method indicates the forecast would be sensitive. The updating combines observations (or truth) with the first guess (predicted) fields. Another experiment also modifies predicted variables at the target time (12 h), but over the areas where the method indicates the forecast errors are less correlated to. The results show that the modification has greatly reduced the forecast error variances at the verification time (48 h) in the first experiment, however it has a very little impact on the variance fields at the forecast hour (48 h) in the second experiment. It is very clear from our experiments, that the proposed method is able to identify sensitive areas, where additional observations can help to reduce hurricane forecast errors from an NWP-model. Received July 19, 1999 Revised November 28, 1999     

热带气旋的路径及登陆预报

用几个非线性数学模型制作热带气旋短期路径预报及热带气旋个数、登陆时段、地段的短期气候预报。5年多的研究和预报试验结果表明：用指数曲线模型制作热带气旋路径预报，准确率较高。24h预报，199次平均误差123km，达到国内先进水平。用多项式等非线性模型，制作登陆我国及登陆广东热带气旋的年、月个数预测，经过3年实际应用检验，准确率达到70％～90％。用非线性预测模型的逐日气压场、逐日雨量场长期预测结果进行分析，制作广东热带气旋登陆时段、地段和南海海面热带气旋出现时间的预报，准确率达到70％～80％，2002年热带气旋的预报，采用长中短期预报相结合，数值预报与统计预报相结合，预报效果较佳。     

Anomalous gradient winds in the subtropical jet stream and interpretations of forecast failures

Summary Dramatic examples of forecast failures in global models of moderate resolution (i.e., T106) have been shown to occur during periods of the negative phase of the Pacific/North American (PNA) pattern in the Northern Hemisphere winter. Specifically, in these periods forecast skills at 500hPa as measured by the standard anomaly correlation index dropped to rather low values by days 4 and 5 of the forecasts. This paper examines systematically some of the factors that may have contributed to the failure of these model forecasts.In particular, strong winds approaching intensities on the order of 100ms^–1 south of Japan at the 200hPa level were degraded by the initialization and data assimilation procedures of the models. These observed winds were found to be supergradient in nature and representative of the anomalous solution of the gradient wind equation. Procedures such as the multivariate optimum interpolation (with its geostrophic constraints) and the normal-mode initialization including several vertical modes apparently were factors that led to the degradation of these strong winds in the initial model states. In this paper, an analysis of these factors is presented, and it is shown that uninitialized analyses (with no constraints) based on a simple successive correction procedure can retain the strong winds evident in the observations. Forecasts thus performed appear to retain wave trains, a characteristic feature of negative PNA initial states, leading to a significant improvement in forecast skill.     

Impact of cloud microphysical processes on hurricane intensity, part 1: Control run

Summary The first part of the paper comprises of a control experiment and its forecasts validations with the observed. The PSU-NCAR mesoscale model MM5 was utilized at a horizontal resolution of 4 km using the data sets for hurricane Charley of 2004. The model configures some of the best available versions for physics and microphysical parameterizations schemes to produce forecasts which are close to the observed trend of hurricane Charley. The basic validations of the control run were carried out in terms of track, intensity (sea-level pressure and surface wind speed), storm propagation speed, precipitation and radar reflectivity with that of observed. The validations were necessary because this control experiment will be considered as a benchmark forecast for comparison with other microphysics sensitive experiments forecasts in the second part of this paper. In general, the control run forecasts closely comply with that of observed track and intensity of the hurricane Charley. We also note that control run manage to reproduce much of the important structural characteristics features of the hurricane as observed.     

An ensemble forecast of the South China Sea monsoon

1.IntroductionThispaperexploresanensembleforecaststrategyforthelarge--scaletropicalpredictionproblem.Thisisgeneralizedfromarecentstudyontheuseofempiricalorthogonalfunction(EOF)--basedperturbationsforhurricanetrackensembleforecasts,(ZhangandKrishnamur...     

A Relocation-based Initialization Scheme to Improve Track-forecasting of Tropical Cyclones简

A relocation procedure to initialize tropical cyclones was developed to improve the representation of the initial conditions and the track forecast for Panasonic Weather Solutions Tropical Operational Forecasts. This scheme separates the vortex perturbation and environment field from the first guess, then relocates the initial vortex perturbations to Lhe observed position by merging them with the environment field. The relationships of wind vector components with stream function and velocity potential are used for separating the vortex disturbance from first guess. For the separation of scalars, a low-pass Barnes filter is employed. The irregular-shaped relocation area corresponding to the specific initial conditions is determined by mapping the edge of the vortex radius in 36 directions.Then, the non-vortex perturbations in the relocation area are removed by a two-pass Barnes filter to retain the vortex perturbations, while the variable fields outside the perimeter of the modified vortex are kept ide.ntical to the original first guess. The potential impacts of this scheme on track forecasts were examined for three hurricane cases in the 2011-12 hurricane season. The experimental results demonstrate that the initialization scheme is able to effectively separate the vortex field from the environment field and maintain a relatively balanced and accurate relocated first guess. As the initial track error is reduced, the following track forecasts are considerably improved. The 72-h average track forecast error was redu,~ed by 32.6% for the cold-start cases, and by 38.4% when using the full-cycling data assimilation because of the accumulatedL improvements from the initialization scheme.     

Upper-air circulation of the Southern Oscillation from the NCEP-NCAR Reanalysis

Summary ?The upper-air circulation characteristics of the Southern Oscillation (SO) are studied from the 1958–97 NCEP-NCAR Reanalysis. The low/high phase of the SO is defined by anomalously small/large values of the Tahiti minus Darwin surface pressure difference. For the fields of 200 mb topography, divergence and divergent flow, 500 mb vertical motion, and 1000 mb topography, during January, April, July and October, differences are computed between ensembles of years of extremely low versus high SO phase, and tested for statistical significance. The 1000 mb topography bears out SO characteristics consistent with earlier work on surface pressure, such as the seesaw between the western and eastern Pacific in all seasons, and in the low SO phase in the boreal winter a northward displaced near-equatorial trough over the Atlantic. In addition, the new dataset afforded documentation of functional upper-air processes. Thus, overall positive/negative 1000 mb topography anomalies tend to be associated with enhanced/reduced subsidence or reduced/enhanced ascending motion. Ensemble differences in mid-tropospheric vertical motion are coherent with those in upper-tropospheric divergence and divergent flow. Characteristic of the low SO phase are significant positive anomalies of upper-tropospheric topographies especially in the tropics and most prominently over the eastern Pacific, reflecting the anomalously warm atmospheric column, and during the boreal winter a train of centers with alternatingly positive and negative topography differences extending from the tropical Pacific across North America to the Atlantic. Received October 10, 2001; revised May 8, 2002; accepted May 31, 2002     

用数字滤波方法进行数值模式的初始化

将数字滤波原理用于Ｔ４２Ｌ９全球谱模式的初始化。选取两个初值进行了实验。通过对模式短时间积分得到的时间序列进行滤波处理，有效地滤去了初始场中的高频重力波振荡，保留了其中有天气意义的扰动，同时对分析场的改变很小。５ｄ预报的结果说明，经过初始化后的预报较为平稳，场较为光滑。对比实验表明，绝热和非绝热初始化的效果很接近。这种方法简便有效，是一种实用的初值化方法     

Assimilation of screen-level observations by variational soil moisture analysis

Summary Inaccurate specification of soil moisture contents can result in forecast errors up to several degrees centigrade. Since direct measurements are rarely available, a variational method has been developed that assimilates synoptic measurements of 2 m-temperature in order to specify the moisture contents of the two soil layers of the Local Model at Deutscher Wetterdienst. The analyzed values minimize a cost functional that expresses the differences between model forecast and observed screen-level temperatures. The minimization is performed highly efficiently and only two additional forecasts are required but neither tangent linear nor adjoint. Background state and background error covariance matrix are updated at each analysis step in a Kalman-filter-like cycled scheme, which takes a model error into account. The soil moisture assimilation shows improved 2 m-temperature forecasts in case of high radiative forcing by up to 3 °C for small areas in the presented 6-week trial run. It proved stability and robustness for general weather conditions and has become operational at DWD for the LM on 14 March 2000. Received August 21, 2000 Revised March 26, 2001     

The application of non‐linear normal mode initialization to an operational forecast model

Abstract

The non‐linear normal mode initialization technique used in shallow water equation models by Baer (1977) and Machenhauer (1977) is now applied to a full baroclinic primitive equations forecast model. The initialization procedure is shown to be capable of completely removing high frequency oscillations from model integrations, even in the presence of topography. The procedure also produces a consistent and physically realistic initial vertical motion field.     

Vortex initialization for typhoon track prediction

Summary A shallow water single level primitive equation model is ideally suited for studying the motion of a tropical cyclone. Three factors seem to be important in the initialization, i.e. the size, intensity and the initial speed and direction of motion of the storm. This study presents the results of sensitivity studies on the above parameters in the definition of a synthetic idealized vortex. The sensitivity studies include results of experimental forecasts for typhoons Betty of 1987 and Dan of 1989. The results of these studies show that the initial size, intensity and direction and speed of motion show considerable sensitivity to the predicted track. Finally a summary of the track forecast errors through 72 hours are presented for these storms.With 8 Figures     

全球数值模式中的台风初始化Ⅱ: 业务应用

由于缺少大量有效的观测资料,台风初始化对数值天气预报业务模式而言,仍然是一个悬而末决的难题.中国国家气象中心自从1996年将台风数值预报系统投入业务运行以来,一直使用经验的人造bogus涡旋台风初始化技术.实际上,不同时期的台风有着不同的环流结构,即使同一个台风在不同的生命期也具有不同的结构特征,而这些结构特征的差异并不能依靠现有的bogus涡旋技术体现出来,这种主观方法的统一性与台风在时空上的差异性形成了强烈的反差.最近,基于国家气象中心全球资料分析同化-预报循环系统,设计和发展了一套新的台风初始化业务方案,它主要由初始涡旋形成、涡旋重定位和涡旋调整3部分过程组成.相比于业务中使用的人造bogus涡旋台风初始化方案,新方案在很大程度上减少了人为因素对台风涡旋结构的影响,而更多地是依靠数值模式自身的动力和物理过程来协调约束产生三维空间的涡旋结构.应用新方案,文中对生成于西北太平洋的2006年0605号台风格美(Kaemi)进行了数值试验,初步分析表明,新方案在实现台风涡旋环流结构的初始化方面效果较好,同时,对台风格美多个时次的预报结果也显示,相比于业务使用的bogus方案而言,新方案对台风路径平均预报误差有了大幅度的降低.     

Reduction of forecast error for global numerical weather prediction by The Florida State University (FSU) Superensemble

Summary The skill of the FSU Superensemble technique as applied to global numerical weather prediction is evaluated extensively in this paper. The global mass and motion fields for year 2000 and precipitation over the domain 55S to 55N for year 2001, as predicted by the Superensemble, the ensemble member models, and the mean of the ensemble members, are evaluated by standard statistical measures of skill to determine the performance of the Superensemble in relation to the other models. The member models are global forecast models from 5 of the worlds operational forecast centers in addition to the FSU global spectral model. For precipitation 5 additional versions of the FSU global model are utilized in the ensemble, as defined by different initial conditions provided by various physical initialization algorithms. Statistical parameters calculated for the mass and motion fields include root mean square (RMS) error, systematic error (or bias), and anomaly correlation. These are applied to the mean sea level pressure, 500hPa heights, and the wind fields at 850hPa and 200hPa. Statistical parameters that were calculated for precipitation include RMS error, correlation, equitable threat score (ETS), and a special definition of bias appropriate for the precipitation field. For the mass and motion fields the performance of the Superensemble was considered for the annual global case, as well as for each hemisphere (north and south) and for each of the four seasons. For precipitation only the annual case was considered over the domain cited above.For the mass and motion fields the RMS calculations showed the Superensemble to be superior (to have the smallest total forecast error) in all comparisons to the ensemble member models, and to be superior to the ensemble mean in the vast majority of comparisons. Performance in comparison to the other models was generally better in the Southern Hemisphere than in the Northern Hemisphere, and better in the transition seasons of fall and spring than in the extreme seasons of winter and summer. The Superensemble had the best success with mean sea level pressure, followed in order by 500hPa geopotential heights, 850hPa winds, and 200hPa winds.In the calculations of 500hPa geopotential height anomaly correlation the Superensemble had higher scores in all comparisons to the ensemble member models, as well as higher scores in the majority of comparisons to the ensemble mean. As with the RMS error results, the Superensemble performed better in the Southern Hemisphere than in the Northern Hemisphere, and better in fall than in summer, in comparison to the other models. The superior anomaly correlation scores of the Superensemble attest to the ability of the model to forecast daily perturbations from the climatological means, perturbations that are associated with transient synoptic scale features, given the horizontal resolution in the forecast models.In terms of systematic error reduction the Superensemble produces its most impressive results. Annual global mean sea-level pressure systematic errors for day 5 forecasts are generally in the range of ±1hPa (compared to errors as high as 8hPa in other models), and day 2 forecasts of 500hPa geopotential height produced systematic errors generally in the range of ±10 meters (compared to errors as high as 60 meters in other models). The Superensemble was able to reduce systematic errors in forecasts of a variety of important features in the global mass and motion fields: surface equatorial trough, wave amplitude in geopotential heights at 500hPa, trade winds and Somali Jet at 850hPa, mid-latitude westerlies, subtropical jet, and Tropical Easterly Jet (TEJ) at 200hPa.In terms of forecasting precipitation the Superensemble outperforms all ensemble member models and the ensemble mean in terms of RMS error, correlation coefficient, equitable threat score, and bias. The superior correlation scores indicate that the Superensemble is more reliable than the other models in predicting perturbations in the area distribution of precipitation, perturbations that are essentially associated with migrant synoptic scale disturbances, considering the horizontal resolution of the forecast models.The Superensemble is a valuable tool for significantly improving upon the global model forecasts of the worlds operational forecast centers. These forecasts are used daily as important guidance in making weather forecasts in all regions of the world. This paper will demonstrate that the Superensemble improves upon the ensemble member model forecasts: (1) in a statistical sense considering broad areas of the globe, (2) in a synoptic climatology sense through focus on the improved forecasts of climatological features seen in the global mass and motion fields, (3) in a synoptic sense through use of anomaly correlation and correlation coefficient where improvement is demonstrated in the forecasts of perturbations from mean fields which are essentially associated with transient synoptic scale disturbances.