Best-track Data Research Articles

A study on radial characteristics of North Indian Ocean tropical cyclones and associated energy indices through numerical modeling

The structural characteristics of North Indian Ocean (NIO) Tropical Cyclones (TCs) are emphasized in this study by adopting Weather Research and Forecasting (WRF) numerical modeling framework. TCs that occurred during 2001–2020, were simulated using WRF model in the CTRL, and scatterometer wind vortices from four satellites were considered for assimilation purpose in the WRFDA experiment. Considering the results from both types of simulations, best-track data sets, and ERA5 global analyses, several aspects including radial parameters, the thermodynamic and dynamic properties, horizontal structure, etc., were analyzed besides developing some robust empirical relationships. In the process, model performance was evaluated too. The study emphasized on sub-basin, sectorial, intensity-based, range-based, and frequency-based analysis for the radial parameters. Also, yearly, and seasonal variations were considered, inter-comparisons were carried out, the inter-relationships with numerous environmental and storm-related characteristics have been explored. The study provides valuable insights into the TC radial characteristics over NIO, and associated thermodynamic and dynamic behavior. It upholds that the relatively greater intensity TCs occurring during October–November months are mostly observed to have genesis over the East Arabian Sea (EAS), Middle Bay of Bengal (MBOB), and North Bay of Bengal (NBOB) sectors. The EAS region exhibits higher values for the energy indices and maximum sustained wind (MSW), compared to the Western Arabian Sea (WAS), while the maximum convective potential values show an opposite trend. The assimilation of scatterometer wind data has a limited impact on model results, although it is more effective over the Arabian Sea (AS) than the Bay of Bengal (BOB). WRFDA could provide a relatively better estimate for most of the TC radial features over both AS and BOB. The extensive analysis of TC radial features facilitated developing several empirical relationships involving radial characteristics, MSW and minimum central pressure, which are valid for NIO region.

The Effects of Upper-Ocean Sea Temperatures and Salinity on the Intensity Change of Tropical Cyclones over the Western North Pacific and the South China Sea: An Observational Study

With increasing air and sea temperatures, the thermodynamic environments over the oceans are becoming more favourable for the development of intense tropical cyclones (TCs) with rapid intensification (RI). The South China coastal region consists of highly densely populated cities, especially over the Pearl River Delta (PRD) region. Intense TCs maintaining their strength or the RI of TCs close to the coastal region can present substantial forecasting challenges and have significant potential impacts on the coastal population. This study investigates the effect of sea-surface and sub-surface temperatures and salinity on the intensification of five TCs, namely Super Typhoon Hato in 2017, Super Typhoon Mangkhut in 2018, and Typhoon Talim, Super Typhoon Saola, and Severe Typhoon Koinu in 2023, which have significantly affected the South China coastal region and triggered high TC warning signals in Hong Kong in the past few years. This analysis utilised the Hong Kong Observatory’s TC best-track and intensity data, along with sea temperature and salinity profiles generated using the China Ocean ReAnalysis version 2 (CORA2) product from the National Marine Data and Information Service of China. It was found that high sea-surface temperatures (SST) of 30 °C or above for a depth of about 20 m, low sea-surface salinity (SSS) levels of 33.8 psu or below for a depth of at least 20 m, and strong salinity stratification of at least 0.6 psu per 100 m depth might offer useful hints for predicting the RI of TCs over the western North Pacific and the South China Sea (SCS) in operational forecasting, while noting other contributing environmental factors and synoptic flow patterns conducive to RI. This study represents the first documentation of sub-surface salinity’s impact on some intense TCs traversing the SCS during 2017–2023 based on an observational study. Our aim is to supplement operational techniques for forecasting RI with some quantitative guidance based on upper-level ocean observations of temperatures and salinity, on top of well-known but more rapidly changing dynamical factors like low-level convergence, weak vertical wind shear, and upper-level divergent outflow, as forecasted with numerical weather prediction models. This study will also encourage further research to refine the analysis of quantitative contributions from different RI factors and the identification of essential features for developing AI models as one way to improve the forecasting of TC RI before the TC makes landfall near the PRD, with due consideration given to the effect of freshwater river discharge from the Pearl River.

Key ingredients in regional climate modelling for improving the representation of typhoon tracks and intensities

Abstract. There is evidence of an increased frequency of rapid intensification events of tropical cyclones (TCs) in global offshore regions. This will not only result in increased peak wind speeds but may lead to more intense heavy precipitation events, leading to flooding in coastal regions. Therefore, high impacts are expected for urban agglomerations in coastal regions such as the densely populated Pearl River Delta (PRD) in China. Regional climate models (RCMs) such as the Weather Research and Forecasting (WRF) model are state-of-the-art tools commonly applied to predict TCs. However, typhoon simulations are connected with high uncertainties due to the high number of parameterization schemes of relevant physical processes (including possible interactions between the parameterization schemes) such as cumulus (CU) and microphysics (MP), as well as other crucial model settings such as domain setup, initial times, and spectral nudging. Since previous studies mostly focus on either individual typhoon cases or individual parameterization schemes, in this study a more comprehensive analysis is provided by considering four different typhoons of different intensity categories with landfall near the PRD, i.e. Typhoon Neoguri (2008), Typhoon Hagupit (2008), Typhoon Hato (2017), and Typhoon Usagi (2013), as well as two different schemes for CU and MP, respectively. Moreover, the impact of the model initialization and the driving data is studied by using three different initial times and two spectral nudging settings. Compared with the best-track reference data, the results show that the four typhoons show some consistency. For track bias, nudging only horizontal wind has a positive effect on reducing the track distance bias; for intensity, compared with a model explicitly resolving cumulus convection, i.e. without cumulus parameterization (CuOFF; nudging potential temperature and horizontal wind; late initial time), using the Kain–Fritsch scheme (KF; nudging only horizontal wind; early initial time) configuration shows relatively lower minimum sea level pressures and higher maximum wind speeds, which means stronger typhoon intensity. Intensity shows less sensitivity to two MP schemes compared with the CuOFF, nudging, and initial time settings. Furthermore, we found that compared with the CuOFF, using the KF scheme shows a relatively larger latent heat flux and higher equivalent potential temperature, providing more energy to typhoon development and inducing stronger TCs. This study could be used as a reference to configure WRF with the model's different combinations of schemes for historical and future TC simulations and also contributes to a better understanding of the performance of principal TC structures.

Reexamining the Estimation of Tropical Cyclone Radius of Maximum Wind from Outer Size with an Extensive Synthetic Aperture Radar Dataset

Abstract The radius of maximum wind Rmax, an important parameter in tropical cyclone (TC) ocean surface wind structure, is currently resolved by only a few sensors so that, in most cases, it is estimated subjectively or via crude statistical models. Recently, a semiempirical model relying on an outer wind radius, intensity, and latitude was fit to best-track data. In this study we revise this semiempirical model and discuss its physical basis. While intensity and latitude are taken from best-track data, Rmax observations from high-resolution (3 km) spaceborne synthetic aperture radar (SAR) and wind radii from an intercalibrated dataset of medium-resolution radiometers and scatterometers are considered to revise the model coefficients. The new version of the model is then applied to the period 2010–20 and yields Rmax reanalyses and trends that are more accurate than best-track data. SAR measurements corroborate that fundamental conservation principles constrain the radial wind structure on average, endorsing the physical basis of the model. Observations highlight that departures from the average conservation situation are mainly explained by wind profile shape variations, confirming the model’s physical basis, which further shows that radial inflow, boundary layer depth, and drag coefficient also play roles. Physical understanding will benefit from improved observations of the near-core region from accumulated SAR observations and future missions. In the meantime, the revised model offers an efficient tool to provide guidance on Rmax when a radiometer or scatterometer observation is available, for either operations or reanalysis purposes.

Simulation of Storm Surge Heights Based on Reconstructed Historical Typhoon Best Tracks Using Expanded Wind Field Information

A numerical model integrating tides, waves, and surges can accurately evaluate the surge height (SH) risks of tropical cyclones. Furthermore, incorporating the external forces exerted by the storm’s wind field can help to accurately reproduce the SH. However, the lack of long-term typhoon best track (BT) data degrades the SH evaluations of past events. Moreover, archived BT data (BTD) for older typhoons contain less information than recent typhoon BTD. Thus, herein, the wind field structure, specifically its relationship with the central air pressure, maximum wind speed, and wind radius, are augmented. Wind formulae are formulated with empirically adjusted radii and the maximum gradient wind speed is correlated with the central pressure. Furthermore, the process is expanded to four quadrants through regression analyses using historical asymmetric typhoon advisory data. The final old typhoon BTs are converted to a pseudo automated tropical cyclone forecasting format for consistency. Validation tests of the SH employing recent BT and reconstructed BT (rBT) indicate the importance of the nonlinear interactions of tides, waves, and surges for the macrotidal west and microtidal south coasts of Korea. The expanded wind fields—rBT—based on the historical old BT successfully assess the return periods of the SH. The proposed process effectively increases typhoon population data by incorporating actual storm tracks.

A CatBoost-Based Model for the Intensity Detection of Tropical Cyclones over the Western North Pacific Based on Satellite Cloud Images

A CatBoost-based intelligent tropical cyclone (TC) intensity-detecting model was built to quantify the intensity of TCs over the Western North Pacific (WNP) with the cloud-top brightness temperature (CTBT) data of Fengyun-2F (FY-2F) and Fengyun-2G (FY-2G) and the best-track data of the China Meteorological Administration (CMA-BST) in recent years (2015–2018). The CatBoost-based model was featured with the greedy strategy of combination, the ordering principle in optimizing the possible gradient bias and prediction shift problems, and the oblivious tree in fast scoring. Compared with the previous studies based on the pure convolutional neural network (CNN) models, the CatBoost-based model exhibited better skills in detecting the TC intensity with the root mean square error (RMSE) of 3.74 m s−1. In addition to the three mentioned model features, there are also two reasons for the model’s design. On one hand, the CatBoost-based model used the method of introducing prior physical factors (e.g., the structure and shape of the cloud, deep convections, and background fields) into its training process. On the other hand, the CatBoost-based model expanded the dataset size from 2342 to 13,471 samples through hourly interpolations of the original dataset. Furthermore, this paper investigated the errors of this model in detecting the different categories of TC intensity. The results showed that the deep learning-based TC intensity-detecting model proposed in this paper has systematic biases, namely, the overestimation (underestimation) of intensities in TCs which were weaker (stronger) than at the typhoon level, and the errors of the model in detecting weaker (stronger) TCs were smaller (larger). This implies that more factors than the CTBT should be included to further reduce the errors in detecting strong TCs.

The effects of intraseasonal oscillations on landfalling tropical cyclones in the Philippines during the boreal summer

As a kind of weather phenomenon with destructive wind and heavy rainfall, tropical cyclones (TCs), especially the landing TCs, can cause severer economic damage and losses of life. Philippines is one of the countries mostly affected by Tropical cyclones (TCs). Based on the best-track TC data and global atmospheric and oceanic reanalysis data, the present paper investigates the isolated and combined effects of two intraseasonal oscillations, the Madden-Julian oscillation (MJO) and the quasi-biweekly oscillation (QBWO), on landfall of TCs in the Philippines during boreal summer (May-September) in 1979-2019. The results show that both the MJO and the QBWO can significantly affect the frequency, landfall intensity, location, and translation speed of TCs that make landfall in the Philippines. During the convective (non-convective) phases of the MJO and the QBWO, more (less) frequent and stronger (weaker) TCs make landfall in the Philippines. This is due to the increased (decreased) frequency of TCs formation in the NWP and environmental factors in the region east of the Philippines that are favorable (unfavorable) for the development of TCs. With the northward propagation of the convective signals of the MJO and QBWO, the Western Pacific Subtropical High (WPSH) shifts eastward, and the steering flow is unfavorable for westward movement of NWP TCs. This, in turn, causes a northward shift in the landfall locations and a decrease in the translation speed of TCs. These results are helpful for the prediction of the TCs affecting the Philippines.

Analysis of Extreme Rainstorms Associated with Tropical Cyclone Rammasun (2014) over Guangxi Province of China

The tropical cyclone (TC) Rammasun (1409) successively caused three extreme rainstorms centered in Hepu, Fangcheng, and Lingshan in Guangxi. In this study, a set of datasets, including the China Meteorological Administration (CMA) TC best-track data, rain gauge stations, radar products, and the latest ERA5 reanalysis, were used to investigate the spatial–temporal characteristics and causes of the three rainstorms. Overall, there are apparent discrepancies among them regarding the triggering and maintenance mechanisms. Prior to its landfall in Guangxi, the rainstorm surge (hourly precipitation 136.9 mm) around Hepu was triggered by the approaching low-level jet and northward intrusion of convective instability in the context of abundant vapor supply and vigorous convection. At the weakening stage, Rammasun gradually crossed over the Shiwan Mountains, with a strong convection zone in the western eyewall moving to Fangcheng, which enhanced the precipitation there under the help of topographic uplift. The friction effect slowed down the TC, and contributed to accumulation of the precipitation, which, however, was somewhat suppressed (hourly precipitation less than 50 mm) by the low-level stratification. In the decaying stage, the Lingshan rainstorm was triggered under the combined efforts of the high-level divergence due to the expansion of the South Asian high and the surface convergence between the southeasterly and southerly winds, accompanied by the westward expansion of the high-energy zone. Finally, the corresponding conceptual models for the three rainstorms are summarized and should provide important implications to both research and forecasting for TC-related heavy precipitation.

A Deep Learning Model for Estimating Tropical Cyclone Wind Radius from Geostationary Satellite Infrared Imagery

Abstract This article developed a deep learning (DL) model for estimating tropical cyclone (TC) 34-, 50-, and 64-kt (1 kt ≈ 0.51 m s−1) wind radii in four quadrants from infrared images in the global ocean. We collected 63 675 TC images from 2004 to 2016 and divided them into three periods (2004–12, 2013–14, and 2015–16) for model training, validation, and testing. First, four DL-based radius estimation models were developed to estimate the TC wind radius for each of the four quadrants. Then, the entire original images and the one-quarter-quadrant subimages were included in the model training for each quadrant. Last, we modified the mean absolute error (MAE) loss function in these DL-based models to reduce the side effect of an unbalanced distribution of wind radii and developed an asymmetric TC wind radius estimation model globally. The comparison of model results with the best-track data of TCs shows that the MAEs of 34-kt wind radius are 18.8, 19.5, 18.6, and 18.8 n mi (1 n mi = 1.852 km) for the northeast, southeast, southwest, and northwest quadrants, respectively. The MAEs of 50-kt wind radius are 11.3, 11.3, 11.1, and 10.8 n mi, respectively, and the MAEs of 64-kt wind radius are 8.9, 9.9, 9.2, and 8.7 n mi, respectively. These results represent a 12.1%–35.5% improvement over existing methods in the literature. In addition, the DL-based models were interpreted with two deep visualization toolboxes. The results indicate that the TC eye, cloud, and TC spiral structure are the main factors that affect the model performance.

The relationship between pre-landfall intensity change and post-landfall weakening of tropical cyclones over China

The accurate prediction of the weakening of landfalling tropical cyclones (TC) is of great importance to the disaster prevention but is still challenging. In this study, based on the 6-hourly TC best-track data and global reanalysis data, the relationship between the intensity change prior to landfall of TCs and the energy dissipation rate after landfall over mainland China is statistically analyzed, and the difference between East and South China is compared. Results show that TCs making landfall over East China often experienced pre-landfall weakening and usually corresponded to a rapid decay after landfall, while most TCs making landfall over South China intensified prior to landfall and weakened slowly after landfall. The key factors affecting both pre-landfall intensity change and post-landfall energy dissipation rate are quantitatively analyzed. It is found that the decreasing sea surface temperature (SST), increasing SST gradient, and increasing environmental vertical wind shear are the major factors favoring high pre-landfall weakening occurrence, leading to rapid TC weakening after landfall over East China. In South China, changes in the large-scale environmental factors are relatively small and contribute little to the post-landfall weakening rate.

Uncertainties in tropical cyclone landfall decay

Understanding the responses of landfalling tropical cyclones to a changing climate has been a topic of great interest and research. Among them, the recently reported slowdown of tropical cyclone landfall decay in a warming climate engenders controversy. Here, the global climatology of landfall decay, based on the tropical cyclone best-track data available, reveals that the reported trends are uncertain and not universal, but spatial, temporal, data, and methodology dependent such that any claim of a climate trend could be misleading at present. The effective area of moisture supply from the ocean, most likely determined by the landfalling track modes, is demonstrated to be an important factor for the decay. This study provides timely essential clarifications of the current contentious understanding.

Application of Microwave Space-Based Environmental Monitoring (SBEM) Data for Operational Tropical Cyclone Intensity Estimation at the Joint Typhoon Warning Center

Abstract The Joint Typhoon Warning Center (JTWC) utilized new space-based environmental monitoring (SBEM) data alongside traditional data to adjust JTWC tropical cyclone (TC) intensity and structure estimates during production of the official 2019 Best Track dataset. Intensity estimates from multiple platforms such as Advanced Microwave Scanning Radiometer-2 (AMSR2), the Soil Moisture Active Passive (SMAP) and Soil Moisture and Ocean Salinity (SMOS) radiometers, and synthetic aperture radar (SAR), along with objective Dvorak and satellite consensus algorithms, not only aided the poststorm best track (BT) process, but also provided robust data that supported real-time analysis and forecasting. This summary attempts to communicate with the TC community the extent to which these new data affected the 2019 official BT data, how JTWC utilized these new data in the poststorm BT process, and provide examples of how these data influenced forecaster decision-making in real time. This paper makes no attempt to validate the accuracy of the wind speed estimates from these methods (SAR, SMAP/SMOS, or AMSR2) and does not outline the entirety of the JTWC process for determining TC intensity, but it does outline, briefly, the impact of these new datasets on the final JTWC BT intensity estimates and on real-time analysis. These methodologies are valuable sources of cyclone intensity estimates in an otherwise data-sparse area of responsibility, and in many cases provide critical data not captured by traditional methods alone, which are detailed further in this summary.

Machine Learning–Based Hurricane Wind Reconstruction

Abstract Here we present a machine learning–based wind reconstruction model. The model reconstructs hurricane surface winds with XGBoost, which is a decision-tree-based ensemble predictive algorithm. The model treats the symmetric and asymmetric wind fields separately. The symmetric wind field is approximated by a parametric wind profile model and two Bessel function series. The asymmetric field, accounting for asymmetries induced by the storm and its ambient environment, is represented using a small number of Laplacian eigenfunctions. The coefficients associated with Bessel functions and eigenfunctions are predicted by XGBoost based on storm and environmental features taken from NHC best-track and ERA-Interim data, respectively. We use HWIND for the observed wind fields. Three parametric wind profile models are tested in the symmetric wind model. The wind reconstruction model’s performance is insensitive to the choice of the profile model because the Bessel function series correct biases of the parametric profiles. The mean square error of the reconstructed surface winds is smaller than the climatological variance, indicating skillful reconstruction. Storm center location, eyewall size, and translation speed play important roles in controlling the magnitude of the leading asymmetries, while the phase of the asymmetries is mainly affected by storm translation direction. Vertical wind shear impacts the asymmetry phase to a lesser degree. Intended applications of this model include assessing hurricane risk using synthetic storm event sets generated by statistical–dynamical downscaling hurricane models.

On the Two Types of Tropical Cyclone Eye Formation: Clearing Formation and Banding Formation

Abstract Two types of eye formation are proposed: “clearing formation (CF)” and “banding formation (BF)”. The objectives are to identify the tropical cyclone (TC) characteristics associated with these two types of eye formation and to clarify how the environment, pattern of synoptic systems and TC structure during initial and development stages affect the eye formation. The satellite imagery and best-track data are used to classify and analyze the TCs named in the western North Pacific from 2007 to 2020. The results show that TCs with CF have significantly higher intensity and intensification rate during the period of the first eye presence, smaller size after genesis and prior to eye formation, smaller eye size when the eye forms and a more westward track. It is noted that CF and BF TCs tend to occur in autumn and summer, respectively. Meanwhile, the composite results using reanalysis data show that the CF TCs are generally characterized by easterly-wave features with a dryer environment, smaller initial size and larger radial gradient of vorticity, while the BF TCs feature a monsoon-depression structure with a wetter environment, larger initial size and flatter vorticity profile. A conceptual hypothesis is thus proposed, as compared to BF TCs, the smaller size and weaker outer wind in CF storms associated with the easterly-wave disturbance are facilitated by inactive outer convection, leading to larger radial gradient of inertial stability. The low-level inflow can penetrate inward close to the center, resulting in more diabatic heating inside the radius of maximum wind with much higher heating efficiency, and also a higher intensification rate.

A New Type of Red-Green-Blue Composite and Its Application in Tropical Cyclone Center Positioning

Weak tropical cyclone (TC) center positioning is difficult work in operational forecasting. In the present study, a TC-red-green-blue (TC-RGB) composite was designed by using satellite multichannel observations (reflectance, brightness temperature, and brightness temperature differences). Compared with single channel images, TC-RGB composites can clearly show the exposed low-level circulation (LLC) of weak TCs under large vertical wind shear. Based on the guidelines of TC-RGB composites for TC center positioning, we repositioned 83 western North Pacific (WNP) TC cases during 2017–2019. Then, the comparisons of TC center positions were made between the TC-RGB composite and the Regional Specialized Meteorological Centre-Tokyo (RSMC-Tokyo), the Joint Typhoon Warning Center (JTWC) and the China Meteorological Administration-Shanghai Typhoon Institute (CMA-STI). Via case analysis of TC Kalmaegi (2019), it was found that the best-track data from the RSMC-Tokyo, JTWC and CMA-STI would have over 100 km biases at the early stage of TC life history. Taking all the 83 TC cases into account, the results show that the average center position biases and standard deviations for weak TCs under small vertical wind shear in the daytime are 5 km larger than those under large vertical wind shear at nighttime. When considering the 83 TC cases with clear LLC centers, the difference of these two biases is 10 km. The average biases are mostly above 20 km in the areas south of 18° N and north of 36° N over the WNP. Conversely, in the areas between 18° N and 36° N over the WNP, they are mostly below 20 km.

Changes in Typhoon Regional Heavy Precipitation Events over China from 1960 to 2018

In earlier studies, objective techniques have been used to determine the contribution of tropical cyclones to precipitation (TCP) in a region, where the Tropical cyclone Precipitation Event (TPE) and the Regional Heavy Precipitation Events (RHPEs) are defined and investigated. In this study, TPE and RHPEs are combined to determine the Typhoon Regional Heavy Precipitation Events (TRHPEs), which is employed to evaluate the contribution of tropical cyclones to regional extreme precipitation events. Based on the Objective Identification Technique for Regional Extreme Events (OITREE) and the Objective Synoptic Analysis Technique (OSAT) to define TPE, temporal and spatial overlap indices are developed to identify the combined events as TRHPE. With daily precipitation data and TC best-track data over the western North Pacific from 1960 to 2018, 86 TRHPEs have been identified. TRHPEs contribute as much as 20% of the RHPEs, but 100% of events with extreme individual precipitation intensities. The major TRHPEs continued for approximately a week after tropical cyclone landfall, indicating a role of post landfall precipitation. The frequency and extreme intensity of TRHPEs display increasing trends, consistent with an observed positive trend in the mean intensity of TPEs as measured by the number of daily station precipitation observations exceeding 100 mm and 250 mm. More frequent landfalling Southeast and South China TCs induced more serious impacts in coastal areas in the Southeast and the South during 1990–2018 than 1960–89. The roles of cyclone translation speed and “shifts” in cyclone tracks are examined as possible explanations for the temporal trends.

Relationships between the Southwest Monsoon Surge and the Heavy Rainfall Associated with Landfalling Super Typhoon Rammasun

Based on the China Meteorological Administration (CMA) best-track data, the ERA5 reanalysis, and the Global Precipitation Measurement (GPM) precipitation data, this paper analyzes the reasons for the heavy rainfall event of Super Typhoon Rammasun in 2014, and the results are as follows: (1) Rammasun was blocked by the western Pacific subtropical high (WPSH), the continental high, and the mid-latitude westerly trough. Such a stable circulation pattern maintained the vortex circulation of Rammasun. (2) During the period of landfall, the southwest summer monsoon surge was reinforced due to the dramatic increase of the zonal wind and the cross-equatorial flow near 108° E. The results of the dynamic monsoon surge index (DMSI) and boundary water vapor budget (BWVB) show that the monsoon surge kept providing abundant water vapor for Rammasun, which led to the enhanced rainfall. (3) The East Asian monsoon manifested an obvious low-frequency oscillation with a main period of 20–40 days in the summer of 2014, which propagated northward significantly. When the low-frequency oscillation reached the extremely active phase, the monsoon surge hit the maximum and influenced the circulation of Rammasun. Meanwhile, the convergence and water vapor flux associated with the low-frequency oscillation significantly contributed to the heavy rainfall.

A Novel Deep Learning Model by BiGRU with Attention Mechanism for Tropical Cyclone Track Prediction in the Northwest Pacific

Abstract Tropical cyclones are among the most powerful and destructive meteorological systems on Earth. In this paper, we propose a novel deep learning model for tropical cyclone track prediction method. Specifically, the track task is regarded as a time series predicting challenge, and then a deep learning framework by a bidirectional gate recurrent unit network (BiGRU) with attention mechanism is developed for track prediction. This proposed model can excavate the effective information of the historical track in a deeper and more accurate way. Data experiments are conducted on tropical cyclone best-track data provided by the Joint Typhoon Warning Center (JTWC) from 1988 to 2017 in the northwestern Pacific Ocean. Results show that our model performs well for tracks of 6, 12, 24, 48, and 72 h in the future. The prediction results show that our proposed combined model is superior to state-of-the-art deep learning models, including a recurrent neural network (RNN), long short-term memory neural network (LSTM), gate recurrent unit network (GRU), and BiGRU without the use of attention mechanism. In comparison with the methods used by the China Meteorological Administration, Japan Meteorological Agency, and the JTWC, our method has obvious advantages in the mid- to long-term track forecasting, especially in the next 72 h.

Factors Controlling Tropical Cyclone Intensification Over the Marginal Seas of China

Tropical cyclone (TC) intensification over marginal seas, especially rapid intensification (RI), often poses great threat to lives and properties in coastal regions and is subject to large forecast errors. It is thus important to understand the characteristics of TC intensification and the involved key factors affecting TC intensification over marginal seas. In this study, the 6-hourly TC best-track data from Shanghai Typhoon Institute of China Meteorological Administration, ERA-Interim reanalysis data, and TRMM satellite rainfall products are used to analyze and compare the climatological characteristics and key factors of different intensification stratifications over the marginal seas of China (MSC) and the western North Pacific (WNP) during 1980–2018. The statistical results show that TC intensification over the MSC is more likely to occur when TCs experience relatively large intensities, weak vertical wind shear, small translation perpendicular to the coastline, relatively high fullness, strong upper-level divergence, low-level relative vorticity, and high inner-core precipitation rate. The box difference index method is used to quantify the relative contributions of these factors to TC RI. Results show that the initial (relative) intensity contributes the most to TC RI over both the MSC and the WNP. The inner-core precipitation rate and translation perpendicular to the coastline are of second importance to TC RI over the MSC, while both vertical wind shear and TC fullness are crucial to TC RI over the WNP. These findings may help understand TC activity over the MSC and provide a basis for improving intensity prediction of TCs in the MSC.

The Performance of Three Exponential Decay Models in Estimating Tropical Cyclone Intensity Change After Landfall Over China

In this study, the performance of three exponential decay models in estimating intensity change of tropical cyclones (TCs) after landfall over China is evaluated based on the best-track TC data during 1980–2018. Results indicate that the three models evaluated can reproduce the weakening trend of TCs after landfall, but two of them (M1 and M2) tend to overestimate TC intensity and one (M3) tends to overestimate TC intensity in the first 12 h and underestimate TC intensity afterwards. M2 has the best performance with the smallest errors among the three models within 24 h after landfall. M3 has better performance than M1 in the first 20 h after landfall, but its errors increase largely afterwards. M1 and M2 show systematic positive biases in the southeastern China likely due to the fact that they have not explicitly included any topographic effect. M3 has better performance in the southeastern China, where it was originally attempted, but shows negative biases in the eastern China. The relative contributions of different factors, including landfall intensity, translational speed, 850-hPa moist static energy, and topography, to model errors are examined based on classification analyses. Results indicate that the landfall intensity contributes about 18%, translational speed, moist static energy and topography contribute equally about 15% to the model errors. It is strongly suggested that the TC characteristics and the time-dependent decay constant determined by environmental conditions, topography and land cover properties, should be considered in a good exponential decay model of TC weakening after landfall.