Explore the vast range of titles available.

Accurate forecasting of tropical cyclone (TC) tracks is critical for disaster preparedness and risk mitigation. While traditional numerical weather prediction (NWP) systems have long served as the backbone of operational forecasting, they face limitations in computational cost and sensitivity to initial conditions. In recent years, deep learning (DL) has emerged as a promising alternative, offering data-driven modeling capabilities for capturing nonlinear spatiotemporal patterns. This paper presents a comprehensive review of DL-based approaches for TC track forecasting. We categorize all DL-based TC tracking models according to the architecture, including recurrent neural networks (RNNs), convolutional neural networks (CNNs), Transformers, graph neural networks (GNNs), generative models, and Fourier-based operators. To enable rigorous performance comparison, we introduce a Unified Geodesic Distance Error (UGDE) metric that standardizes evaluation across diverse studies and lead times. Based on this metric, we conduct a critical comparison of state-of-the-art models and identify key insights into their relative strengths, limitations, and suitable application scenarios. Building on this framework, we conduct a critical cross-model analysis that reveals key trends, performance disparities, and architectural tradeoffs. Our analysis also highlights several persistent challenges, such as long-term forecast degradation, limited physical integration, and generalization to extreme events, pointing toward future directions for developing more robust and operationally viable DL models for TC track forecasting. To support reproducibility and facilitate standardized evaluation, we release an open-source UGDE conversion tool on GitHub.

Tropical cyclones have always been a concern of meteorologists, and there are many studies regarding the axisymmetric structures, dynamic mechanisms, and forecasting techniques from the past 100 years. This research demonstrates the ongoing progress as well as the many remaining problems. Machine learning, as a means of artificial intelligence, has been certified by many researchers as being able to provide a new way to solve the bottlenecks of tropical cyclone forecasts, whether using a pure data-driven model or improving numerical models by incorporating machine learning. Through summarizing and analyzing the challenges of tropical cyclone forecasts in recent years and successful cases of machine learning methods in these aspects, this review introduces progress based on machine learning in genesis forecasts, track forecasts, intensity forecasts, extreme weather forecasts associated with tropical cyclones (such as strong winds and rainstorms, and their disastrous impacts), and storm surge forecasts, as well as in improving numerical forecast models. All of these can be regarded as both an opportunity and a challenge. The opportunity is that at present, the potential of machine learning has not been completely exploited, and a large amount of multi-source data have also not been fully utilized to improve the accuracy of tropical cyclone forecasting. The challenge is that the predictable period and stability of tropical cyclone prediction can be difficult to guarantee, because tropical cyclones are different from normal weather phenomena and oceanographic processes and they have complex dynamic mechanisms and are easily influenced by many factors.

This paper describes a novel machine learning (ML) framework for tropical cyclone intensity and track forecasting, combining multiple ML techniques and utilizing diverse data sources. Our multimodal framework, called Hurricast, efficiently combines spatial–temporal data with statistical data by extracting features with deep learning encoder–decoder architectures and predicting with gradient-boosted trees. We evaluate our models in the North Atlantic and eastern Pacific basins in 2016–19 for 24-h lead-time track and intensity forecasts and show they achieve comparable mean absolute error and skill to current operational forecast models while computing in seconds. Furthermore, the inclusion of Hurricast into an operational forecast consensus model could improve upon the National Hurricane Center’s official forecast, thus highlighting the complementary properties with existing approaches. In summary, our work demonstrates that utilizing machine learning techniques to combine different data sources can lead to new opportunities in tropical cyclone forecasting.

The operational Canadian Global Deterministic Prediction System suffers from a weak-intensity bias for simulated tropical cyclones. The presence of this bias is confirmed in progressively simplified experiments using a hierarchical system development technique. Within a semi-idealized, simplified-physics framework, an unexpected insensitivity to the representation of relevant physical processes leads to investigation of the model’s semi-Lagrangian dynamical core. The root cause of the weak-intensity bias is identified as excessive numerical dissipation caused by substantial off-centering in the two time-level time integration scheme used to solve the governing equations. Any (semi)implicit semi-Lagrangian model that employs such off-centering to enhance numerical stability will be afflicted by a misalignment of the pressure gradient force in strong vortices. Although the associated drag is maximized in the tropical cyclone eyewall, the impact on storm intensity can be mitigated through an intercomparison-constrained adjustment of the model’s temporal discretization. The revised configuration is more sensitive to changes in physical parameterizations and simulated tropical cyclone intensities are improved at each step of increasing experimental complexity. Although some rebalancing of the operational system may be required to adapt to the increased effective resolution, significant reduction of the weak-intensity bias will improve the quality of Canadian guidance for global tropical cyclone forecasting.

The National Hurricane Center (NHC) uses a variety of guidance models for its operational tropical cyclone track, intensity, and wind structure forecasts, and as baselines for the evaluation of forecast skill. A set of the simpler models, collectively known as the NHC guidance suite, is maintained by NHC. The models comprising the guidance suite are briefly described and evaluated, with details provided for those that have not been documented previously. Decay-SHIFOR is a modified version of the Statistical Hurricane Intensity Forecast (SHIFOR) model that includes decay over land; this modification improves the SHIFOR forecasts through about 96 h. T-CLIPER, a climatology and persistence model that predicts track and intensity using a trajectory approach, has error characteristics similar to those of CLIPER and D-SHIFOR but can be run to any forecast length. The Trajectory and Beta model (TAB), another trajectory track model, applies a gridpoint spatial filter to smooth winds from the National Centers for Environmental Prediction (NCEP) Global Forecast System (GFS) model. TAB model errors were 10%–15% lower than those of the Beta and Advection model (BAM), the model it replaced in 2017. Optimizing TAB’s vertical weights shows that the lower troposphere’s environmental flow provides a better match to observed tropical cyclone motion than does the upper troposphere’s, and that the optimal steering layer is shallower for higher-latitude and weaker tropical cyclones. The advantages and disadvantages of the D-SHIFOR, T-CLIPER, and TAB models relative to their earlier counterparts are discussed.

The radius of maximum wind (RMW) defines the location of the maximum winds in a tropical cyclone and is critical to understanding intensity change as well as hazard impacts. A comparison between the Hurricane Analysis and Forecast System (HAFS) models and two statistical models based off the National Hurricane Center official forecast is conducted relative to a new baseline climatology to better understand whether models have skill in forecasting the RMW of North Atlantic tropical cyclones. On average, the HAFS models are less skillful than the climatology and persistence baseline and two statistically derived RMW estimates. The performance of the HAFS models is dependent on intensity with better skill for stronger tropical cyclones compared to weaker tropical cyclones. To further improve guidance of tropical cyclone hazards, more work needs to be done to improve forecasts of tropical cyclone structure. Plain Language Summary The radius of maximum wind (RMW) is a key structural parameter of tropical cyclones that describes how far the strongest winds are from the storm's center. The RMW is closely tied to significant hazards such as wind, storm surge, and rainfall. However, little forecast guidance is provided for the RMW resulting in forecasters using climatological estimates to help communicate hazard risk. In order to better forecast the RMW, we need to understand the performance of the few guidance techniques available. We compare RMW forecasts from the Hurricane Analysis and Forecast System (HAFS) to two statistical models and a climatological estimate. Forecasts of the RMW from HAFS are not competitive with statistical derivations of the RMW with marginally better to comparable skill for stronger tropical cyclones. The results indicate that there is a strong need for future improvements to better predict tropical cyclone structure in addition to track and intensity. Key Points Forecasting the radius of maximum wind (RMW) is important for forecasting tropical cyclone hazards A RMW climatology and persistence model is created to determine forecast skill Statistical RMW forecasts are skillful and outperform dynamical model guidance

The horizontal turbulence parameterization is vital for the intensity and structure forecasting of tropical cyclone (TC) in numerical weather prediction (NWP) models. The default two‐dimensional (2D) standard Smagorinsky model with a single universal constant in Weather and Research Forecasting (WRF) model has been proven to be over dissipative for TC, leading to underprediction of TC intensity. This study provides the first attempt to implement the physically based 2D dynamic Smagorinsky model (DSM) for horizontal turbulence parameterization in WRF model for TC forecasts. The DSM dynamically computes the Smagorinsky coefficient as a function of the resolved flow during the simulation, avoiding the need to prescribe the coefficient a prior. The test results of the DSM in a TC NWP model show that the DSM can significantly improve the wind intensity forecasts compared to the standard Smagorinsky model. Plain Language Summary The representation of horizontal turbulent mixing in numerical models is important for the tropical cyclone (TC) forecasting. However, existing horizontal turbulence models (i.e., traditional Smagorinsky model with a constant coefficient) in numerical models underpredict the observed maximum surface wind speed. A dynamic Smagorinsky model with dynamically calculated coefficient according to the state of flow avoids the need for case‐by‐case tuning of the coefficient. The simulations of five TC cases using the dynamic Smagorinsky model present the improved wind intensity forecasts compared to the traditional Smagorinsky model. Key Points Standard Smagorinsky model with default constant coefficient is overly dissipative for tropical cyclone (TC), underpredicting TC intensity We are the first to attempt to implement the dynamic Smagorinsky model for horizontal turbulence parameterization in TC mesoscale numerical weather prediction model Dynamic Smagorinsky model significantly reduces the bias in TC intensity forecasts

We investigate the run-to-run consistency (jumpiness) of ensemble forecasts of tropical cyclone tracks from three global centers: ECMWF, the Met Office, and NCEP. We use a divergence function to quantify the change in cross-track position between consecutive ensemble forecasts initialized at 12-h intervals. Results for the 2019–21 North Atlantic hurricane season show that the jumpiness varied substantially between cases and centers, with no common cause across the different ensemble systems. Recent upgrades to the Met Office and NCEP ensembles reduced their overall jumpiness to match that of the ECMWF ensemble. The average divergence over the set of cases provides an objective measure of the expected change in cross-track position from one forecast to the next. For example, a user should expect on average that the ensemble mean position will change by around 80–90 km in the cross-track direction between a forecast for 120 h ahead and the updated forecast made 12 h later for the same valid time. This quantitative information can support users’ decision-making, for example, in deciding whether to act now or wait for the next forecast. We did not find any link between jumpiness and skill, indicating that users should not rely on the consistency between successive forecasts as a measure of confidence. Instead, we suggest that users should use ensemble spread and probabilistic information to assess forecast uncertainty, and consider multimodel combinations to reduce the effects of jumpiness.

We evaluate the ability of Navy's Earth System Prediction Capability (Navy ESPC) to forecast tropical cyclone (TC) activity in the Western North Pacific basin in the subseasonal timeframe. Navy ESPC forecasts of TCs were added to a logistic regression model that also incorporated the Madden‐Julian Oscillation (MJO) and El Niño Southern Oscillation to forecast basin‐wide TC activity. The skill of the statistical‐dynamical model at 14‐day lead times doubled when forecasted accumulated cyclone energy from Navy ESPC was incorporated into the model. This suggests that Navy ESPC TC tracks contain information that is independent from the MJO. An analysis of forecasted TC positions also suggests that the model has skill in forecasting TC track out to 21 days. These results suggest that combining TC tracks and large‐scale environmental information in a statistical‐dynamical model offers a promising path forward for predictions of basin‐scale or even sub‐basin‐scale subseasonal TC activity.

The combination of moderate vertical wind shear (VWS) and dry environments can produce the most uncertain scenarios for tropical cyclone (TC) genesis and intensification. We investigated the sources of increased uncertainty of TC development under moderate VWS and dry environments using a set of Weather Research and Forecasting (WRF) ensemble simulations. Statistical analysis of ensemble members for precursor events and time-lagged correlations indicates that successful TC development is dependent on a specific set of precursor events. A deficiency in any of these precursor events leads to a failure of TC intensification. The uncertainty of TC intensification can be largely attributed to the probabilistic characteristics of precursor events lining up together before TC intensification. The critical bifurcation point between successful and failed trials in these idealized simulations is the sustained vortex alignment process. Even for the failed intensification cases, most simulations showed deep organized convection, which reformed a midlevel vortex. However, for the failed cycles, the new midlevel vortex could not sustain vertical alignment with the low-level center and was carried away by VWS shortly. Under the most uncertain setup (VWS = 7.5 m s−1 and 50% moisture), the latest-developing ensemble member had seven events of tilt decreasing and increasing again that occurred during the 8 days before genesis. Some unsuccessful precursor events looked very close to the successful ones, implying limits on the intrinsic predictability for TC genesis and intensification in moderately sheared and dry environments.