Cyclone Intensity Prediction Research Articles

Deep-TCP: Multi-source data fusion for deep learning-powered tropical cyclone intensity prediction to enhance urban sustainability

Tropical cyclones (TC) exert a profound impact on cities, causing extensive damage and losses. Thus, TC Intensity Prediction is crucial for creating sustainable cities as it enables proactive measures to be taken, including evacuation planning, infrastructure reinforcement, and emergency response coordination. In this study, we propose a Deep learning-powered TC Intensity Prediction (Deep-TCP) framework. In particular, Deep-TCP contains a data constraint module for fusing data features from multiple sources and establishing a unified global representation. To capture the spatiotemporal attributes, a Spatial-Temporal Attention (ST-Attention) module is built to distill insights from environmental variables. To improve the robustness and stability of the predictions, an encoder-decoder module that utilizes the ConvGPU unit is introduced to enhance feature maps. Then, a novel feature enhancement module is built to bolster the generalization capability and solve the dependency attenuation. The results demonstrate that the Deep-TCP framework significantly outperforms various benchmarks. Additionally, it effectively predicts multiple TC categories within the 6–24 h timeframe, showing strong capability in predicting changing trends. The reliable prediction results are potentially beneficial for disaster management and urban planning, significantly enhancing urban sustainability by improving preparedness and response strategies.

Lifetime Performance of the Operational Hurricane Weather Research and Forecasting Model (HWRF) for North Atlantic Tropical Cyclones

Abstract The Hurricane Weather Research and Forecasting Model (HWRF) was the flagship hurricane model at NOAA’s National Centers for Environmental Prediction for 16 years and a state-of-the-art tool for tropical cyclone (TC) intensity prediction at the National Weather Service and across the globe. HWRF was a joint development between NOAA research and operations, specifically the Environmental Modeling Center and the Atlantic Oceanographic and Meteorological Laboratory. Significant support also came from the National Hurricane Center, Developmental Testbed Center, University Corporation for Atmospheric Research, universities, cooperative institutes, and the TC community. In the North Atlantic basin, where most improvement efforts focused, HWRF intensity forecast errors decreased by 45%–50% at many lead times between 2007 and 2022. These large improvements resulted from increases in horizontal and vertical resolution, as well as advances in model physics and data assimilation. HWRF intensity forecasts performed particularly well over the Gulf of Mexico in recent years, providing useful guidance for a large number of impactful landfalling hurricanes. Such advances were made possible not only by significant gains in computing but also through substantial investment from the Hurricane Forecast Improvement Program.

Forecasting of tropical cyclones ASANI (2022) and MOCHA (2023) over the Bay of Bengal - real time challenges to forecasters

This study examines the track and intensity forecasts of two typical Bay of Bengal tropical cyclones (TC) ASANI and MOCHA. The analysis of various Numerical Weather Prediction (NWP) model forecasts [ECMWF (European Centre for Medium range Weather Forecast), NCEP (National Centers for Environmental Prediction), NCUM (National Centre for Medium Range Weather Forecast-Unified Model), IMD (India Meteorological Department), HWRF (Hurricane Weather Research and Forecasting)], MME (Multi-model Ensemble), SCIP (Statistical Cyclone Intensity Prediction) model, and OFCL (Official) forecasts shows that intensity forecasts of ASANI and track forecasts of MOCHA were reasonably good, but there were large errors and wide variation in track forecasts of ASANI and in intensity forecasts of MOCHA. Among all model forecasts, the track forecast errors of IMD model and MME were least in general for ASANI and MOCHA respectively. Also, the landfall point forecast errors of IMD were least for ASANI, and the MME and OFCL forecast errors were least for MOCHA. No model is found to be consistently better for landfall time forecast for ASANI, and the errors of ECMWF, IMD and HWRF were least and of same order for MOCHA. The intensity forecast errors of OFCL and SCIP were least for ASANI, and the forecast errors of HWRF, IMD, NCEP, SCIP and OFCL were comparable and least for MOCHA up to 48 h forecast and HWRF errors were least thereafter in general. The ECMWF model forecast errors for intensity were found to be highest for both the TCs. The results also show that although there is significant improvement of track forecasts and limited or no improvement of intensity forecast in previous decades but challenges still persists in real time forecasting of both track and intensity due to wide variation and inconsistency of model forecasts for different TC cases.

Cyclone Intensity Estimation Using Deep Learning

Cyclone Intensity Estimation based on Deep Learning focuses on cyclone intensity estimation, enhancing early warning systems and empowering decision-makers to take proactive measures to safeguard vulnerable communities and infrastructure, which aims to minimize the devastating impacts of cyclones worldwide. The challenge in cyclone intensity prediction persists due to the complex and dynamic nature of these storms. Traditional methods fall short of capturing rapid changes. This research project leverages the power of deep learning to enhance the accuracy of cyclone intensity prediction by utilizing both satellite images and grayscale representations as input datasets. It involves preprocessing and feature extraction from satellite images captured during cyclonic events. Convolutional Neural Networks (CNN) are employed to automatically learn and extract relevant patterns from these complex datasets. In parallel, grayscale images derived from the original satellite images are utilized to capture essential structure information that contributes to cyclone intensity. The fusion of information from both modalities is achieved through a novel deep learning patterns that go beyond the capabilities of traditional intensity estimation methods, and architecture, fostering a comprehensive understanding of cyclonic patterns that go beyond the capabilities of traditional intensity estimation methods. A new approach seeks to automate cyclone estimation, streamlining timelines and increasing efficiency by merging deep learning with hurricane-focused satellite data. This approach might result in more precise forecasts, Reducing the amount of casualties and property damage in prone areas.

Advanced hybrid CNN-Bi-LSTM model augmented with GA and FFO for enhanced cyclone intensity forecasting

Predicting cyclone intensity is an important aspect of weather forecasting since it influences disaster preparation and response. This framework addresses the pressing need for precise cyclone intensity prediction by presenting a unique predictive model based on a hybrid CNN and Bi-LSTM architecture optimized using a Genetic Algorithm (GA) enhanced Fruit Fly Optimizer (FFO). Existing methods have primarily relied on traditional machine learning models and meteorological data, demonstrating limitations in capturing the complex spatial-temporal patterns inherent in cyclone evolution. These drawbacks include insufficient feature extraction abilities, underutilization of convolutional neural networks (CNN), and poor model tuning. This unique method incorporates a hybrid CNN and Bi-LSTM architecture that is tuned by a Genetic Algorithm (GA) enhanced Fruit Fly Optimizer (FFO), resulting in higher cyclone intensity prediction accuracy. The experimental results are implemented in Python software, and they reveal that this method outperforms current models by an average of 21% when compared to existing methods such as VGG-16 achieved an accuracy of 78% and Ty 5-CNN (95.23%). The suggested CNN-Bi-LSTM model predicts cyclone strength with an excellent accuracy of 99.4%. This unique approach offers a possible avenue for increasing cyclone intensity prediction, hence improving disaster preparedness and risk mitigation efforts in sensitive locations.

Operational tropical cyclone intensity prediction in the bay of bengal - An updated empirical technique

Given its role in preventing cyclone-related property damage and fatalities, effective weather forecasting is of utmost importance. Numerous methods for forecasting the weather have been developed via extensive research. To improve precise and trustworthy weather forecast techniques, however, there is a constant need. This study suggests an improved empirical methodology for forecasting tropical cyclone (TC) intensity, particularly in the Bay of Bengal (BoB) region. The study is based on an analysis of 32 cyclones that occurred between 2003 and 2019. The model predicts that TC has significantly strengthened. This article's methodology has the potential to raise forecast precision at 12-hour intervals. However, the forecast's precision decreases over time.

User-Responsive Diagnostic Forecast Evaluation Approaches: Application to Tropical Cyclone Predictions

Abstract The Hurricane Forecast Improvement Project (HFIP; renamed the “Hurricane Forecast Improvement Program” in 2017) was established by the U.S. National Oceanic and Atmospheric Administration (NOAA) in 2007 with a goal of improving tropical cyclone (TC) track and intensity predictions. A major focus of HFIP has been to increase the quality of guidance products for these parameters that are available to forecasters at the National Weather Service National Hurricane Center (NWS/NHC). One HFIP effort involved the demonstration of an operational decision process, named Stream 1.5, in which promising experimental versions of numerical weather prediction models were selected for TC forecast guidance. The selection occurred every year from 2010 to 2014 in the period preceding the hurricane season (defined as August–October), and was based on an extensive verification exercise of retrospective TC forecasts from candidate experimental models run over previous hurricane seasons. As part of this process, user-responsive verification questions were identified via discussions between NHC staff and forecast verification experts, with additional questions considered each year. A suite of statistically meaningful verification approaches consisting of traditional and innovative methods was developed to respond to these questions. Two examples of the application of the Stream 1.5 evaluations are presented, and the benefits of this approach are discussed. These benefits include the ability to provide information to forecasters and others that is relevant for their decision-making processes, via the selection of models that meet forecast quality standards and are meaningful for demonstration to forecasters in the subsequent hurricane season; clarification of user-responsive strengths and weaknesses of the selected models; and identification of paths to model improvement. Significance Statement The Hurricane Forecast Improvement Project (HFIP) tropical cyclone (TC) forecast evaluation effort led to innovations in TC predictions as well as new capabilities to provide more meaningful and comprehensive information about model performance to forecast users. Such an effort—to clearly specify the needs of forecasters and clarify how forecast improvements should be measured in a “user-oriented” framework—is rare. This project provides a template for one approach to achieving that goal.

Enhancing Cyclone Intensity Prediction for Smart Cities Using a Deep-Learning Approach for Accurate Prediction

Enhancing Cyclone Intensity Prediction for Smart Cities Using a Deep-Learning Approach for Accurate Prediction

A tropical cyclone intensity prediction model using conditional generative adversarial network

This study proposes a model for predicting tropical cyclone (TC) intensity evolution by utilizing the conditional Wasserstein generative adversarial network with gradient penalty (CWGAN-GP) for TC-related risk assessment. The TC intensity evolution is treated as a random variable, which is the output of a nonlinear system conditioned on a series of input random variables that characterize both the TC state (e.g., translation speed and potential intensity) and environment (e.g., vertical wind shear and ocean mixed layer depth). To represent the high-dimensional non-Gaussian probabilistic characteristics, the system is modeled by the CWGAN-GP and calibrated concerning the dataset of 1010 historical TCs from the western North Pacific basin. The over-ocean and overland branches of the model are calibrated separately and then merged. The suitability of the established model is validated by comparing the predicted TC intensity to the observations and the results of existing models such as those based on linear regression. Numerical results indicate that the proposed model successfully replicates the probabilistic properties such as the non-Gaussian marginal distribution of the intensity change and the input-output joint distribution and moments between the TC intensity and predictors, which the linear regression model cannot provide. Further application examples are also carried out by using the model to simulate the evolution of historical TC events and the extreme TC intensities in Southern China, demonstrating its potential role in the TC hazard assessment.

Cyclone trajectory and intensity prediction with uncertainty quantification using variational recurrent neural networks

Cyclone trajectory and intensity prediction with uncertainty quantification using variational recurrent neural networks

Evaluation of a Mesoscale Coupled Ocean‐Atmosphere Configuration for Tropical Cyclone Forecasting in the South West Indian Ocean Basin

AbstractThe performance in term of tropical cyclone track and intensity prediction of the new coupled ocean‐atmosphere system based on the operational atmospheric model AROME‐Indian Ocean and the ocean model NEMO is assessed against that of the current operational configuration in the case of seven recent tropical cyclones. Five different configurations of the forecast system are evaluated: two with the coupled system, two with an ocean mixed layer parameterization and one with a constant sea surface temperature. For each ocean‐atmosphere coupling option, one is initialized directly with the MERCATOR‐Ocean PSY4 product as in the current operational configuration and the other with the ocean state that is cycled in the AROME‐NEMO coupled suite since a few days before the cyclone intensification. The results show that the coupling with NEMO generally improves the intensity of cyclones in AROME‐IO, reducing the mean intensity bias of the 72 hr forecast of about 10 hPa. However, the impact is especially significant when the TCs encounter a slow propagation phase. For short‐term forecasts (less than 36 hr), the presence of a cooling in the initial state that has been triggered by the AROME high‐resolution cyclonic winds in a previous coupled forecast already improves the tropical cyclone intensity bias of 2–3 hPa for both coupled or uncoupled configurations.

Tropical Cyclone intensity prediction based on hybrid learning techniques

Tropical Cyclone intensity prediction based on hybrid learning techniques

Tropical Cyclone Intensity and Track Prediction in the Bay of Bengal Using LSTM-CSO Method

Tropical Cyclone Intensity and Track Prediction in the Bay of Bengal Using LSTM-CSO Method

Tropical cyclone intensity prediction by inter- and intra-pattern fusion based on multi-source data

Tropical cyclones (TCs) are one of the most destructive natural disasters, which can bring huge life and economic losses to the global coastal areas. Accurate TC intensity prediction is critical for disaster prevention and loss reduction, but the dynamic processes involved in TCs are complicated and not adequately understood, which make the intensity prediction is still a challenging task. In recent years, several deep-learning (DL)-based methods have been developed for TC prediction by mining TC intensity series or related environmental factors. However, information hidden between the two different data sources is generally ignored. Here, a novel DL-based TC intensity prediction network named Pre_3D is proposed, which aimed to mine of inter- and intra-patterns of TC intensity and related external factors independently by separate feature extraction sub-networks. An MLP network is adopted to achieve adaptive fusion of the two patterns for accurate TCs intensity prediction. TC records from several agencies were used to evaluate generalizability of the proposed framework and extensive experiments were conducted validate its effectiveness. The experimental results demonstrate that the models based on the Pre_3D framework achieved considerable performance. ConvGRU-based Pre_3D yields a significant improvement of over 15% in prediction accuracy in 24 h prediction relative to official agencies.

A Machine Learning (ML)-Based Approach to Improve Tropical Cyclone Intensity Prediction of NCMRWF Ensemble Prediction System

A Machine Learning (ML)-Based Approach to Improve Tropical Cyclone Intensity Prediction of NCMRWF Ensemble Prediction System

Improving Analysis and Prediction of Tropical Cyclones by Assimilating Radar and GNSS-R Wind Observations: Ensemble Data Assimilation and Observing System Simulation Experiments Using a Coupled Atmosphere–Ocean Model

Abstract This study investigates the impact of assimilating ground-based radar reflectivity and wind data on tropical cyclone (TC) intensity prediction. The effect on a high-impact TC in the western North Pacific region that penetrated the Bashi Channel is examined. A multiscale correction based on the successive covariance localization (SCL) method is adopted to improve the analysis and forecast performance. In addition, GNSS-R wind speed is assimilated jointly in the rapid update assimilation framework to complement the TC boundary layer where radar data are limited. Model experiments are conducted and evaluated using the observing system simulation experiments (OSSEs) framework with a coupled atmosphere–ocean model nature run. Taking the experiment without data assimilation as the baseline, assimilating the radar data with the standard localization and SCL methods reduces the wind speed analysis error by 12% and 44%, respectively. The SCL method dominates the improvement in TC intensity prediction with a lead time longer than 2 days and the TC’s peak intensity forecast is improved by 18 hPa. The additional assimilation of the GNSS-R wind speed observation further reduces the wind speed error in the low-level analysis by 12% and 5% under the standard and SCL radar assimilation framework, respectively. GNSS-R wind assimilation leads to a further 6-hPa improvement in TC’s peak intensity. However, the sampling error introduced by the SCL method restrains the effect of GNSS-R assimilation. Sensitivity experiments with different GNSS-R data arrangements show that better GNSS-R wind coverage and additional wind direction information can further improve the TC analysis. Significance Statement Tropical cyclone (TC) intensity prediction over the western North Pacific (WNP) region remains a significant challenge due to limited observations. This study aims to improve the TC intensity prediction in WNP by assimilating the ground-based radar data using a multiscale correction framework and incorporating with the satellite ocean surface wind speed observation. We particularly focus on a high-impact TC like Typhoon Hato (2017), which penetrated the Bashi Channel and later made landfall in China, causing great damage. Our results showed that the assimilation strategy improved the TC intensity prediction for a lead time longer than 2 days. These results demonstrate the great potential of these observations and can provide guidance for future applications in operation centers.

An Upper Ocean Thermal Field Metrics Dataset

The upper ocean provides a source of thermal energy for tropical cyclone development and maintenance through a series of complex interactions. In this work, we develop a seventeen-year dataset of upper ocean thermal field metrics for use in tropical cyclone studies and development of tropical cyclone intensity prediction models. These metrics include the surface temperature, two different measures of vertically integrated heat content, and four different measures of vertically averaged temperature. Some metrics have been used to study upper-ocean energy response to tropical cyclone passage, while others have been employed to improve operational tropical cyclone intensity prediction models. The vertically integrated ocean heat content has been used to improve tropical cyclone intensity forecasts at U.S. tropical cyclone forecast centers and is an integral part of several operational intensity forecast models. A static 2005–2021 dataset that includes all twelve metrics described within is available on the Naval Research Laboratory web server, and a subset of six metrics have been produced in real-time at Fleet Numerical Meteorology and Oceanography Center and provided to the public via the GODAE server since 2021.

Tropical Cyclone Intensity Prediction Using Deep Convolutional Neural Network

In this study, deep convolutional neural network (CNN) models of stimulated tropical cyclone intensity (TCI), minimum central pressure (MCP), and maximum 2 min mean wind speed at near center (MWS) were constructed based on ocean and atmospheric reanalysis, as well Best Track of tropical hurricane data over 2014–2018. In order to explore the interpretability of the model structure, sensitivity experiments were designed with various combinations of predictors. The model test results show that simplified VGG-16 (VGG-16 s) outperforms the other two general models (LeNet-5 and AlexNet). The results of the sensitivity experiments display good consistency with the hypothesis and perceptions, which verifies the validity and reliability of the model. Furthermore, the results also suggest that the importance of predictors varies in different targets. The top three factors that are highly related to TCI are sea surface temperature (SST), temperature at 500 hPa (TEM_500), and the differences in wind speed between 850 hPa and 500 hPa (vertical wind shear speed, VWSS). VWSS, relative humidity (RH), and SST are more significant than MCP. For MWS and SST, TEM_500, and temperature at 850 hPa (TEM_850) outweigh the other variables. This conclusion also implies that deep learning could be an alternative way to conduct intensive and quantitative research.

Tropical cyclone track and intensity prediction skill of GFS model over NIO during 2019 & 2020

The Tropical Cyclone (TC) track prediction using different NWP models and its verification is the critical task to provide prior knowledge about the model errors, which is beneficial for giving the model guidance-based real-time cyclone warning advisories. This study has attempted to verify the Global Forecast System (GFS) model forecasted tropical cyclone track and intensity over the North Indian Ocean (NIO) for the years 2019 and 2020. GFS is one of the operational models in the India Meteorological Department (IMD), which provides the medium-range weather forecast up to 10 days. The forecasted tracks from the GFS forecast are obtained using a vortex tracker developed by Geophysical Fluid Dynamics Laboratory (GFDL). A total of 13 tropical cyclones formed over the North Indian Ocean, eight during 2019 and five in 2020 have been considered in this study. The accuracy of the model predicted tracks and intensity is verified for five days forecasts (120 h) at 6-h intervals; the track errors are verified in terms of Direct Position Error (DPE), Along Track Error (ATE) and Cross-Track Error (CTE). The annual mean DPE over NIO during 2019 (51–331 km) is lower than 2020 (82–359 km), and the DPE is less than 150 km up to 66 h during 2019 and 48 h during 2020. The positive ATE (76–332 km) indicates the predicted track movement is faster than the observed track during both years. The positive CTE values for most forecast lead times suggest that the predicted track is towards the right side of the observed track during both years. The cyclone Intensity forecast for the maximum sustained wind speed (MaxWS) and central mean sea level pressure (MSLP) are verified in terms of mean error (ME) and root mean square error (RMSE). The errors are lead time independent. However, most of the time model under-predicted the cyclone intensity during both years. Finally, there is a significant variance in track and intensity errors from the cyclone to cyclone and Bay of Bengal basin to the Arabian Sea basin.

A neural network framework for fine-grained tropical cyclone intensity prediction

Huge losses of life and economy are brought by Tropical Cyclones(TCs). Accurate TC intensity prediction is crucial for disaster prevention and emergency decision-making, but the chaotic nature of TC makes the intensity prediction a challenging task. Recently, TC intensity prediction has been regarded as a spatio-temporal prediction problem by a surging number of researchers, and deep learning methods have been used to model both spatial and temporal characteristics of which. However, existing studies on TC intensity prediction based on deep learning methods are still inadequate in spatio-temporal feature modeling and prediction granularity. In this study, a neural network framework specifically for TC intensity prediction named TC-Pred is proposed. To be Specific, a novel feature extraction and aggregation approach is designed considering the characteristics of multi-source environmental variables, and the sequence-to-sequence architecture is innovatively introduced to make fine-grained predictions. Moreover, a module inspired by the convolutional transformer is developed that aims to alleviate the long-term dependency decay problem and improve model performance. Extensive experiments are performed on the environmental variables dataset to verify the effectiveness and practicability of the proposed framework. The experimental results demonstrate that most of the models based on TC-Pred have achieved better performance, and the TC-Pred(ConvGRU) model obtains a significant improvement by 18.31%, 7.01%, 14.17%, and 11.95% improvements compared with baselines at 6h, 12h, 18h, and 24h intervals, respectively.