A Rare Tropical Cyclone Necessitating the Issuance of Gale or Storm Wind Warning Signal in Hong Kong in Late Autumn in 2022—Severe Tropical Storm Nalgae

Since 1970, tropical cyclone (TC) track forecasts have improved steadily in the Atlantic basin. This improvement has been linked primarily to advances in numerical weather prediction (NWP) models. Concurrently, with few exceptions, the development and operational use of statistical track prediction schemes have experienced a relative decline. Statistical schemes provided the most accurate TC track forecasts until approximately the late 1980s. In this note, it is shown that increased reliance on the global NWP models does not always guarantee the best forecast. Here, Hurricane Ivan is used from the 2004 Atlantic TC season as a classical example, and reminder, of how strong climatological signals still can add substantial value to TC track forecasts, in the form of improved accuracy and increased timeliness at minimal computational cost. In an 8-day period in early September 2004, Hurricane Ivan was repeatedly, and incorrectly, forecast by 12 operational NWP models to move with a significant northward (poleward) component. It was found that the mean 24-h trajectory forecasts of a consensus of five commonly used NWP track prediction aids had a statistically significant right-of-track bias. Furthermore, the official track forecasts, which relied heavily on erroneous numerical guidance over this period, were also found to have significant poleward trajectory errors. At the same time, a climatology-based prediction technique, drawn entirely from the historical record of motion characteristics of TCs in geographical locations similar to Ivan, correctly and consistently indicated a more westward motion component, had a small directional spread, and was supported by a large number of archived cases. This climatological signal was in conflict with the deterministic NWP model output, and it is suggested that the large errors in the official track forecast for TC Ivan could have been reduced considerably by taking into greater account such a strong climatological signal. The potential impact of such an error reduction is a saving of lives and billions of dollars in both actual damage and unnecessary evacuations costs, for just this one hurricane. We also suggest that this simple strategy of examining the strength of the climatological signal be considered for all TCs to identify cases where the NWP and official forecasts differ significantly from strong, persistent climatological signals.

Stochastic parameterization has become commonplace in numerical weather prediction (NWP) models used for probabilistic prediction. Here a specific stochastic parameterization will be related to the theory of stochastic differential equations and shown to be affected strongly by the choice of stochastic calculus. From an NWP perspective the focus will be on ameliorating a common trait of the ensemble distributions of tropical cyclone (TC) tracks (or position); namely, that they generally contain a bias and an underestimate of the variance. With this trait in mind the authors present a stochastic track variance inflation parameterization. This parameterization makes use of a properly constructed stochastic advection term that follows a TC and induces its position to undergo Brownian motion. A central characteristic of Brownian motion is that its variance increases with time, which allows for an effective inflation of an ensemble’s TC track variance. Using this stochastic parameterization the authors present a comparison of the behavior of TCs from the perspective of the stochastic calculi of Itô and Stratonovich within an operational NWP model. The central difference between these two perspectives as pertains to TCs is shown to be properly predicted by the stochastic calculus and the Itô correction. In the cases presented here these differences will manifest as overly intense TCs, which, depending on the strength of the forcing, could lead to problems with numerical stability and physical realism.

A tropical cyclone (TC) Vayu developed over the Arabian Sea during June, 2019. It followed a northward track from southeast Arabian Sea to northeast Arabian Sea close to Gujarat coast during 10–12 June 2019 as a very severe cyclonic storm. It skirted south Gujarat coast by recurving west-northwestwards during 13th–14th June and again made a northeastward recurvature on 16th June towards Gujarat coast. However, it weakened over Sea on 17th. There was large divergence among various models in predicting the track of TC Vayu leading to over warning for Gujarat state and also delay in dewarning leading to evacuation of people from coastal region. Hence, a study has thus been taken up to analyze the performance of various numerical weather prediction (NWP) models in forecasting the track of TC Vayu so as to find out the reason for above limitation of NWP models. Results suggest that there is a need to relook into the existing multi-model ensemble (MME) technique which outperforms individual models in track forecasting. There is also a need to improve the individual deterministic model guidance so as to suitably represent the interaction between mid-latitude westerlies with the TC and steering anticyclone by improving the initial and boundary conditions through augmented direct and remotely sensed observations over the Arabian Sea and their assimilation in NWP models.

Machine Learning Weather Prediction (MLWP) models have recently demonstrated remarkable potential to rival physics‐based Numerical Weather Prediction (NWP) models, offering global weather forecasts at a fraction of the computational cost. However, thorough evaluations are essential before considering MLWP models as replacements for NWP models. This study presents a comprehensive evaluation of four leading MLWP models—GraphCast, PanguWeather, Aurora, and FourCastNet—against observations and three state‐of‐the‐art NWP models in predicting tropical cyclones (TCs) across all tropical ocean basins. All MLWP models exhibited strong skill in forecasting TC tracks, achieving an average track error of less than 200 km at a 96‐hr forecast lead time. However, they consistently underestimated maximum sustained wind speeds compared to NWP models and observations. The low bias in TC intensity forecasts by MLWP models is linked to similar bias in their training data, along with the double penalization effect. MLWP models realistically captured the absolute vorticity patterns and their advection, demonstrating their ability to represent the dynamics underlying TC translation. They also captured the low‐level convergence and vertical warm core structure of TCs, although the magnitudes were weaker than observed, highlighting the linkage between dynamical and thermodynamical processes. The consistency in magnitude between various physical fields in the MLWP models suggests that they intuitively learn the interrelationships among different physical fields during the evolution of weather systems, demonstrating their ability to capture complex physical interactions. Among the MLWP models, Aurora showed superior performance, surpassing GraphCast, PanguWeather, and FourCastNet.

In this study, an experiment based on the Dynamical-Statistical-Analog Ensemble Forecast model for Landfalling Typhoon Gale (DSAEF_LTG model) was conducted to predict tropical cyclone (TC)-induced potential maximum gales in South China for the first time. A total of 21 TCs with maximum gales greater than or equal to 17.2 m/s (at least one station) during 2011–2018 were selected for this experiment. Among them, 16 TCs in 2011–2015 were selected as the training samples aimed at identifying the best forecast scheme, while 5 TCs in 2016–2018 were selected as the independent samples to verify the best forecast scheme. Finally, the forecast results were compared with four numerical weather prediction (NWP) models (i.e., CMA, ECMWF, JMA and NCEP) based on four forecasting skill scores (Threat Score, False Alarm Ratio, Missing Ratio and Bias Score) at thresholds above Beaufort Scale 7 and 10, and two more indicators (Mean Absolute Error and pearson correlation coefficient). The results revealed encouraging forecasting ability in South China for the DSAEF_LTG model. In general, the DSAEF_LTG model showed higher forecasting skill than the NWP models above the critical thresholds. While the DSAEF_LTG model was prone to false alarms, the NWP models were prone to missing alarms, especially for an intense scale (≥Beaufort Scale 10). In addition, the DSAEF_LTG model also performed best with the smallest forecasting error. Furthermore, the DSAEF_LTG model had distinct advantages in predicting target TCs with typical tracks and widespread gales, both in terms of the wind field pattern and the magnitude of central wind speeds. However, for sideswiping TCs with small-scale gales, the DSAEF_LTG model tended to over-predict and held no advantage over the NWP models, which could perhaps be improved by introducing more reasonable ensemble forecast schemes in further research.

Data‐driven artificial intelligence weather prediction (AIWP) models show great potential in weather forecasts, facilitating paradigm shift of prediction from a deductive to an inductive inference. However, this shift raises concerns regarding the performance of the AIWP models in severe weather forecasting. Tropical cyclones (TCs) are one of the most typical cases of severe weather prediction. In this study, we compare Western Pacific TCs in 2023 produced by the AIWP model, Pangu‐Weather, with those generated by numerical weather prediction (NWP) models, specifically the European Center for Medium‐Range Weather Forecasts (ECMWF) and the National Centers for Environmental Prediction (NCEP), in the operational context. We analyze the impact of different initial conditions (ICs) on AIWP models, representative by Pangu‐Weather, in TC forecasting. Our analysis includes statistical evaluation of forecast skill related to TC activity, track, intensity, and a case study on the physical structure of a TC. The Pangu‐Weather model exhibits superior forecast skills compared to the NWP model regarding TC tracks and environmental variables within TC activity domains, particularly at longer forecast lead times. However, the overly smooth forecasts of Pangu‐Weather and the coarse‐resolution ICs with reduced information of TCs potentially lead to the underestimation of intensity and a weakened dynamic‐thermodynamic structure of TCs. Also, Pangu‐Weather shows low sensitivity to ICs concerning TC structure and intensity. Hybrid models combining physical processes with data‐driven approaches may enhance AIWP performance for severe weather forecasting.

A new algorithm to generate wave heights consistent with tropical cyclone official forecasts from the Joint Typhoon Warning Center (JTWC) has been developed. The process involves generating synthetic observations from the forecast track and the 34-, 50-, and 64-kt wind radii. The JTWC estimate of the radius of maximum winds is used in the algorithm to generate observations for the forecast intensity (wind), and the JTWC-estimated radius of the outermost closed isobar is used to assign observations at the outermost extent of the tropical cyclone circulation. These observations are then interpolated to a high-resolution latitude–longitude grid covering the entire extent of the circulation. Finally, numerical weather prediction (NWP) model fields are obtained for each forecast time, the NWP model forecast tropical cyclone is removed from these fields, and the new JTWC vortex is inserted without blending zones between the vortex and the background. These modified fields are then used as input into a wave model to generate waves consistent with the JTWC forecasts. The algorithm is applied to Typhoon Yagi (2006), in anticipation of which U.S. Navy ships were moved from Tokyo Bay to an area off the southeastern coast of Kyushu. The decision to move (sortie) the ships was based on NWP model-driven long-range wave forecasts that indicated high seas impacting the coast in the vicinity of Tokyo Bay. The sortie decision was made approximately 84 h in advance of the high seas in order to give ships time to steam the approximately 500 n mi to safety. Results from the new algorithm indicate that the high seas would not affect the coast near Tokyo Bay within 84 h. This specific forecast verifies, but altimeter observations show that it does not outperform, the NWP model-driven wave analysis and forecasts for this particular case. Overall, the performance of the new algorithm is dependent on the JTWC tropical cyclone forecast performance, which has generally outperformed those of the NWP model over the last several years.

Artificial intelligence and numerical weather prediction models: A technical survey

Forecasts of precipitation and water vapor made by the Met Office global numerical weather prediction (NWP) model are evaluated using products from satellite observations by the Special Sensor Microwave Imager/Sounder (SSMIS) and Tropical Rainfall Measuring Mission (TRMM) Multisatellite Precipitation Analysis (TMPA) for June–September 2011, with a focus on tropical areas (30°S–30°N). Consistent with previous studies, the predicted diurnal cycle of precipitation peaks too early (by ~3 h) and the amplitude is too strong over both tropical ocean and land regions. Most of the wet and dry precipitation biases, particularly those over land, can be explained by the diurnal-cycle discrepancies. An overall wet bias over the equatorial Pacific and Indian Oceans and a dry bias over the western Pacific warm pool and India are linked with similar biases in the climate model, which shares common parameterizations with the NWP version. Whereas precipitation biases develop within hours in the NWP model, underestimates in water vapor (which are assimilated by the NWP model) evolve over the first few days of the forecast. The NWP simulations are able to capture observed daily-to-intraseasonal variability in water vapor and precipitation, including fluctuations associated with tropical cyclones.

This paper presents a single-channel atmospheric correction method for remotely sensed infrared (wavelength of 3–15 μm) images with various observation angles. The method is based on basic radiative transfer equations with a simple absorption-focused regression model to calculate the optical thickness of each atmospheric layer. By employing a simple regression model and re-organization of atmospheric profiles by considering viewing geometry, the proposed method conducts atmospheric correction at every pixel of a numerical weather prediction model in a single step calculation. The Visible Infrared Imaging Radiometer Suite (VIIRS) imaging channel (375 m) I4 (3.55~3.93 μm) and I5 (10.50~12.40 μm) bands were used as mid-wavelength and thermal infrared images to demonstrate the effectiveness of the proposed single-channel atmospheric correction method. The estimated sea surface temperatures (SSTs) obtained by the proposed method with high resolution numerical weather prediction models were compared with sea-truth temperature data from ocean buoys, multichannel-based SST products from VIIRS/MODIS, and results from MODerate resolution atmospheric TRANsmission 5 (MODTRAN 5), for validation. High resolution (1.5 km and 12 km) numerical weather prediction (NWP) models distributed by the Korea Meteorological Administration (KMA) were employed as input atmospheric data. Nighttime SST estimations with the I4 band showed a root mean squared error (RMSE) of 0.95 °C, similar to that of the VIIRS product (RMSE: 0.92 °C) and lower than that of the MODIS product (RMSE: 1.74 °C), while estimations with the I5 band showed an RMSE of 1.81 °C. RMSEs from MODTRAN simulations were similar (within 0.2 °C) to those of the proposed method (I4: 0.81 °C, I5: 1.67 °C). These results demonstrated the competitive performance of a regression-based method using high-resolution numerical weather prediction (NWP) models for atmospheric correction of single-channel infrared imaging sensors.

There is a growing need for improvement in tropical cyclone (TC) vital parameters (Knaff 2011) in view of the requirements of numerical weather prediction (NWP) models and various stake holders. As the damage due to a TC is directly proportional to the square of the maximum sustained wind (MSW) and loss due to a TC is proportional to cube of MSW, the surface wind structure associated with a TC serves insurance agencies to assess the damage due to a TC. The disaster managers who prepare for the impact of a landfalling TC may use the wind field information as guidance as to where the most severe wind or surge damage may occur. The TC Vital parameters also serve as input to NWP models and storm surge models that are run prior to landfalling events to create synthetic vortex (Chourasia et al. 2013), as most of the NWP models fail to simulate accurately the location and intensity of the TC. The creation of synthetic vortex helps in improving the track and intensity forecast of the model. In the parametric storm surge prediction models, the surface wind structure in the quadrant base form alongwith the radius of maximum wind (RMW) and pressure drop (ΔP) at the centre are utilised to create the wind stress and hence predict the storm surge (Dube et al. 2013). In post-event cases, these wind structure data are utilised for diagnosis of TC and to better plan for future TC forecasts. Engineers and planners rely on historical TC information to determine long-term risks to facilities and infrastructure and to ensure the resilience of communities to potential disasters. Another most important use of this product is the determination of ship avoidance area over the sea due to a TC.

High resolution wind speed forecasts are crucial for a range of applications, including the management of onshore wind power generation. Conventional wind speed forecasting is bound to the coarse spatial resolution of NWP models of 2-30 km. The wind speed complementarity model (WiCoMo)* is a high-resolution wind downscaling model that provides distributions of annual wind speeds across Germany at a horizontal resolution of 25 m x 25 m. This work aims to combine high resolution wind downscaling with numerical weather prediction (NWP) models to improve accuracy and resolve local effects, particularly in complex terrain. Quantile mapping was used to derive a transfer function at each 25 m x 25 m grid cell based on annual historical wind speeds calculated by WiCoMo and the NWP models respectively. The function was then applied to hourly time series of NWP models to simulate downscaled predictions. In addition, power curves of wind turbines were used to calculate the onshore wind power output of Germany from the high-resolution forecast. Validation metrics were used to compare the performance of the WiCoMo-enhanced NWP models with raw NWP outputs. The analysis demonstrates that the WiCoMo-enhanced NWP models outperform raw NWP across all tested models. For the year 2022, the MAE of NEMS4 was reduced from 1.66 m/s to 1.13 m/s and for NEMSGLOBAL it improved by almost 33%. The MBE was reduced to near 0 in all cases. Furthermore, spatial evaluations show that local wind speed effects often falling below the grid size in NWP models, such as hilltop speed-up or sheltering valleys, are resolved by the downscaling. The study suggests that localized wind speeds at wind turbine sites improve the accuracy of wind power output predictions. However, several limitations are identified, including challenges in applying corrections during specific weather conditions. Additionally, the modelled wind power output could not be validated at single turbine sites, limiting the validity of estimates for the entire country. The study demonstrates the potential of WiCoMo-enhanced NWP models in improving wind speed forecasting capabilities. The findings have important implications for various applications, including renewable energy planning and risk assessment. * Christopher Jung and Dirk Schindler. Introducing a new wind speed complementarity model. Energy, 265:126284, 2023. ISSN 0360-5442. doi: https://doi.org/10.1016/j.energy.2022.126284.

Improving the Turbulence Coupling between High Resolution Numerical Weather Prediction Models and Lagrangian Particle Dispersion Models

A comprehensive assessment of the forecast skill of various meteorological indices over the Indian region from different numerical weather prediction (NWP) models is lacking in the literature. In this study the performance of four NWP models, namely Global Ensemble Forecast System (GEFSv12), European Center for Medium Range Forecasting (ECMWF), Climate Forecast System (CFSv2) and Indian Institute of Tropical Meteorology (IITM) towards forecasting of precipitation, temperature and associated meteorological indices, is evaluated at short to medium timescales across the Indian region. Further, the effect of ocean atmospheric (OA) oscillations on the precipitation/temperature forecast skill from the different NWP models is also assessed. Results show that the NWP models are better in predicting the meteorological indices than the quantitative forecasts. The ECMWF model was found to be the best for PCP forecasting with CFSv2 performing poorly. For temperature the GEFSv12 model performance was the lowest, compared to rest of the models. The models show poor skill in forecasting monsoon season precipitation compared to non‐monsoon and the temperature forecasts from the NWP models are particularly poor for the northern basins. Skilful temperature forecasts are observed in the Northwestern, Indo‐Gangetic and Central basins for the CFSv2 and ECMWF models. The forecast skill of precipitation indices are higher in the northwestern, central and Indo‐Gangetic basins compared to the rest. The skill of precipitation indices, namely rainy days, extreme rainy days and consecutive wet days, is higher during the monsoon seasons while the prediction skill of consecutive dry days is higher during the non‐monsoon season. OA analysis revealed that the ENSO phases have dominant effect on the forecast skill of the precipitation only. The temperature and meteorological indices forecasts are not affected significantly by the OA phases. The outcomes of this study have implications towards irrigation scheduling and water resources management decision making in India.

We examine the merit of atmosphere–ocean coupled models for tropical cyclone (TC) predictions in the western North Pacific (WNP), where accurate TC predictions remain challenging. The UK Met Office operational atmospheric global numerical weather prediction (NWP) model is compared with two trial coupled configurations, in which the operational atmospheric model is coupled to a one‐dimensional mixed‐layer ocean model and a three‐dimensional dynamical ocean model. Reforecasts for the 2016 TC season show that the coupled models outperform the NWP model for TC location predictions, with a systematic improvement of 50–100 km over the seven‐day forecasts, but the coupled models amplify the underestimation of TC intensity in the NWP model. Nearly identical TC predictions (for both location and intensity) are found in the two coupled models, indicating the dominance of thermodynamic processes at the air–sea interface for TC predictions on these timescales. The improved prediction of the TC position in the coupled models is associated with an enhanced Western North Pacific Subtropical High (WNPSH), which introduces an anticyclonic steering flow anomaly that shifts TC tracks further west in the southern part of the region and further east in the northern part. Based on sensitivity experiments, we show that these improvements in the coupled models are due mainly to colder initial sea‐surface temperatures (SSTs). Air–sea feedbacks do not change the WNPSH or TC tracks noticeably. Apart from the effect of the initial SSTs, tropical ocean warming due to air–sea interaction in the coupled forecasts can also reduce the predicted TC intensity, presumably due to a stronger regional Hadley circulation with increased subtropical subsidence.