Storm Size Research Articles

Storm Dynamics Control Sedimentation and Shelf‐Bay‐Marsh Sediment Exchange Along the Louisiana Coast

AbstractHurricanes can benefit wetland accretion by augmenting the delivery of mineral sediment, an essential process allowing marshes to offset submergence during rising sea levels. Using Hurricane Gustav (2008, Louisiana) as a control, we examined eight synthetic storms with varying characteristics (track, speed, intensity, size) to evaluate sediment exchange between the inner shelf and bay and bay‐to‐marsh interfaces. All storms showed net landward sediment exchange from the inner shelf to the bay to the marsh—storms with closer proximity, higher intensity, and slower forward speed positively correlated with net sediment exchange; storm size had little impact. Except for slow‐moving storms (½ speed of Gustav), our analyses suggest that most hurricane scenarios cause net bay erosion, because more sediment is conveyed to landward wetlands than is replenished from erosion of the inner shelf. Our results suggest that the ongoing deepening of the bay will likely worsen because of rising sea levels.

Systematic evaluation of swale length, shape, and longitudinal slope with simulated highway runoff for better swale design.

Swales are a low-cost, conveyance and treatment system to manage roadway runoff, but available design guidance is limited. Eight grass swales were constructed in Raleigh, North Carolina, USA, to systematically evaluate the effects of design factors: length, shape, and longitudinal slope under two different storm sizes. Water from an onsite reservoir was used to generate synthetic runoff and simulate flow through the swales. Inflow volume, total suspended sediment (TSS), nitrogen, phosphorus, and four total metals (copper, lead, zinc, and cadmium) were tested with simulated levels representing highway runoff. Efficiency ratios were used to estimate the reductions in inflow volume, pollutant concentrations, and mass loads. Swale length, shape, longitudinal slope, and storm size significantly influenced runoff volume reduction. The longer (30m) trapezoidal swale constructed on the flatter (1%) longitudinal slope provided maximum reductions in sediment and heavy metal concentrations during small-medium storms. Larger storms had modestly reduced pollutant and volume mitigation. Effluent nutrient concentrations generally exceeded the influent exporting nitrogen and phosphorus from all swale configurations. Significantly better pollutant load reductions were provided by the longer swales for all pollutants, except dissolved phosphorus. Therefore, to optimize swale function, designers could maximize the swale length to the greatest extent practicable, particularly when swales receive inflow from end-of-pipe systems draining roadway surfaces. The trapezoidal cross-section was superior to the triangular cross-section for stormwater treatment. Proper vegetation establishment and maintaining optimal grass height are key to proper swale functioning.

Bivariate cubic normal distribution for non-Gaussian problems

Probabilistic models play critical role in various engineering fields. Numerous critical issues exist in probabilistic modeling, especially for non-Gaussian correlated random variables. Traditional parameter-based bivariate distribution models are typically developed for specific types of random variables, which limits their flexibility and applicability. In this study, a flexible bivariate distribution model is proposed, in which the joint cumulative distribution function (JCDF) is derived by expressing the probability as the summation of three basic probabilities corresponding to simple functions. These probabilities are computed using a univariate cubic normal distribution, and thus the proposed model is named as bivariate cubic normal (BCN) distribution. The proposed BCN distribution has been applied in modeling several common bivariate distributions and actual engineering datasets. Results show that the BCN distribution accurately fits the JCDFs of both theoretical distributions and practical datasets, offering significant improvement over existing models. Furthermore, the proposed BCN distribution is utilized in seismic reliability assessment and the calculation of the mean recurrence interval and hazard curve of hurricane wind speed and storm size. Results demonstrate that the BCN distribution excels in modeling and matching capabilities, proving its accuracy and effectiveness in civil engineering applications.

TC‐GEN: Data‐Driven Tropical Cyclone Downscaling Using Machine Learning‐Based High‐Resolution Weather Model

AbstractSynthetic downscaling of tropical cyclones (TCs) is critically important to estimate the long‐term hazard of rare high‐impact storm events. Existing downscaling approaches rely on statistical or statistical‐deterministic models that are capable of generating large samples of synthetic storms with characteristics similar to observed storms. However, these models do not capture the complex two‐way interactions between a storm and its environment. In addition, these approaches either necessitate a separate TC size model to simulate storm size or involve post‐processing to capture the asymmetries in the simulated surface wind. In this study, we present an innovative data‐driven approach for TC synthetic downscaling. Using a machine learning‐based high‐resolution global weather model (ML‐GWM), our approach can simulate the full life cycle of a storm with asymmetric surface wind that accounts for the two‐way interactions between the storm and its environment. This approach consists of multiple components: a data‐driven model for generating synthetic TC seeds, a blending method that seamlessly integrates storm seeds into the surrounding while maintaining the seed structure, and a model based on a recurrent neural network to correct for biases in storm intensity. Compared to observations and synthetic storms simulated using existing statistical‐deterministic and statistical downscaling approaches, our method shows the ability to effectively capture many aspects of TC statistics, including track density, landfall frequency, landfall intensity, and outermost wind extent. Leveraging the computational efficiency of ML‐GWM, our approach shows substantial potential for TC regional hazard and risk assessment.

Changes in Four Decades of Near‐CONUS Tropical Cyclones in an Ensemble of 12 km Thermodynamic Global Warming Simulations

AbstractWe evaluate tropical cyclones (TCs) in a set of thermodynamic global warming (TGW) simulations over the continental United States (CONUS). A 12 km simulation forced by ERA5 provides a 40‐year historical (1980–2019) control. Four complimentary future scenarios are generated using thermodynamic deltas applied to lateral boundary, interior, and surface forcing. We curate a data set of 4,498 6‐hourly TC snapshots in the control and find a corresponding “twin” in each counterfactual, permitting a paired comparison. Warming results in an increase in mean dynamical TC intensity and moisture‐related quantities, with the latter being more pronounced. TC inner cores contract slightly but outer storm size remains unchanged. The frequency with which TCs become more intense is only moderately consistent, with snapshots having increased hazards ranging from 50% to 80% depending on warming level. The fractions of TCs undergoing rapid intensification and weakening both increase across all warming simulations, suggesting elevated short‐term intensity variability.

An Analytical Model for Tropical Cyclone Outer-Size Expansion on the f Plane

Abstract Tropical cyclones are known to expand to an equilibrium size on the f plane, but the expansion process is not understood. In this study, an analytical model for tropical cyclone outer-size expansion on the f plane is proposed. Conceptually, the storm expands because the imbalance between latent heating and radiative cooling drives a lateral inflow that imports absolute vorticity. Volume-integrated latent heating increases more slowly with size than radiative cooling, and hence, the storm expands toward an equilibrium size. The predicted expansion rate is given by the ratio of the difference in size from its equilibrium value rt,eq to an environmentally determined time scale τrt of 10–15 days. The model is fully predictive if given a constant rt,eq, which can also be estimated environmentally. The model successfully captures the first-order size evolution across a range of numerical simulation experiments in which the potential intensity and f are varied. The model predictions of the dependencies of lateral inflow velocity and expansion rate on latent heating rate are also compared well with numerical simulations. This model provides a useful foundation for understanding storm size dynamics in nature.

Numerical simulation of size contraction of Typhoon Cempaka (2021)

In 2021, Cempaka, a tiny tropical cyclone, made landfall in China. As the TC intensified prior to landfall, the tropical cyclone size measured with precipitation decreased significantly. A numerical simulation was conducted to examine the possible processes modulating the storm size. Azimuthally mean potential vorticity (PV) was found to decrease mainly in the middle to upper troposphere between 50- and 80-km radii. The PV budget results indicate that the advection and generation of mean PV associated with asymmetric processes, rather than the symmetric processes, primarily contributed to the decrease in mean PV. These asymmetric processes leading to a negative PV tendency were likely associated with inactive outer rainbands. In contrast, the tangential winds simultaneously expanded radially outward, possibly related to inner-core diabatic heating. The findings here emphasize the importance of outer rainband activity in tropical cyclone size change.摘要2021年台风“查帕卡”增强过程中风场向外扩张, 但卫星云图显示降水范围减小. 高分辨率数值模拟结果显示, “查帕卡”降水范围减小对应着50-80km半径处对流层中层位涡的减小. 位涡诊断结果表明, 非对称过程是导致上述位涡减小的关键物理过程, 而非对称过程主要同外雨带活动减弱有关. 上述结果强调了外雨带活动对热带气旋尺度变化的潜在重要影响.

AN INVESTIGATION INTO PREDICTION FORECASTING METHODS IN VARIOUS CHHATTISGARH INDIAN WEATHER CONDITIONS

Existing climate prediction models are physics-based and use supercomputers to predict future climate evolution. Physics-based predictions require improved atmospheric models, which can be challenging to find and develop, and improving the fundamental resolution of simulations can be computationally demanding. Measuring rainfall around the world has many scientific and social benefits. These benefits include of the processes that cause global climate change, increased precipitation and hydrological processes, improved weather information, Better identification of severe storms, including predictions of storm size and landfall. The Global Precipitation Measure (GPM) has completed the work of defining the scientific and metric goals and framework needed to achieve those goals. Precipitation forecasting is the prediction of the amount and type of precipitation (rain, snow, sleet, etc.) that will fall in a specific location over a certain period of time. This is typically done using numerical weather prediction models, which take into a wide range of atmospheric and oceanic data to make predictions about the weather. Precipitation forecast is the process that predicts the possibility of weather conditions at an exact time and place. An artificial neural network is a type of machine learning model that can be used for long-term precipitation forecasting. This model is trained on a large amount of historical weather data and uses this data to predict future rainfall. Artificial neural networks can be used to model complex and linear relationships between atmospheric and oceanic variables and precipitation. The Model may help us in forecasting long-range precipitation, approximately before one month of monsoon season over Kabirdham.

Storm-induced nitrogen transport via surface runoff, interflow and groundwater in a pomelo agricultural watershed, southeast China

The storm-induced export of nitrogen (N) from agricultural watersheds significantly impacts aquatic ecosystems, yet the mechanisms of source supply and transport behind N species remain unclear. Here, we investigated the hydrological factors influencing the timing and magnitude of river N species export in a Chinese pomelo agricultural watershed. We conducted continuous observations of watershed hydrology, N species, and their isotopic ratios along a soil-groundwater-river continuum during two storm events in 2018–2019. We found the export flux of river NO3–N covers ∼80% of the total N flux during storms, and the rest for other N species. Our results further revealed distinct pathways and timing of N transport among different N species, especially between ammonium N (NH4–N) and nitrate N (NO3–N). NH4–N in stormflow predominantly originates from sewage and soil leachate, rapidly transported via surface runoff and interflow. Orchard fertilization (contributed 41–56% based on SIAR analysis) was the major source of river NO3–N, which underwent initial dilution via surface runoff and subsequently became enriched through delayed discharge of soil leachate and groundwater. The variations in timing and magnitude of N transport between storms can be explained by antecedent conditions such as precipitation, soil N pools, and storm size. These findings emphasize the hydrological controls on N export from agricultural watersheds, and highlight the variations in source supply and transport pathways among different N species. The insights gained from this study hold significance for managing agricultural pollution and restoring impaired aquatic systems.

Assessing RRFS versus HRRR in Predicting Widespread Convective Systems over the Eastern CONUS

Abstract This study provides a comparison of the operational HRRR version 4 and its eventual successor, the experimental Rapid Refresh Forecast System (RRFS) model (summer 2022 version), at predicting the evolution of convective storm characteristics during widespread convective events that occurred primarily over the eastern United States during summer 2022. In total 32 widespread convective events were selected using observations from the MRMS composite reflectivity, which includes an equal number of MCSs, quasi-linear convective systems (QLCSs), clusters, and cellular convection. Each storm system was assessed on four primary characteristics: total storm area, total storm count, storm area ratio (an indicator of mean storm size), and storm size distributions. It was found that the HRRR predictions of total storm area were comparable to MRMS, while the RRFS overpredicted total storm area by 40%–60% depending on forecast lead time. Both models tended to underpredict storm counts particularly during the storm initiation and growth period. This bias in storm counts originates early in the model runs (forecast hour 1) and propagates through the simulation in both models indicating that both miss storm initiation events and/or merge individual storm objects too quickly. Thus, both models end up with mean storm sizes that are much larger than observed (RRFS more so than HRRR). Additional analyses revealed that the storm area and individual storm biases were largest for the clusters and cellular convective modes. These results can serve as a benchmark for assessing future versions of RRFS and will aid model users in interpreting forecast guidance.

Automated and Objective Thunderstorm Identification and Tracking Using Geostationary Lightning Mapper (GLM) Data

Abstract A unique storm identification and tracking method is analyzed in varying storm environments within the United States spanning 273 hours in 2018. The methodology uses a quantity calculated through fusion of radar-based vertically integrated liquid (VIL) and satellite-based GLM flash rate density (FRD) called VILFRD to identify storms in space and time. This research analyzes GLM data use within VILFRD for the first time (method original: O), assesses four modifications to VILFRD implementation to find a more stable storm size with time (method new: N), larger storms (method original dilated: OD), or both (method new dilated: ND), and compares VILFRD methods with storm tracking using the 35-dBZ isosurface at −10°C (method non-VILFRD: NV). A case study analysis from 2019 is included to assess methods on a smaller scale and introduce a “lightning only” (LO) version of VILFRD. Large study results highlight that VILFRD-based storm identification produces smaller storms with more lightning than the NV method, and the NV method produces larger storms with a more stable size over time. Methods N and ND create smaller storm size fluctuations, but size changes more often. Dilation (OD, ND) creates larger storms and almost double the number of storms identified relative to nondilated methods (O, N, NV). The case study results closely resemble the large sample results and show that the LO method identifies storms with more lightning and shorter durations. Overall, these findings can aid in choice of storm tracking method based on desired user application and promote further testing of a lightning-only version of VILFRD.

Buffering Effect of Atmosphere–Ocean Coupling on Intensity Changes of Tropical Cyclones Under a Changing Climate

AbstractThe effect of atmosphere–ocean coupling on intensity changes of tropical cyclones (TCs) under global warming was examined using a regional high‐resolution three‐dimensional atmosphere–ocean coupled model. A storyline event attribution approach was applied to four historical intense TCs in the western North Pacific. The results indicate that atmosphere–ocean coupling buffers TC intensification as global warming progresses. This buffering effect increased as storms traveled northward. Moreover, the effect intensified as warming progressed, because reductions in sea surface temperature induced by the storm increased as the storm strengthened in future warmer climates. The magnitude of the buffering effect depended on the storm's size and translation speed; a large, slow‐moving storm had significant resilience against global warming, whereas a compact, fast‐moving storm was sensitive to global warming. A high‐resolution atmosphere–ocean coupled model is important for more reliable future projections of TC intensity under the changing climate.

Hindcast Insights from Storm Surge Forecasting of Super Typhoon Saola (2309) in Hong Kong with the Sea, Lake and Overland Surges from Hurricanes Model

Super Typhoon Saola (2309) skirted past south-southeast of Hong Kong within 40 km on the night of 1 September 2023, posing a significant storm surge threat to Hong Kong. Given the close proximity of Saola with a peak intensity of about 210 km/h within 300 km of Hong Kong, a close call of the “super typhoon direct-hit” scenario, this case provides valuable insights from a hindcast review of storm surge forecasts and warning operation using the Sea, Lake and Overland Surges from Hurricanes (SLOSH) model, which is the operational storm surge model adopted by the Hong Kong Observatory (HKO). The performance of the HKO’s PRobabilistic Inundation Map Evaluation System (PRIMES) using both statistical and model ensemble approaches was also reviewed in this paper. Saola was a challenging case for operational forecasting of a compact TC structure with changes in storm size and intensity when it came close to Hong Kong. With major observations of storm structure using weather radar and dense automatic weather station, tide gauge and water level gauge networks, the high sensitivity of storm surge forecasts to the storm size parameter and the distance of closest approach was clearly revealed in the case of Saola. Even with a circularly symmetric TC parametric model like SLOSH, the hindcast review results illustrated that the model outputs were reasonably accurate during the closest approach of Saola given an accurate storm size and distance of closest approach were input, and using a highly computationally efficient storm surge model made it possible for the nowcasting of storm surges to handle compact and intense TC direct-hit cases in operational TC forecasting. Taking a nowcasting approach not only helps provide more reliable storm tide forecasts, but also facilitates the formulation of a better warning strategy when making final-call decisions in emergency response actions, based on the more frequent real-time analysis of TC position, intensity and storm size and the more accurate prediction of these parameters. A nowcasting workflow for storm surge operation was proposed in this paper.

Operational Storm Surge Forecasting at the National Hurricane Center: The Case for Probabilistic Guidance and the Evaluation of Improved Storm Size Forecasts Used to Define the Wind Forcing

Abstract The primary source of guidance used by the Storm Surge Unit (SSU) at the National Hurricane Center (NHC) for issuing storm surge watches and warnings is the Probabilistic Tropical Storm Surge model (P-Surge). P-Surge is an ensemble of Sea, Lake, and Overland Surges from Hurricanes (SLOSH) model forecasts that is generated based on historical error distributions from NHC official forecasts. A probabilistic framework is used for operational storm surge forecasting to account for uncertainty related to the tropical cyclone track and wind forcing. Previous studies have shown that the size of a storm’s wind field is an important factor that can affect storm surge. A simple radius of maximum wind (RMW) prediction scheme was developed to forecast RMW based on NHC forecast parameters. Verification results indicate this scheme is an improvement over the RMW forecasts used by previous versions of P-Surge. To test the impact of the updated RMW forecasts in P-Surge, retrospective cases were selected from 25 storms from 2008 to 2020 that had an adequate number of observations. Evaluation of P-Surge forecasts using these improved RMW forecasts shows that the probability of detection is higher for most probability of exceedance thresholds. In addition, the forecast reliability is improved, and there is an increase in the number of high probability forecasts for extreme events at longer lead times. The improved RMW forecasts were recently incorporated into the operational version of P-Surge (v2.9), and serve as an important step toward extending the lead time of skillful and reliable storm surge forecasts.

Impact of particle size on heavy metal contamination in human health from sandstorms in Iraq

BackgroundThe Middle East, including Iraq, is undergoing significant climate change, which has accelerated desertification, drought, and dust storms. Rapid industrialization and urbanization, which frequently contaminate the atmosphere with hazardous metals, are to blame for these changes. MethodsThe human health risks of six common heavy metals, including lead (Pb), chromium (Cr), cadmium (Cd), copper (Cu), zinc (Zn), and nickel (Ni), were examined in dust particles that were collected from five major administrative districts in Iraq in order to understand the impact of particle size on heavy metals in a sandy dust storm. The hazard quotient (HQ), hazard index (HI), and carcinogenic risk (RI) were used to evaluate the health risk for both carcinogenic and non-carcinogenic substances. Resultssmaller particles' Cu and Cd concentrations generally increased from 63 μm to 75 μm, and the maximum Cr content was discovered in the particles (75 μm). The HI values of heavy metals were less than 1 and suggested that these elements did not offer a non-cancer risk to either adults or children. TCR values for Ni were higher than 1 × 10−6 in particle size (63 μm, 75 μm, and 200 μm) for children. Moreover, TCR values for Cr were higher than 1 × 10−6 in particle size (63 μm and 200 μm) with the exception that Cr in all sites were higher than TCR 1 × 10−4 in particle size (75 μm) for children ConclusionThe amounts of six metals varied according on the size of the dust storm's particles. Given their health hazard, these metals' health concerns, particularly those of chromium and nickel for children, should receive significant attention.

Separating impacts of storm characteristics and mitigation activities on suspended sediment: Implications for sampling to evaluate mitigation success

Mitigation activities are often implemented to reduce anthropogenic sediment loads to sensitive ecosystems, including coral reefs, but effectiveness is rarely documented due to a lack of pre-mitigation data and to the difficulty of separating the impacts of mitigation from those of storm size and intra- and inter-storm variability. Here we present a method using stormwise metrics to quantify the effectiveness of a sediment mitigation project in American Samoa, where an aggregate quarry covering <1 % of a small watershed (drainage area 1.8 km2) generated approximately half of the annual suspended sediment load to the coast. Rainfall, runoff, turbidity and suspended sediment concentrations (SSC, mg L–1), yield (SSY, tons km−2 per unit time) and loads (SSL, tons per unit time) were monitored periodically over a five-year period upstream and downstream of the quarry. Mitigation included revegetation and placement of gravel on roads and processing platforms followed by installation of retention ponds. Storms were identified using automated storm separation, and a simple regression model using maximum area-normalized storm discharge (Qmax) estimated SSC, SSY and SSL for each mitigation stage (post-gravel, post-ponds) for a reference storm sequence. Regression parameters, and SSL and SSC for the reference storm sequence were calculated for a range of storm separation parameters. Monte Carlo sampling of the regression errors was used to calculate a mean and standard deviation of SSL and its change following mitigation. SSC was nearly 100x higher during large storms than during small storms, demonstrating the need to account for storm size to detect impacts of mitigation. After mitigation, SSL downstream of the quarry decreased by 63–82 % for all storms and by 88–92 % for small storms, and SSY downstream of the quarry was similar to SSY from the forested watershed upstream. Mitigation of sediment loads can target a small fraction of a watershed area that accounts for the majority of the human impact. Combining stormwise metrics with simple regression modeling and uncertainty analysis is a simple and novel way to detect and quantify the effectiveness of mitigation activities on sediment loads.

The Variation of Ionospheric O+ and H+ Outflow on Storm Timescales

AbstractGeomagnetic storms are primarily driven by stream interaction regions (SIRs) and coronal mass ejections (CMEs). Since SIR and CME storms have different solar wind and magnetic field characteristics, the magnetospheric response may vary accordingly. Using FAST/TEAMS data, we investigate the variation of ionospheric O+ and H+ outflow as a function of geomagnetic storm phase during SIR and CME magnetic storms. The effects of storm size and solar EUV flux, including solar cycle and seasonal effects, on storm time ionospheric outflow, are also investigated. The results show that for both CME and SIR storms, the O+ and H+ fluences peak during the main phase, and then declines in the recovery phase. However, for CME storms, there is also significant increase during the initial phase. Because the outflow starts during the initial phase in CME storms, there is time for the O+ to reach the plasma sheet before the start of the main phase. Since plasma is convected into the ring current from the plasma sheet during the main phase, this may explain why more O+ is observed in the ring current during CME storms than during SIR storms. We also find that outflow fluence is higher for intense storms than moderate storms and is higher during solar maximum than solar minimum.

Hydrological evaluation of conservation bench terrace system in outer foothills of Himalayas using soft computing techniques

Soil erosion has been problematic in the outer foothills of the Himalayas. Therefore, it is necessary to implement suitable conservation measures to check erosion issues. The present study experimented on conservation bench terraces (CBT) as a conservation measure in the outer foot Himalayas in Dehradun, India. The study evaluates the effectiveness of different CBTs of embankment heights 10 cm (CBT-10), 15 cm (CBT-15), and 20 cm (CBT-20) to runoff and soil loss under different rainfall storms. Further, the estimation of Soil loss (Sloss) was evaluated using soft computing techniques (machine learning models). The soft computing techniques were applied to the data obtained from CBT plots for hydrological assessment. The methods used for analysis were support vector machine (SVM), Bagging, and Random Forest (RF) models. The performance of developed models was evaluated using mean absolute error (MAE), root mean square error (RMSE), Nash-Sutcliffe efficiency (NSE), Pearson's correlation coefficient (PCC), and Wilmott Index (WI) along with graphical representation (Scatter plot and Taylor diagram). The results showed that the CBT techniques were observed to reduce runoff and soil loss. Among CBT, CBT-20 was the most significant as the number of runoff events was less than others. Considering the storm size, the maximum percentage of runoff generated from the storm size is more important than 100 mm. Moreover, soil loss (Sloss) estimation showed that the SVM models perform better in all three scenarios, which are results of three scenarios such as R2 = 0.50, 0.52 and 0.80; MAE = 0.03, 0.03 and 0.10; RMSE = 0.08, 0.07 and 0.27; NSE = 0.39, 0.12 and 0.21; PCC = 0.71, 0.725 and 0.89; and WI = 0.67, 0.794, and 0.49, respectively. Therefore, SVM techniques can be applied to the estimation of Sloss. Results can be recommended for precision farming under climate changes in the hilly and plain areas and similar models can be applied to other areas.

Effect of Storm Size on Sea Surface Cooling and Tropical Cyclone Intensification in the Western North Pacific

Abstract The effect of tropical cyclone (TC) size on TC-induced sea surface temperature (SST) cooling and subsequent TC intensification is an intriguing issue without much exploration. Via compositing satellite-observed SST over the western North Pacific during 2004–19, this study systematically examined the effect of storm size on the magnitude, spatial extension, and temporal evolution of TC-induced SST anomalies (SSTA). Consequential influence on TC intensification is also explored. Among the various TC wind radii, SSTA are found to be most sensitive to the 34-kt wind radius (R34) (1 kt ≈ 0.51 m s−1). Generally, large TCs generate stronger and more widespread SSTA than small TCs (for category 1–2 TCs, R34: ∼270 vs 160 km; SSTA: −1.7° vs −0.9°C). Despite the same effect on prolonging residence time of TC winds, the effect of doubling R34 on SSTA is more profound than halving translation speed, due to more wind energy input into the upper ocean. Also differing from translation speed, storm size has a rather modest effect on the rightward shift and timing of maximum cooling. This study further demonstrates that storm size regulates TC intensification through an oceanic pathway: large TCs tend to induce stronger SST cooling and are exposed to the cooling for a longer time, both of which reduce the ocean’s enthalpy supply and thereby diminish TC intensification. For larger TCs experiencing stronger SST cooling, the probability of rapid intensification is half of smaller TCs. The presented results suggest that accurately specifying storm size should lead to improved cooling effect estimation and TC intensity prediction. Significance Statement Storm size has long been speculated to play a crucial role in modulating the TC self-induced sea surface temperature (SST) cooling and thus potentially influence TC intensification through ocean negative feedback. Nevertheless, systematic analysis is lacking. Here we show that larger TCs tend to generate stronger SST cooling and have longer exposure to the cooling effect, both of which enhance the strength of the negative feedback. Consequently, larger TCs undergo weaker intensification and are less likely to experience rapid intensification than smaller TCs. These results demonstrate that storm size can influence TC intensification not only from the atmospheric pathway, but also via the oceanic pathway. Accurate characterization of this oceanic pathway in coupled models is important to accurately forecast TC intensity.

Diagnosing Storm Mode with Deep Learning in Convection-Allowing Models

Abstract While convective storm mode is explicitly depicted in convection-allowing model (CAM) output, subjectively diagnosing mode in large volumes of CAM forecasts can be burdensome. In this work, four machine learning (ML) models were trained to probabilistically classify CAM storms into one of three modes: supercells, quasi-linear convective systems, and disorganized convection. The four ML models included a dense neural network (DNN), logistic regression (LR), a convolutional neural network (CNN), and semisupervised CNN–Gaussian mixture model (GMM). The DNN, CNN, and LR were trained with a set of hand-labeled CAM storms, while the semisupervised GMM used updraft helicity and storm size to generate clusters, which were then hand labeled. When evaluated using storms withheld from training, the four classifiers had similar ability to discriminate between modes, but the GMM had worse calibration. The DNN and LR had similar objective performance to the CNN, suggesting that CNN-based methods may not be needed for mode classification tasks. The mode classifications from all four classifiers successfully approximated the known climatology of modes in the United States, including a maximum in supercell occurrence in the U.S. Central Plains. Further, the modes also occurred in environments recognized to support the three different storm morphologies. Finally, storm mode provided useful information about hazard type, e.g., storm reports were most likely with supercells, further supporting the efficacy of the classifiers. Future applications, including the use of objective CAM mode classifications as a novel predictor in ML systems, could potentially lead to improved forecasts of convective hazards. Significance Statement Whether a thunderstorm produces hazards such as tornadoes, hail, or intense wind gusts is in part determined by whether the storm takes the form of a single cell or a line. Numerical forecasting models can now provide forecasts that depict this structure. We tested several automated algorithms to extract this information from forecast output using machine learning. All of the automated methods were able to distinguish between a set of three convective types, with the simple techniques providing similarly skilled classifications compared to the complex approaches. The automated classifications also successfully discriminated between thunderstorm hazards, potentially leading to new forecast tools and better forecasts of high-impact convective hazards.