Storm Modes Research Articles

An Artificially Intelligent System for the Automated Issuance of Tornado Warnings in Simulated Convective Storms

AbstractThe utility of employing artificial intelligence (AI) to issue tornado warnings is explored using an ensemble of 128 idealized simulations. Over 700 tornadoes develop within the ensemble of simulations, varying in duration, length, and associated storm mode. Machine-learning models are trained to forecast the temporal and spatial probabilities of tornado formation for a specific lead time. The machine-learning probabilities are used to produce tornado warning decisions for each grid point and lead time. An optimization function is defined, such that warning thresholds are modified to optimize the performance of the AI system on a specified metric (e.g., increased lead time, minimized false alarms, etc.). Using genetic algorithms, multiple AI systems are developed with different optimization functions. The different AI systems yield unique warning output depending on the desired attributes of the optimization function. The effects of the different optimization functions on warning performance are explored. Overall, performance is encouraging and suggests that automated tornado warning guidance is worth exploring with real-time data.

Political Agenda Setting in the Hybrid Media System: Why Legacy Media Still Matter a Great Deal

This article examines the roles of the media in the process of political agenda setting. There is a long tradition of studies on this topic, but they have mostly focused on legacy news media, thus overlooking the role of other actors and the complex hybrid dynamics that characterize contemporary political communication. In contrast, through an in-depth case study using mixed-methods and multiplatform data, this article provides a detailed analysis of the roles and interactions between different types of media and how they were used by political and advocacy elites. It explores what happened in the different parts of the system, and thus the paths to attention that led to setting this issue in the political and media agendas. The analysis of the case, a partial policy reversal in the United Kingdom provoked by an immigration scandal known as the “Windrush scandal” reveals that the issue was pushed into the agenda by a campaign assemblage of investigative journalism, political and advocacy elites, and digitally enabled leaders. The legacy news media came late but were crucial. They greatly amplified the salience of the issue and, once in “storm mode,” they were key for sustaining attention and pressure, eventually compelling the government to respond. It shows that they often remain at the core of the “national conversation” and certainly in the eye of a media storm. In the contemporary context, characterized by fierce battles for attention, shortening attention spans and fractured audiences, this is key and has important implications for agenda setting and beyond.

Neighborhood- and Object-Based Probabilistic Verification of the OU MAP Ensemble Forecasts during 2017 and 2018 Hazardous Weather Testbeds

Abstract An object-based probabilistic (OBPROB) forecasting framework is developed and applied, together with a more traditional neighborhood-based framework, to convection-permitting ensemble forecasts produced by the University of Oklahoma (OU) Multiscale data Assimilation and Predictability (MAP) laboratory during the 2017 and 2018 NOAA Hazardous Weather Testbed Spring Forecasting Experiments. Case studies from 2017 are used for parameter tuning and demonstration of methodology, while the 2018 ensemble forecasts are systematically verified. The 2017 case study demonstrates that the OBPROB forecast product can provide a unique tool to operational forecasters that includes convective-scale details such as storm mode and morphology, which are typically lost in neighborhood-based methods, while also providing quantitative ensemble probabilistic guidance about those details in a more easily interpretable format than the more commonly used paintball plots. The case study also demonstrates that objective verification metrics reveal different relative performance of the ensemble at different forecast lead times depending on the verification framework (i.e., object versus neighborhood) because of the different features emphasized by object- and neighborhood-based evaluations. Both frameworks are then used for a systematic evaluation of 26 forecasts from the spring of 2018. The OBPROB forecast verification as configured in this study shows less sensitivity to forecast lead time than the neighborhood forecasts. Both frameworks indicate a need for probabilistic calibration to improve ensemble reliability. However, lower ensemble discrimination for OBPROB than the neighborhood-based forecasts is also noted.

Observations of Extreme Short-Term Precipitation Associated with Supercells and Mesovortices

Abstract Extreme hourly rainfall accumulations (e.g., exceeding 75 mm h−1) in several noteworthy flash flood events have suggested that the most intense accumulations were attendant with discrete mesoscale rotation or rotation embedded within larger organized systems. This research aims to explore how often extreme short-term rain rates in the United States are associated with storm-scale or mesoscale vortices. Five years of METAR observations and three years of Stage-IV analyses were obtained and filtered for hourly accumulations over 75 and 100 mm, respectively, clustered into events, and subjectively identified for rotation. The distribution of the short-term, locally extreme events shows the majority of the events were located along the Atlantic and Gulf of Mexico coastlines with additional events occurring in the central plains and into the Midwest. Nearly 50% of the cases were associated with low-level rotation in high-precipitation supercells or mesoscale vortices embedded in organized storm modes. Rotation events occurred more clearly in the warm sector, while nonrotation events tended to occur along a surface boundary. The rotation events tended to produce higher hourly accumulations over a larger region, but were associated with somewhat stronger synoptic-to-mesoscale forcing for ascent and more total column moisture. These results support recent modeling results suggesting that rotationally induced dynamic vertical pressure perturbations should not be ignored when it comes to extreme precipitation and can potentially enhance the short-term rain rates.

Tornado Fatalities: An Environmental Perspective

Abstract Deadly tornadoes are rare events, but that level of rarity varies with many factors. In this work, we summarize and update past research on tornado fatalities, and also discuss the environments of deadly tornadoes both from the perspective of proximity soundings (i.e., point-based) and self-organizing maps (i.e., two-dimensional). In our study of 16 232 tornado events from 2003 to 2017, we find that deadly tornadoes are disproportionately likely to have high (E)F-scale ratings, to have right-moving supercell parent storm modes (deadly QLCS tornadoes are exceptionally rare and tend to result in only one death when they do occur), and to occur during the winter and spring. Warning skill is generally higher for deadly tornadoes than for nondeadly tornadoes: 87% of deadly tornadoes were warned in advance, and nearly 95% of tornado deaths occurred within an active warning. The same environments are warned well for both deadly and nondeadly tornadoes, but the deadly tornadoes tend to occur in environments that are less conducive to weaker (E)F0–1 tornadoes. We identify four prototypical deadly tornado scenarios using self-organizing maps, ranging from marginal environments resulting in relatively few fatalities to major deadly outbreak events. Overall results indicate that the most dangerous tornadoes (i.e., those with high numbers of deaths per deadly tornado) also generally occur in environments and under conditions in which warning skill is high. While, generally speaking, the correct storms are being warned, we include some recommendations for additional research and further improvement.

Microphysical and Dynamical Effects of Mixed-Phase Hydrometeors in Convective Storms Using a Bin Microphysics Model: Melting

AbstractThe dynamics of convective systems are inherently linked to microphysical processes through phase changes that result in warming or cooling. This is especially true of near-surface cooling via evaporation and melting of falling hydrometeors. In most numerical simulations, the melting of frozen hydrometeors (e.g., hail, graupel, snow) is computed within parameterized bulk microphysics schemes, many of which lack the ability to accurately represent mixed-phase hydrometeors (i.e., partially melted ice), which can affect hydrometeor sedimentation, melting, and evaporation of shed drops. To better understand the microphysical and dynamical effects of melting in convective storms, a bin microphysics scheme was used in the Weather Research and Forecasting Model for two idealized cases: a supercell storm and a squall line. Physically based predicted liquid fraction, instantaneous melting, and instantaneous shedding schemes were used to examine the role and importance of melting hydrometeors for these two storm modes. The results suggest that the amount of precipitation is dependent on the representation of melting. Moreover, the dynamic and thermodynamic characteristics of the simulated storms are found to differ substantially between the melting scenarios, resulting in varied storm system evolution; these differences are found to be dependent on the ambient aerosol concentration, although the differences induced by changing the representation of melting generally outweigh those of changing the aerosol loading. The results highlight the large role of melting in convective storm characteristics and suggest that further model improvements are needed in the near future.

Using Overshooting Top Area to Discriminate Potential for Large, Intense Tornadoes

AbstractRecent work established strong links between storm updraft width and the tornado intensity, suggesting that updraft width could be used to gauge potential tornado intensity. It was also posited that overshooting top area (OTA) could be used as an analog for updraft width and, thus, as a means to assess potential tornado intensity in observed storms. The implementation of new high‐resolution GOES‐R series satellites presents a unique opportunity to investigate these findings in severe weather observations. Herein, a method using GOES‐16 longwave infrared satellite data to quantify OTA of tornadic storms is explored. A comparison between observed tornado strength and OTA yields a strong correlation (R2 = 0.54). These results show the potential of these quantifications to be used with real‐time observations of tornadic storms, irrespective of storm mode, seasonality, or geographic location, allowing forecasters to determine which storms pose the highest risk to life and property.

Interpretable Deep Learning for Spatial Analysis of Severe Hailstorms

AbstractDeep learning models, such as convolutional neural networks, utilize multiple specialized layers to encode spatial patterns at different scales. In this study, deep learning models are compared with standard machine learning approaches on the task of predicting the probability of severe hail based on upper-air dynamic and thermodynamic fields from a convection-allowing numerical weather prediction model. The data for this study come from patches surrounding storms identified in NCAR convection-allowing ensemble runs from 3 May to 3 June 2016. The machine learning models are trained to predict whether the simulated surface hail size from the Thompson hail size diagnostic exceeds 25 mm over the hour following storm detection. A convolutional neural network is compared with logistic regressions using input variables derived from either the spatial means of each field or principal component analysis. The convolutional neural network statistically significantly outperforms all other methods in terms of Brier skill score and area under the receiver operator characteristic curve. Interpretation of the convolutional neural network through feature importance and feature optimization reveals that the network synthesized information about the environment and storm morphology that is consistent with our understanding of hail growth, including large lapse rates and a wind shear profile that favors wide updrafts. Different neurons in the network also record different storm modes, and the magnitude of the output of those neurons is used to analyze the spatiotemporal distributions of different storm modes in the NCAR ensemble.

Techniques and Thresholds of Significance for Using WSR-88D Velocity Data to Anticipate Significant Tornadoes

Significant tornadoes (EF2+) make up a very small percentage of the total United States tornado events, but produce the overwhelming majority of tornado fatalities. Identifying significant tornado events in a shortfused warning environment has been a particular focus of the United States National Weather Service’s severe weather program in recent years, with the goal of reducing the loss of life from significant events to the greatest extent possible. This study aims to further this effort by identifying and quantifying the skill of key signals present in Weather Service Radar-1998 Doppler velocity data in the minutes prior to the onset of significant tornado damage. When separated by storm mode, several radar velocity signals are identified that show operationally useful skill in differentiating between significant and weak/nontornadic events in supercells—with lead time. The highest skill scores are achieved by combining maximum volumetric rotational speed and depth of the storm’s mesocyclone, as well as the overall change in rotational speed in the final minutes prior to the onset of significant damage. Very little, if any, predictive skill was found when only the lowest elevation scan was considered, including more frequent supplementary scans in between full volumetric scans. The same signals that showed noteworthy skill for supercells failed to discriminate between significant and weak/nontornadic events in quasi-linear convective Systems and bow echo/mesoscale convective vortex events.

Cardiovascular response as a marker of environmental stress caused by variations in geomagnetic field and local weather

We report the results of a physiological study that include ECG analysis, capillary blood velocity (CBV) data, and blood pressure (BP) measurements obtained under conditions of modified external magnetic field (MF). Each of eight volunteers was sequentially exposed to MFs of three different types for 22 h. A Helmholtz-like MF exposure system was used. The system was specially designed for long-term exposures of human beings to static and low frequency MFs. The MF of the first type reproduced an initially recorded geomagnetic storm (GS). The MF inductions of the other two types were about 55 and 49 μT, which corresponded to the natural local magnetic background and a slightly decreased MF, respectively. In all three cases, the environmental magnetic fluctuations were suppressed by the active Helmholtz system. The variability of RR intervals was considered as a key indicator of cardiovascular response to the MFs and weather conditions. A time-frequency representation (TFR) and a frequency-frequency representation (FFR) were used for processing of the time series of RR intervals. It turned out that the transformation to TFR makes it possible to obtain the characteristic matrix of RR intervals; such a matrix can be used as a subject identifier. FFR was used to calculate correlations of RR intervals with the components of the MFs. The averaged correlation coefficients of RR intervals with the Bx and By components appeared significantly higher (p < 0.001) during the Storm mode in comparison with control. Thus, it is shown that artificial geomagnetic storm can cause a detectable cardiovascular response.

Cold-Season Tornadoes: Climatological and Meteorological Insights

Abstract Tornadoes that occur during the cold season, defined here as November–February (NDJF), pose many societal risks, yet less attention has been given to their climatological trends and variability than their warm-season counterparts, and their meteorological environments have been studied relatively recently. This study aims to advance the current state of knowledge of cold-season tornadoes through analysis of these components. A climatology of all (E)F1–(E)F5 NDJF tornadoes from 1953 to 2015 across a domain of 25°–42.5°N, 75°–100°W was developed. An increasing trend in cold-season tornado occurrence was found across much of the southeastern United States, with a bull’s-eye in western Tennessee, while a decreasing trend was found across eastern Oklahoma. Spectral analysis reveals a cyclic pattern of enhanced NDJF counts every 3–7 years, coincident with the period of ENSO. La Niña episodes favor enhanced NDJF counts, but a stronger relationship was found with the Arctic Oscillation (AO). From a meteorological standpoint, the most-tornadic and least-tornadic NDJF seasons were compared using NCEP–NCAR reanalysis data of various severe weather and tornado parameters. The most-tornadic cold seasons are characterized by warm and moist conditions across the Southeast, with an anomalous mean trough across the western United States. In addition, analysis of the convective mode reveals that NDJF tornadoes are common in both discrete and linear storm modes, yet those associated with discrete supercells are more deadly. Taken together, the perspectives presented here provide a deeper understanding of NDJF tornadoes and their societal impacts, an understanding that serves to increase public awareness and reduce human casualty.

Evaluation of combined effects of lunar gravity simulation and the altered magnetic field on cardiovascular system of healthy volunteers

Evaluation of combined effects of lunar gravity simulation and the altered magnetic field on cardiovascular system of healthy volunteers

Investigating the Relationship between Lightning and Mesocyclonic Rotation in Supercell Thunderstorms

Abstract Relationships between lightning and lightning jumps and physical updraft properties are frequently observed and generally understood. However, a more intensive characterization of how lightning relates to traditional radar-based metrics of storm intensity may provide further operational utility. This study addresses the supercell storm mode because of the intrinsic relationship between a supercell’s characteristic rotating updraft–downdraft couplet, or mesocyclone, and its prolific ability to produce severe weather. Lightning and radar measurements of a diverse sample of 19 supercell thunderstorms were used to assess the conceptual model that lightning and the mesocyclone may be linked by the updraft’s role in the formation and enhancement of each. Analysis of early stages of supercell development showed that the initial lightning jump occurred prior to the time of mesocyclogenesis inferred from three methods by median values of 5–10 min. Comparison between lightning jumps and subsequent increases in mesocyclonic rotation indicated that lightning can also be used to infer or confirm imminent strengthening or reintensification of the mesocyclone. Stronger relationships emerged in supercells that exhibited more robust updrafts, in which 85% of lightning jumps were associated with at least one increase in rotation and 77% of observed increases in rotation were temporally associated with a lightning jump. Preliminary results from analysis of the relationship between lightning jumps and intensification of the low-level mesocyclone in tornadic supercells also offer motivation for the future analysis of lightning data with respect to downdraft-related processes.

Diurnal Preconditioning of Subtropical Coastal Convective Storm Environments

AbstractBoundary layer evolution in response to diurnal forcing is manifested at the mesobeta and smaller scales of the atmosphere. Because this variability resides on subsynoptic scales, the potential influence upon convective storm environments is often not captured in coarse observational and modeling datasets, particularly for complex physical settings such as coastal regions. A detailed observational analysis of diurnally forced preconditioning for convective storm environments of South East Queensland, Australia (SEQ), during the Coastal Convective Interactions Experiment (2013–15) is presented. The observations used include surface-based measurements, aerological soundings, and dual-polarization Doppler radar. The sea-breeze circulation was found to be the dominant influence; however, profile modification by the coastward advection of the continental boundary layer was found to be an essential mechanism for favorable preconditioning of deep convection. This includes 1) enhanced moisture in the city of Brisbane, potentiality due to an urban heat island–enhanced land–sea thermal contrast, 2) significant afternoon warming and moistening above the sea breeze resulting from the advection of the inland convective boundary layer coastward under prevailing westerly flow coupled with the sea-breeze return flow, and 3) substantial variations in near-surface moisture likely associated with topography and land use. For the 27 November 2014 Brisbane hailstorm, which caused damages exceeding $1.5 billion Australian dollars (AUD), the three introduced diurnal preconditioning processes are shown to favor a mesoscale convective environment supportive of large hailstone growth. The hybrid high-precipitation supercell storm mode noted for this event and previous similar events in SEQ is hypothesized to be more sensitive to variations in near-surface and boundary layer instability in contrast to contemporary supercell storms.

High-Resolution Hail Observations: Implications for NWS Warning Operations

Abstract A field research campaign, the Hail Spatial and Temporal Observing Network Effort (HailSTONE), was designed to obtain physical high-resolution hail measurements at the ground associated with convective storms to help address several operational challenges that remain unsatisfied through public storm reports. Field phases occurred over a 5-yr period, yielding hail measurements from 73 severe thunderstorms [hail diameter ≥ 1.00 in. (2.54 cm)]. These data provide unprecedented insight into the hailfall character of each storm and afford a baseline to explore the representativeness of the climatological hail database and hail forecasts in NWS warning products. Based upon the full analysis of HailSTONE observations, hail sizes recorded in Storm Data as well as hail size forecasts in NWS warnings frequently underestimated the maximum diameter hailfall occurring at the surface. NWS hail forecasts were generally conservative in size and at least partially calibrated to incoming hail reports. Storm mode played a notable role in determining the potential range of maximum hail size during the life span of each storm. Supercells overwhelmingly produced the largest hail diameters, with smaller maximum hail sizes observed as convection became progressively less organized. Warning forecasters may employ a storm-mode hail size forecast philosophy, in conjunction with other radar-based hail detection techniques, to better anticipate and forecast hail sizes during convective warning episodes.

The Nonlinear Effect of Information on Political Attention: Media Storms and U.S. Congressional Hearings

Agenda-setting scholars have claimed that the typical punctuated pattern of governmental attention is a consequence of disproportionate information processing. Yet these claims remain unsubstantiated. We tackle this challenge by considering mass media coverage as a source of information for political actors and by examining the relationship between preceding media information and subsequent governmental attention. Employing data capturing U.S. media attention and congressional hearings (1996–2006), we find that the effects of media attention on congressional attention are conditioned by the presence of “media storms”—sudden and large surges in media attention to a given topic. A one-story increase in media attention has a greater effect on congressional attention in the context of a media storm, since media storms surpass a key threshold for catching policymakers’ attention. We find evidence that the influence of media attention on political attention is nonlinear; agenda-setting operates differently when the media are in storm mode.

Storm event representation and analysis based on a directed spatiotemporal graph model

Large amounts of time series of spatial snapshot data have been collected or generated for the monitoring and modeling of environmental systems. Those time series of data also provide the opportunity to study the movements and dynamics of many different natural phenomena. While the snapshot organization is conceptually simple and straightforward, it does not directly capture or represent the dynamic characteristics of the phenomena. This study presents computational methods to identify dynamic events from time series of spatial snapshots. Events are represented as directed spatiotemporal graphs to characterize their initiation, development, movement, and cessation. Graph-based algorithms are used to analyze the dynamics of the events. The method is applied to time series of high-resolution radar reflectivity images during one of the deadliest storm outbreaks that impacted 15 states of southeastern United States between 23 and 29 April 2011. As shown in this case study, convective storm events identified using our methods are consistent with previous studies, and our analysis confirms that the left split/merger occurs more than right split/merger in those convective storm events, which confirms theory, numerical simulations, and other observed case studies. While this study does not differentiate between storm modes, the method shows potential for capturing a more detailed climatology of precipitation characteristics.

Spatial distribution of lightning associated with low-shear thunderstorm environments in the central Appalachian region

Differential heating, the preferential warming of terrain by insolation based on its gradient and aspect, has been observed to play an instrumental role in thunderstorm formation in weak flow regimes. In this study, spatiotemporal cluster analysis was used to group total lightning observations into likely thunderstorms, which were then categorized into lightning-defined (LD) storm modes. Flashes associated with the 351 LD single-cell storms identified were then imported into a geographic information system to compare their frequency against the presence of east- and south-facing surfaces with above-average gradients, termed “enhanced diurnal heating surfaces” (EDHSs). The influence of EDHSs on LD single-cell total lightning and storm frequency was examined via a series of pooled t-tests at eight different spatial scales. The results indicate that EDHSs possessed a significantly greater frequency of LD single-cell storms, 37.6% greater than non-EDHSs, with the relationship best observed at a 15-km resolution. The analysis was then repeated but with terrain aspect ignored. The elimination of aspect as an EDHS requirement removed all statistical significance from the initial results, suggesting that aspect is indeed a key contributor to single-cell thunderstorm formation within areas of complex terrain.

A Preliminary Assessment of Using Spatiotemporal Lightning Patterns for a Binary Classification of Thunderstorm Mode

Abstract This study provides a preliminary, regional assessment of the viability of using spatiotemporal lightning patterns to classify storms into single- versus multi- and supercell storm modes. Total lightning flashes (intracloud and cloud-to-ground flashes) occurring during the afternoon and evening of the period May–August 2012 within an area of the central Appalachian Mountains region were grouped based on their spatial and temporal characteristics using single-linkage clustering. The resulting discrete thunderstorm clusters were characterized in terms of duration, motion, areal extent, and shape. These values were used to formulate four individual attribute scores representing the similarity to the expected values for a typical single-cell thunderstorm. The four scores were then combined into a storm index (SI) using relative weights determined through the analytic hierarchy process (AHP) performed on input from operational forecasters. Of the study days, 89 (72.4%) possessed appreciable lightning, of which 36 (40%) possessed a defined minimum amount of lightning activity required for further analysis. These 36 storm days were divided into two tiers according to the distribution of median daily SI values. The tier containing the 24 storm days (66.7%) with the largest median SI values possessed statistically significant smaller values of 0–6-km wind shear [13.8 knots (kt; 1 kt = 0.51 m s−1)] versus the 12 days in the lower tier of SI values (26.5 kt). This consistency between the total lightning-based classification scheme and increased vertical wind shear associated with lightning-defined multi- and supercells, also evident in synoptic atmospheric composites, lends credibility to the procedure.

Tornado Probability of Detection and Lead Time as a Function of Convective Mode and Environmental Parameters

Abstract The ability to provide advanced warning on tornadoes can be impacted by variations in storm mode. This research evaluates 2 yr of National Weather Service (NWS) tornado warnings, verification reports, and radar-derived convective modes to appraise the ability of the NWS to warn across a variety of convective modes and environmental conditions. Several specific hypotheses are considered: (i) supercell morphologies are the easiest convective modes to warn for tornadoes and yield the greatest lead times, while tornadoes from more linear, nonsupercell convective modes, such as quasi-linear convective systems, are more difficult to warn for; (ii) parameters such as tornado distance from radar, population density, and tornado intensity (F scale) introduce significant and complex variability into warning statistics as a function of storm mode; and (iii) tornadoes from stronger storms, as measured by their mesocyclone strength (when present), convective available potential energy (CAPE), vertical wind shear, and significant tornado parameter (STP) are easier to warn for than tornadoes from weaker systems. Results confirmed these hypotheses. Supercell morphologies caused 97% of tornado fatalities, 96% of injuries, and 92% of damage during the study period. Tornado warnings for supercells had a statistically higher probability of detection (POD) and lead time than tornado warnings for nonsupercells; among supercell storms, tornadoes from supercells in lines were slightly more difficult to warn for than tornadoes from discrete or clusters of supercells. F-scale intensity and distance from radar had some impact on POD, with less impact on lead times. Higher mesocyclone strength (when applicable), CAPE, wind shear, and STP values were associated with greater tornado POD and lead times.