Tornado Events Research Articles

Examining Intraethnic Disparities in Tornado Hazard Understanding, Reception, and Response in Latinx/e Communities in the Southeast United States

Abstract The Southeast region in the United States has a high risk for tornadoes and tornado-related losses (e.g., fatalities, injuries, damages). For the public in tornado-vulnerable areas, receiving warning information is essential for seeking protective action during a tornado event. However, for Latinx/e households in the Southeast, social, cultural, and political factors pertaining to language barriers, cultural differences, and trust in public officials often inhibit the effectiveness of warning communication strategies. The current study utilized a survey among 820 Latinx/e adults across 11 states in the U.S. Southeast. Using the Wilcoxon–Mann–Whitney (WMW) test, we examined if language, nativity, and immigration status impacted how Latinx/e populations receive, understand, and act upon severe weather information in the Southeast region. Results found intraethnic disparities related to tornado hazard understanding, reception, and response among Latinx/e adults based on language, nativity, and U.S. citizenship status. These findings suggest that NWS and its partners work with Latinx/e communities to adjust their communication strategies to consider ethnic heterogeneity in future hazard research, alert systems, and targeted outreach campaigns. Significance Statement We sought to determine if Latinx/e populations in the Southeast United States face distinct social and systematic challenges and exhibit varying levels of preparedness and awareness in the face of tornado hazards. We analyzed tornado warning comprehension, weather information sources, and barriers to protective actions among 820 Latinx/e participants, along with intraethnic disparities or vulnerabilities that may impact an individual’s or community’s ability to prepare for, respond to, and adapt to tornado hazards. These findings are important for forecasters, broadcasters, emergency managers, and others responsible for alerting the public during severe weather situations in the Southeast. Understanding differences in human vulnerability in the face of tornado threats can help inform better disaster planning and response for vulnerable and underrepresented populations.

Spatio-temporal patterns of disaster impact and recovery in YouTube content

ABSTRACT This study investigates the utilization of YouTube as a data source in crisis management, focusing on human behavior during a tornado event in the Southern Moravia Region, Czech Republic, in 2021. Utilizing YouTube's Data API, we harvested video metadata and content to explore spatio-temporal patterns in disaster impact and recovery. Our analysis reveals how YouTube content reflects the immediate and long-term responses to the disaster, offering insights into community reactions, damage extent, and recovery processes. By integrating automated and manual content analysis, we uniquely demonstrate YouTube's potential in enhancing spatial analyses and visualizations for disaster management. This research underscores the value of social media data in providing multiple perspectives on human behavior during and after disaster, contributing to a nuanced understanding of disaster impact and community resilience.

The Propagation, Evolution, and Rotation in Linear Storms (PERiLS) Project

Abstract Quasi-linear convective systems (QLCSs) are responsible for approximately a quarter of all tornado events in the United States, but no field campaigns have focused specifically on collecting data to understand QLCS tornadogenesis. The Propagation, Evolution, and Rotation in Linear Storms (PERiLS) project was the first observational study of tornadoes associated with QLCSs ever undertaken. Participants were drawn from more than 10 universities, laboratories, and institutes, with over 100 students participating in field activities. The PERiLS field phases spanned 2 years, late winters and early springs of 2022 and 2023, to increase the probability of intercepting significant tornadic QLCS events in a range of large-scale and local environments. The field phases of PERiLS collected data in nine tornadic and nontornadic QLCSs with unprecedented detail and diversity of measurements. The design and execution of the PERiLS field phase and preliminary data and ongoing analyses are shown.

Pressure Measurements and Video Observations Near and Inside Three EF2 Tornadoes

From the spring of 2016 through the spring of 2019, three EF2-rated tornadoes were sampled with in-situ tornado probes as part of the Pressure Acoustics Recordings Inside Tornadoes Experiment. The probe meteorological instrumentation resolved the temperature, humidity, and pressure-deficit characteristics near or inside all three tornadoes. Probe recorded pressure and visual observations during each tornado event show unique attributes, including multiple strong pressure irregularities before and just after the first EF2 tornado, a rotating bowl-shaped cloud feature located some distance ahead of the second EF2 tornado, independent of the tornado core and immediate inflow layer. Pressure measurements of the bowl feature reveal a brief pressure perturbation lasting roughly 4 s as it passed near or directly over the research team before the documentation of a 41 hPa pressure deficit. During the third EF2 near in-situ deployment, pressure trace observations also reveal pressure fluctuations ahead of the tornado core lasting some 65 s before the documentation of a 13 hPa pressure deficit. These pressure measurements and video observations are consistent in location with recent ultra-high-resolution simulations of the Streamwise Vorticity Current (SVC), or more precisely, the related Vertical Vorticity Sheet (VVS) and Pressure Deficit Lobe (PDL) regions. This paper describes the measured pressure deficit traces during three EF2 tornadoes near and in situ, augmented with brief, one-of-a-kind, near and in-situ video observations of a unique bowl feature, tornado inflow, corner-flow, and tornado core regions. The recorded data and instrumentation were analyzed; results are presented and discussed.

Climatological Aspects of Notable Tornado Events in Chile

Abstract Tornadoes in Chile seem to develop in what are called “high-shear, low-CAPE” (HSLC) environments. An analysis of convective parameters from the ERA5 reanalysis during 16 notable tornadoes in Chile showed that several increased markedly before the time of the reports. The significant tornado parameter (STP) was able to discriminate the timing and location of the tornadoes, even though it was not created with that goal. We established thresholds for the severe hazards in environments with reduced buoyancy (SHERBE) parameter (≥1) and the STP (≤−0.3) to further identify days favorable for tornado activity in Chile. The SHERBE and STP parameters were then used to conduct a climatological analysis from 1959 to 2021 of the seasonal, interannual, and latitudinal variations of the environments that might favor tornadoes. Both parameters were found to have a strong annual cycle. The largest magnitudes of STP were found to be generally confined to south-central Chile, in agreement with the (sparse) tornado record. The probability of a day with both SHERBE and STP values beyond their thresholds was greatest between May and August, which aligns with the months with the most tornado reports. The number of days with both SHERBE and STP beyond their respective thresholds was found to fluctuate interannually. This result warrants further study given the known interannual variability of synoptic and mesoscale weather in Chile. The results of this study extend our understanding of tornado environments in Chile and provide insight into their spatiotemporal variability.

Multi‐Task Learning for Tornado Identification Using Doppler Radar Data

AbstractTornadoes, as highly destructive weather events, require accurate detection for effective decision‐making. Traditional radar‐based tornado detection algorithms (TDA) face challenges with limited tornado feature extraction capabilities, leading to high false alarm rates and low detection probabilities. This study introduces the Multi‐Task Identification Network (MTI‐Net), leveraging Doppler radar data to enhance tornado recognition. MTI‐Net integrates tornado detection and estimation tasks to acquire comprehensive spatial and locational information. As part of MTI‐Net, we introduce a novel backbone network of Multi‐Head Convolutional Block (MHCB), which incorporates Spatial and Channel Attention Units (SAU and CAU). SAU optimizes local tornado feature extraction, while CAU reduces false alarms by enhancing dependencies among input variables. Experiments demonstrate the superiority of MTI‐Net over TDA, with a decrease in false alarm rates from 0.94 to 0.46 and an increase in hit rates from 0.23 to 0.81, highlighting the effectiveness of MTI‐Net in handling small‐scale tornado events.

Predicting the Severity of Tornado Events by Learning a Statistical Manifold for Tornado Property Losses

Predicting the Severity of Tornado Events by Learning a Statistical Manifold for Tornado Property Losses

Assessing the impact and resilience of the community in the aftermath of a small-scale tornado: A field survey-based analysis

Extreme weather can lead to the formation of small-scale tornadoes. Tornadoes are generated by cumulonimbus clouds resulting from convection or orographic lifting of an unstable air mass interacting with a relatively warm air mass within a frontal boundary. The village of Donohudan experienced a small-scale tornado event that resulted in damage to buildings, yards, power infrastructure, and injuries to individuals. This was the first time the community faced a small-scale tornado event, leading to a lack of understanding in handling the aftermath. The aim of this research is to analyze the impact of and resilience to small-scale tornadoes among the population. The research methodology involves field surveys to determine the extent of damage based on the Beaufort scale. The Connor Davidson Resilience Scale (CD-RISC) is employed as a measurement criterion to assess community resilience. The predominant damages were light structural damages and flying roofs caused by the small-scale tornado. Donohudan village demonstrated a moderate level of resilience to small-scale tornadoes, influenced by factors such as flexibility in adapting to changes and challenges, support from family and social networks, spiritual influence, and having a goal-oriented life.

Methods for Validating HRRR Simulated Cloud Properties for Different Weather Phenomena Using Satellite and Radar Observations

Abstract In this study, we evaluate the ability of the High-Resolution Rapid Refresh (HRRR) model to forecast cloud characteristics through comparison of observed and simulated satellite brightness temperatures (BTs) and radar reflectivity during different weather phenomena in December 2021: the Mayfield, Kentucky, tornado on 11 December, a heavy snow event in Minnesota from 10 to 11 December, and the Midwest derecho on 15 December. This is done to illustrate the importance of examining model accuracy across a range of weather phenomena. Observation and forecast objects were created using the Method for Object-Based Diagnostic Evaluation (MODE). HRRR accurately depicted the spatial displacements between observation cloud (defined using BTs) and radar reflectivity objects, namely, the centers of cloud objects are to the east of the radar objects for the tornado and derecho events, and generally west of the radar objects for the snow event. However, HRRR had higher (less intense) simulated BTs and higher (more intense) radar reflectivity than the observations for the tornado event. Simulated radar reflectivity is higher and BTs are lower than the observations during the middle of the snow event. Also, simulated radar reflectivity is higher and BTs are lower than the observations during the derecho event. Of the three weather events, the HRRR forecasts are most accurate for the snow event, based on the object-based threat score, followed by the derecho and tornado events. The tornado event has lower accuracy because matches between paired simulated and observation objects are worse than for the snow event, with less similarity in size forecast objects and greater distance between paired object centers. Significance Statement The purpose of this study is to assess the accuracy of forecast cloud and radar objects, defined using simulated satellite brightness temperatures and radar reflectivity, from the High-Resolution Rapid Refresh (HRRR) model. This assessment was conducted for a tornado, snow, and derecho event from December 2021. Results from these three events indicate that the HRRR model accurately represents the observed displacement between the center of cloud and radar objects for the tornado and derecho events, and is the most accurate overall for the snow event.

Numerical study on wind-loading characteristics of a high-speed train running over the bridge under tornado-like vortices

With global warming intensifying, weather patterns become more volatile and extremes more common. Tornadoes are the most destructive natural disasters causing significant damage to infrastructure. Meanwhile, high-speed railways now face greater risks from tornado events as the national railway network and mass transit trains expand. Thus, studying the tornado flow characteristics and associated effects on high-speed trains is necessary. A study is presented regarding the wind-loading characteristics of a high-speed train running over a railway bridge induced by a tornado belonging to the future railway network. The wind-loading characteristics analyses are performed using the improved delayed detached eddy simulation method. After verifying the numerical approach and mesh strategy, computational studies are conducted to produce a tornado-like vortex and investigate the tornado-induced wind-loading characteristics of a high-speed train running on the bridge by combining a tornado simulation with a moving mesh technique. For the wind-loading parameters studied herein, the selected train's velocity range is between 50 and 350 km/h, the typical operation speed of either regular or high-speed trains. The numerical results show that the time histories of aerodynamic forces on the train revealed a pattern in tornadic flow variability, the time evolutions of the wind loads on the train were affected by train speeds, and the fluctuation was the greatest when the train ran at 50 km/h. Moreover, the train is subjected to larger aerodynamic forces and moments when it operates along with the rotating vortex flow, especially in the core region, and the train is more dangerous when it runs at a lower speed. The results in this study provide references for assessing operation safety, while a train running on the bridge encounters tornadoes.

Tornadoes in Southeast South America: Mesoscale to Planetary-Scale Environments

Abstract A multiscale analysis of the environment supporting tornadoes in southeast South America (SESA) was conducted based on a self-constructed database of 74 reports. Composites of environmental and convective parameters from ERA5 were generated relative to tornado events. The distribution of the reported tornadoes maximizes over the Argentine plains, while events are rare close to the Andes and south of Sierras de Córdoba. Events are relatively common in all seasons except in winter. Proximity environment evolution shows enhanced instability, deep-layer vertical wind shear, storm-relative helicity, reduced convective inhibition, and a lowered lifting condensation level before or during the development of tornadic storms in SESA. No consistent signal in low-level wind shear is seen during tornado occurrence. However, a curved hodograph with counterclockwise rotation is present. The Significant Tornado Parameter (STP) is also maximized prior to tornadogenesis, most strongly associated with enhanced CAPE. Differences in the convective environment between tornadoes in SESA and the U.S. Great Plains are discussed. On the synoptic scale, tornado events are associated with a strong anomalous trough crossing the southern Andes that triggers lee cyclogenesis, subsequently enhancing the South American low-level jet (SALLJ) that increases moisture advection to support deep convection. This synoptic trough also enhances vertical shear that, along with enhanced instability, sustains organized convection capable of producing tornadic storms. At planetary scales, the tornadic environment is modulated by Rossby wave trains that appear to be forced by convection near northern Australia. Madden–Julian oscillation phase 3 preferentially occurs 1–2 weeks ahead of tornado occurrence. Significance Statement The main goal of this study is to describe what atmospheric conditions (from local to global scales) are present prior to and during tornadic storms impacting southeast South America (SESA). Increasing potential for deep convection, wind shear, and potential for rotating updrafts, as well as reducing convective inhibition and cloud-base height, are predominant a few hours before and during the events in connection to low-level northerly winds enhancing moisture transport to the region. Remote convective activity near northern Australia appears to influence large-scale atmospheric circulation that subsequently triggers convective storms supporting tornadogenesis 1–2 weeks later in SESA. Our findings highlight the importance of accounting for atmospheric processes occurring at different scales to understand and predict tornado occurrences.

Climatological occurrences of hail and tornadoes associated with mesoscale convective systems in the United States

Abstract. Hail and tornadoes are hazardous weather events responsible for significant property damage and economic loss worldwide. The most devastating occurrences of hail and tornadoes are commonly produced by supercells in the United States. However, these supercells may also grow upscale into mesoscale convective systems (MCSs) or be embedded within them. The relationship between hail and tornado occurrences with MCSs in the long-term climatology has not been thoroughly examined. In this study, radar features associated with MCSs are extracted from a 14-year MCS tracking database across the contiguous United States, and hazard reports are mapped to these MCS features. We investigate the characteristics of hail and tornadoes in relation to MCSs, considering seasonal and regional variabilities. On average, 8 %–17 % of hail events and 17 %–32 % of tornado events are associated with MCSs, depending on the criteria used to define MCSs. The highest total and MCS-associated hazard events occur from March to May, while the highest MCS-associated portion (23 % for hail and 45 % for tornadoes) is observed in winter (December–February) due to the dominance of MCSs caused by strong synoptic forcing. As hailstone size increases, the fraction associated with MCS decreases, but there is an increasing trend for tornado severity from EF0 to EF3 (Enhanced Fujita Scale). Violent tornadoes at EF4 and EF5 associated with MCSs were also observed, which are generated by supercells embedded within MCSs.

An early indicator index of tornadic storms for Euro-Mediterranean region

Tornadoes are the most violent and destructive of all the severe weather phenomena that localized convective storms produce. There is a requirement in operational meteorology increasing nowadays that an indicator index which allows to reduce the uncertainty of severe convective storms and tornadoes in the scope of climate change adaptation strategies. The main intention is not to replace or substitute mesoscale modeling approaches, or composite indexes, but to warn operationally to draw attention to the Eastern Mediterranean and Türkiye in particular a few days in advance. The development of some indicators using atmospheric variables can undertake a crucial role by enabling such numerical models to be run only at certain time intervals, thus enduring lower computational costs. In this study, Eastern Mediterranean oscillation index (EMEDOi) has been developed in order to be able to detect the presence of ULLs (upper-level low) and frontogenesis approach is employed for selected tornadic storm events in Türkiye. EMEDOi has 7 different its variations (members) which these members have been developed to detect differences depending on the entry directions of cyclones and storms influencing Türkiye from the west of the country. In line with the GDAS data analysis, values of geopotential height are derived for the requirement of EMEDOi in a limited area. A few of the results from the study are as in the following: 86% of the trained tornado events revealed that the EMEDO-Oper index was in negative phase at the time a tornado was reported, regardless of whether the events featured a supercell mesoscale convective storm or a frontal movement. The hourly period until the local minimum is obtained can be described and characterized as the process by which the EMEDO-Oper index value decreases continuously. The time required to reach the local minimum varies based on the tornado occurrence. Based on the tornadic storm scenario in the test cluster in 2022 and the train cluster, this timeframe is predicted to be roughly 33.2 h on average. In western Türkiye, there is a 79% chance of a tornado occurring between six and forty-two hours after the EMEDO-Oper index reaches its local minimum. In particular, the projected chance for this period is 63% between 12 and 30 h after the local minimum is obtained. Besides, the majority of the tornado incidents with EMEDO-Oper values below − 0.75 were evaluated. After an EMEDO-Oper index value falls below that threshold, it is likely to forecast the risk period of a tornado in Türkiye with a probability of 79% and the local minimum point must be identified.

Radar Characteristics and Causal Analysis of Two Consecutive Tornado Events Associated with Heavy Precipitation during the Mei-Yu Season

This paper comprehensively analyzed two consecutive tornado events associated with heavy precipitation during the Mei-yu season (a period of continuous rainy weather that occurs in the middle and lower reaches of the Yangtze River in China from mid-June to mid-July each year) and detailed the formation and development process of the tornadoes using Doppler weather radar, wind profiler radar, ERA5 reanalysis data, ground automatic station data and other multi-source data. The results showed that: (1) Small-scale vortices were triggered and developed during the eastward movement of the low vortex, forming two tornadoes successively on the eastern section of the Mei-yu front. (2) The presence of a gap on the front side of the reflectivity factor profile indicated that strong incoming airflow entered the updraft. Mesocyclones were detected with decreasing heights and increasing shear strengths. The bottom height of the tornado vortex signature (TVS) dropped to 0.7 km, and the shear value increased to 55.4 × 10−3 s−1. Tornado debris signatures (TDSs) could be seen with a low cross-correlation coefficient (CC) value area of 0.85–0.9 in the mesocyclone. The difference between the lowest-level difference velocity (LLDV) and the maximum difference velocity (MXDV) reached the largest value when a tornado occurred. (3) The continuously enhanced low-level jet propagated downward to form a super-low-level jet, and the strong wind direction and wind speed convergence in the boundary layer created a warm, moist and unstable atmosphere in Suzhou. With the entrainment of dry air, the northwest dry jet and the southeast moist jet stimulated the formation of a miniature supercell. (4) The low-level vertical wind shear of 0–1 km increased significantly upon tornado occurrence, which was more conducive to the formation and intensification of horizontal vorticity tubes. Encountering updrafts and downdrafts, the vorticity tubes might have been stretched and intensified. The first lightning jumps appeared 15 min and 66 min earlier than the Kunshan Bacheng tornado and the Taicang Liuhe tornado. The Liuhe tornado occurred during the stage when the lightning frequency reached its peak and then fell back.

Assessment of Tornado Alerting Performance for Canada

ABSTRACT The Northern Tornadoes Project (NTP) completed a first independent assessment of national tornado warning alerting (watches and warnings) in Canada covering the 2019–2021 period. The NTP undertook this study in the spirit of open data, understanding tornado warning issues unique to this country, and improving tornado warning performance. Utilizing the NTP tornado event database for verification, tornado alerts were reviewed for accuracy and timeliness. For the 250 tornadoes that occurred during the study period – and using a definition of what constitutes a warning ‘hit’ developed for the study – the standard 2 × 2 contingency table scores were Probability of Detection = 0.23, FAR = 0.78, and CSI = 0.13. Over 70% of tornadoes had no tornado warning, including 35 EF2 tornadoes. The tornado warning results were compared with US National Weather Service tornado warning scores for the US and US states along the southern Canadian border to provide context. NTP also developed a ‘report card’ aimed at public and media consumption that took into consideration Environment and Climate Change Canada’s national performance targets for tornado warning Probability of Detection (POD) and lead time as well as tornado watch issuance. Using weighted scores for these criteria, NTP assigned a total score of 33.3/100, indicating significant room for improvement. A follow-up assessment was conducted for the 2022 tornado season in Canada following the same established procedures. It was found that the number of both tornado watches and tornado warnings had roughly doubled, resulting in a significant increase in the POD for tornado warnings to 0.35. The report card score also improved to a passing grade of 56.6/100. Further exploration of the results showed enhanced performance for tornadoes that occurred within Doppler radar range, when the parent thunderstorm involved supercell processes, and for tornadoes rated EF2 or higher. A number of recommendations are made aimed at further improvements to tornado alerting performance.

The Impact of Human‐Induced Climate Change on Future Tornado Intensity as Revealed Through Multi‐Scale Modeling

AbstractA novel, multi‐scale climate modeling approach is used to show the potential for increases in future tornado intensity due to anthropogenic climate change. Historical warm‐ and cool‐season (WARM and COOL) tornado events are virtually placed in a globally warmed future via the “pseudo‐global warming” method. As hypothesized based on meteorological arguments, the tornadic‐storm and associated vortex of the COOL event experiences consistent and robust increases in intensity in an ensemble of imposed climate‐change experiments. The tornadic‐storm and associated vortex of the WARM event experiences increases in intensity in some of the experiments, but the response is neither consistent nor robust, and is overall weaker than in the COOL event. An examination of environmental parameters provides further support of the disproportionately stronger response in the cool‐season event. These results have implications on future tornadoes forming outside of climatologically favored seasons.

Analysis of a multiple non-supercell tornado event in Mexico

Multiple tornado events frequently occur in different parts of the world. A region that seems favorable for recurring multiple tornado events due to its geographical location and topography is the eastern sector of the Trans-Mexican Volcanic Belt. A notable multiple tornado event that occurred in this region is investigated. It consisted of five simultaneous tornadoes during the convergence of three air masses: a cold air mass displaced from the north with moderately high relative humidity and specific humidity, a comparatively warm air mass with moderate relative humidity from the southeast, and a hotter air mass with low relative humidity from the southwest. This situation generated a line of intense horizontal wind shear through which the five tornadoes occurred. Numerical simulations revealed three cyclonic vortices and other minor disturbances along the shear line. Vertical cross-sections show rising air in these disturbances suggesting a cyclonic character. The results indicate that the relatively warm mass from the southeast moves over the one associated with a cold front coming from the north, generating multiple cyclonic rotation zones. Experiments with Lagrangian particles revealed vertical movements in the cyclonic rotation zones implying that the maximum heights of the tornadoes varied between 1000 and 1500 m. Moisture flux convergence calculations display structures resembling Kelvin-Helmholtz instabilities. This research proposes a conceptual model where the convergence of masses induced by topography can lead to multiple tornado generation.

Time trends in losses from major tornadoes in the United States

Damage from tornadoes imposes substantial costs on society. This study provides an analysis of time trends in the severity of losses from tornadoes in the United States for the period 1954–2018. Based on information provided by the Storm Prediction Centre (SPC) of the U.S. National Weather Service, we create a dataset of normalized losses from tornadoes spanning 65 years. We then analyse patterns and trends in the total annual losses from tornadoes as well as distributional properties of the damage from individual tornadoes. Our approach allows us to combine observations from the period 1954–1996, when losses from tornadoes were typically reported in a range (e.g. $500,000–$5,000,000) with observations from 1997 onwards when an actual estimate of the damage for an event is provided. Our findings suggest an overall national significant decline in normalized losses from tornado events. At the country level, both the severity of damage from individual events and the total annual losses from tornadoes are seen to have reduced over time. We also find spatial variations in time trends for the damage from tornadoes: while for most U.S. states the declining trend in severity is confirmed, an increasing trend of total annual losses from tornadoes is observed for Alabama.

Probabilistic assessment of concurrent tornado and storm‐related flash flood events

AbstractThe concurrent occurrence of extreme events can significantly impact societies and infrastructure systems, especially when the instructions provided to the exposed communities demand conflicting responses. In this study, a probabilistic assessment of concurrent tornado and storm‐related flash flood (TORFF) events is performed across southern Canada. We quantify the interdependencies between tornadoes and flash floods (extreme precipitation as proxy) using ground‐based and reanalysis datasets. Windspeed values corresponding to tornado events, categorized based on the recorded Fujita rating, are derived through a resampling approach. The TORFF events are clustered and the corresponding bivariate probability distributions of the resampled windspeed and associated precipitation are characterized based on Copula framework. The individual and joint return periods of concurrent tornadoes and flash floods are then assessed under the AND (when both variables exceed predefined thresholds), OR (when either one of the two variables exceeds predefined thresholds), and conditional hazard scenarios across Canada. Results show positive strong dependencies between resampled windspeed and associated precipitation in Saskatchewan, and weaker dependencies followed by Alberta, Manitoba, Ontario, and Quebec. The Saskatchewan region shows the highest risk underestimation considering the independence scenario compared to the other regions. Higher precipitation is also expected during extreme windspeed, as observed in the conditional assessment of precipitation given windspeed. This study provides insights for more robust recurrence interval estimation for tornadoes and flash floods to aid in emergency planning of evacuation decision‐making process.

The Hunt for Missing Tornadoes: Using Satellite Imagery to Detect and Document Historical Tornado Damage in Canadian Forests

Abstract Historical tornado events from 1982 to 2020 were documented within Canada’s forested regions using high-resolution satellite imagery. Tornado forest disturbances were identified using a three-step process: 1) detecting, 2) assessing, and 3) dating each event. A grid of 120 km × 120 km boxes was created covering Canada (excluding the extreme north). Of the 484 boxes, 367 were manually searched. Once a long, narrow region of tree damage was detected, it was first cross-referenced with known tornado databases to ensure it was a unique event. Once events were classified as either tornadic or downburst, the coordinates of the start, worst damage, and end locations were documented, as well as the direction of motion, damage indicators, degree of damage, estimated maximum wind speed, and F/EF-scale rating. In total, 231 previously unknown tornadoes were identified. In Ontario, 103 events were discovered, followed by 98 in Quebec, 9 in Manitoba, 6 in Saskatchewan, 9 in Alberta, 5 in British Columbia, and 1 in New Brunswick. The largest number of discovered tornadoes occurred in 2015, and the largest number of strong F2 tornadoes occurred in 2005. Most of the discovered tornadoes occurred in July for both F/EF1 and F/EF2 ratings. Most tornado tracks had widths between 200 and 400 m, and more than 50% of the tornadoes had a pathlength of less than 10 km. Of all the events that were discovered, 125 events could be fully dated, 19 were dated only by month, 41 were dated only by year, and 46 remained undated.