Jason C. Senkbeil

WHO'S GOOGLING WHAT? What Internet Searches Reveal about Hurricane Information Seeking

Two freely available, searchable databases that track the normalized interest in specific search queries, Google Trends and Google Insights, were used to illustrate spatial and temporal patterns in hurricane information seeking. Searches for the word “hurricane” showed a seasonal pattern with spikes in hurricane searches that corresponded to the severity of the storms making landfall. Regional variation in “hurricane” searches was largely driven by the location and magnitude of hurricane landfalls. Catastrophic hurricanes such as Hurricane Katrina captured national attention. A great deal of regional variation in search volume existed prior to Hurricane Ikes landfall. Not as much variation was seen before Hurricane Gustav and Tropical Storm Fay. This variation appeared to be related to changes in the 5-day track forecast as well as other factors such as issuance of watches and warnings. Searches from Louisiana experienced a sharp decrease after the 5-day track forecast shifted away from the state, but be...

Shelter Seeking Plans of Tuscaloosa Residents for a Future Tornado Event

AbstractThe enhanced Fujita scale category 4 (EF4) Tuscaloosa, Alabama, tornado on 27 April 2011 produced 64 fatalities along its 130-km track. Hybrid survey/interviews were conducted with a sample of 211 Tuscaloosa-area residents to determine how the 27 April tornado might change future shelter-seeking plans. Despite a history of tornadoes in the area, only 47% of Tuscaloosa residents had shelter plans in place prior to 27 April, but 62% intend to change their shelter plans or have shelters plans for the future. Changes in shelter-seeking plans were divided into four groups and discussed according to commonalities. Logistic regression with demographic variables was then used to predict those likely to have shelter plans before 27 April and those likely to change their shelter plans in the future. Among these variables, residents over age 55 [odds ratio (OR) 8.9, 95%; confidence interval (CI): 2.167–36.352] and those having a bachelors degree (OR 5.1, CI: 1.342–19.316) were more likely to have had shelte...

A Postlandfall Hurricane Classification System for the United States

Abstract The Saffir–Simpson scale is useful for evaluating maximum sustained hurricane winds and storm surge over open water in the prelandfall window, but it fails to accurately account for the observed impacts over land. A new postland-fall hurricane classification system (HCS) is proposed that redistributes the categorization of hurricanes into types according to six variables: open water storm surge, rainfall, duration of hurricane force winds, maximum sustained winds, gust score, and minimum central pressure. Hurricanes are assigned values for each variable and summed for a numerical 0–100 grade. Principal component analysis and hierarchical cluster analysis are also performed on the six variables to categorize U.S. land-falling hurricanes into storm types. A total of 41 land-falling hurricanes in the United States since 1960 have been analyzed. The summation scores show many hurricanes that are of strength similar to their Saffir–Simpson classifications, with several notable exceptions. The cluster analysis identifies five different hurricane types. These types can be arranged to identify hurricane strength and structure more effectively than the Saffir–Simpson scale. In focusing on the observed storm intensity over land and the resulting human experience, the HCS allows a way to compare hurricane impacts across different periods.

Hydrometeorological application of an extratropical cyclone classification scheme in the southern United States

Extratropical cyclones (ETCs) in the southern United States are often overlooked when compared with tropical cyclones in the region and ETCs in the northern United States. Although southern ETCs are significant weather events, there is currently not an operational scheme used for identifying and discussing these nameless storms. In this research, we classified 84 ETCs (1970–2009). We manually identified five distinct formation regions and seven unique ETC types using statistical classification. Statistical classification employed the use of principal components analysis and two methods of cluster analysis. Both manual and statistical storm types generally showed positive (negative) relationships with El Niño (La Niña). Manual storm types displayed precipitation swaths consistent with discrete storm tracks which further legitimizes the existence of multiple modes of southern ETCs. Statistical storm types also displayed unique precipitation intensity swaths, but these swaths were less indicative of track location. It is hoped that by classifying southern ETCs into types, that forecasters, hydrologists, and broadcast meteorologists might be able to better anticipate projected amounts of precipitation at their locations.

Facebook and Twitter, communication and shelter, and the 2011 Tuscaloosa tornado

This paper represents one of the first attempts to analyse the many ways in which Facebook and Twitter were used during a tornado disaster. Comparisons between five randomly selected campus samples and a city of Tuscaloosa, Alabama, sample revealed that campus samples used Facebook and Twitter significantly more both before and after the tornado, but Facebook usage was not significantly different after the event. Furthermore, differences in social media usage and other forms of communication before the tornado were found for age, education, and years lived in Tuscaloosa. Generally, age and education were inversely proportionate to social media usage. Influences on shelter-seeking actions varied between social media users and three random samples of non-social media users; however, it appears that social media respondents were likely to be using a smartphone simultaneously to access warning polygon information, to receive text message alerts, and to listen or respond to environmental cues.

Ethnic and Racial Differences in Tornado Hazard Perception, Preparedness, and Shelter Lead Time in Tuscaloosa

The 27 April 2011 EF4 Tuscaloosa tornado directly impacted more than 50,000 residents, causing forty-five fatalities within the city and sixty-five in total. It was a rare urban tornado with varying impacts on the three major ethnic and racial groups within the city. A hybrid survey and interview of open-ended and closed questions was conducted with 211 Tuscaloosa area residents in a two-week period after the tornado. Results indicate significant differences in risk perception, preparedness, and shelter lead time among the three ethnic and racial groups. Furthermore, results were still significant for perception after controlling for the effects of age, education, and experience.

Suggestions for alternative tropical cyclone warning graphics in the USA

Purpose – The cone of uncertainty (COU) warning graphic has created confusion for people trying to make evacuation and safety decisions. The purpose of this research was to create several alternative tropical cyclone graphics and present them to the public and college students via face‐to‐face surveys and polling.Design/methodology/approach – Surveys depicting hypothetical landfall scenarios were administered in Pensacola and Jacksonville, FL. Respondents ranked five graphics in order of preference, and were encouraged to discuss their rankings. Following this initial field research, the most popular graphic of these five was compared to a graphic resembling the one used by The Australian Bureau of Meteorology. Comments were recorded for respondents favoring or disliking the Australian graphic in two separate analyses. A final graphic emphasizing post‐landfall hazards was also created as a suggestion for future research and evaluated directly against the most popular graphics from field research.Findings ...

A Tornado Watch Scale to Improve Public Response

A tornado refuge rubric was revised into a six-level, hierarchical Tornado Watch Scale (TWS) from level 0 to level 5 based on the likelihood of high or low-impact tornadic events. Levels correspond to an estimate of the maximum potential tornado intensity for a given day and include refuge/sheltercategories of ‘‘adequate,’’ ‘‘questionable,’’ or ‘‘inadequate,’’ which encompass a range of refuge/shelter locations taken from the Enhanced Fujita scale. Ratings are based on a conservative estimate of damage indicators in high winds and the safetyofapersontakingrefugeinsidebuildingsofvaryingstructuraldesign.Audiorecordingssimilartothose used in current NOAA weather radio communications were developed for each TWS intensity level. Recordings representing an existing tornado watch, existing particularly dangerous situation (PDS) tornado watch, and three proposed levels from the TWS were then used in interviews with Alabama residents to determine how changes to the information contained in the watch statements would affect each participant’s tornado safety actions and risk perception. Participants were also questioned about their knowledge and past experience with tornado hazards and their preference between the existing NWS tornado watches and the TWS. Results indicate a strong preference for the TWS when compared to existing products. The TWS was favored for providing additional information, containing descriptions of expected severity, and being easy to understand. The TWS also elicits more adequate safety decisions and more appropriate risk perception when compared to existing products, and these increases in safety were statistically significant.

Implications of the 2011 Tuscaloosa EF4 tornado for shelter and refuge decisions

This research provides an overview and discussion of language used in tornado safety recommendations along with development of a rubric for scaled tornado safety recommendations. Residents living in affected areas and those temporarily housed at relief stations were surveyed to collect information on their experiences during a 2-week period following the April 27, 2011 Tuscaloosa, Alabama EF4 tornado. Respondents were asked about their refuge plans during the storm and about any future changes to those plans. A specific focus of this research evaluated the adequacy of each respondent’s plan. Each refuge plan was compared using a tornado refuge rubric developed through the use of enhanced Fujita (EF) scale degree of damage ratings for available damage indicators. There was a significant difference in the counts of refuge adequacy for Tuscaloosa residents when holding the locations during the April 27 tornado constant and comparing adequacy ratings for weak (EF0–EF1), strong (EF2–EF3), and violent (EF4–EF5) tornadoes. There was also a significant difference when comparing the future tornado refuge plans of those same participants to the adequacy ratings for weak, strong, and violent tornadoes. This research introduces renewed discussion on proper refuge and shelter alternatives for days when violent tornadoes are forecasted.

Changes in Summer Weather Type Frequency in Eastern North America

In this research, the Spatial Synoptic Classification (SSC), a weather type scheme, is used as an alternative method of demonstrating evidence of climate change in the Eastern United States and southern Canada. Changes in frequencies for the seven SSC weather types were assessed for summer trends (May–September) at thirty-eight stations and also at four regions of latitude between 1950 and 2015. Using the SSC, results show significant summer decreases in dry polar (DP) days and transitional (TR) days and significant increases in moist tropical (MT) days. The North region exhibited the greatest breadth of significant results among all weather types. The DP and TR decline was strongest at higher latitudes and weakened approaching the subtropics. The MT gain was strongest across the midlatitudes but statistically significant in all four regions. The four remaining SSC weather types showed more localized statistically significant trends. Results suggest that these trends in weather type frequency are an indicator of summer climate change, with some stations losing over 50 percent of their DP frequency, losing over 40 percent of their TR frequency, and gaining over 30 percent of their MT frequency since 1950.