Graduation Year

2020

Document Type

Dissertation

Degree

Ph.D.

Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Geography

Major Professor

Jennifer Collins, Ph.D.

Committee Member

Kamal Alsharif, Ph.D.

Committee Member

Robin Ersing, Ph.D.

Committee Member

Rebecca Morss, Ph.D.

Committee Member

Yujie Hu, Ph.D.

Keywords

interviews, mixed methods, risk perception, scenario, surveys

Abstract

A weather radar display is a tool that provides spatially oriented, timely information about an impending weather event. While radar is frequently used by meteorologists, emergency managers, and pilots, this tool is now readily available for individuals to use on a variety of platforms including television, computer/laptop, smartphones and tablets. Most importantly, there are hundreds of mobile weather applications available as well as online sources that provide a weather radar display. However, little is known about how individuals use a weather radar display. Therefore, the purpose of this dissertation is to understand why radar is sought out as a tool and how useful it is perceived to be as a source of weather information.

This research uses a mixed methods approach to answer the many unknowns surrounding weather radar use. An online survey collected 510 responses from residents within seven counties within the Tampa Bay Area. Overall a weather radar display was found to be a very useful tool. Survey responses were analyzed to determine several factors that increased the overall usefulness of a radar display including age, wealth, and gender (female). Respondents who said that weather radar provided them with enough information for decision making, that trust radar data, that were more weather salient, and found the accuracy of the location of radar data to be greater were more likely to find a radar display to be a useful tool. Respondents reported locating a hazard watch or warning and the location of precipitation as being the most important information provided by a weather radar display. This survey also determined that a smartphone was used most often to view a radar display over television, computers, or tablets.

As smartphones were found to be used most often, the respondents’ preferred mobile weather applications were evaluated. Mobile weather applications were also grouped by the type of information that was displayed first in the app and by how prominent the radar display was in each app. Findings showed that the average age of the user was significantly different between specific apps. More men than women used apps that solely served as a radar display or map focused display, while women preferred apps that delivered a forecast and/or current conditions first. Specific features of each mobile weather app were analyzed to discover what users liked most about their preferred mobile weather apps. Finally, a visual comparison of seven mobile weather apps showed several differences in layout, colors used to display reflectivity values, and legend types. This showed major differences for how each weather radar display looks, including the way reflectivity values are contoured and smoothed.

The second phase focusses on understanding the construal of situational risks and outcomes and applies several theories from social psychology and geography to address research objectives. Using radar involves interpreting space and time while simultaneously evaluating meteorological attributes (reflectivity etc.). Using six radar-based scenarios (three severe and three non-severe), 30 participants took part in semi-structured interviews in order to test how they perceived a radar display. Findings showed that participants found a radar display to be less useful during weather events where directionality was unclear or stationary. Radar was described most often as a tool used to anticipate what will occur in the near future. This study also reveals several possible misconceptions for what participants thought reflectivity values display such as inferring lightning or wind when ‘red’ or ‘orange’ reflectivity values were present. Time was also overestimated in most scenarios. Finally, the broader impacts, limitations, and future research are discussed.

Download

Share

COinS