Detail Matters: Drivers’ Responses to Flood Warnings in Simulated Driving

Abstract

Flood risk communication is imperative to aiding people’s decision making in flood situations. These warnings can be communicated through navigation applications on mobile devices. The current study investigated how flood-depth information affected drivers’ actions given flood warnings from a mobile navigation application in a driving simulator. This study manipulated the type of flood warning presented to the participants in the driving scenarios and measured their actions given a potentially flooded roadway. Participants experienced six drives with different flood warning conditions. Results indicated that providing flood depth information helped drivers accurately estimate the depth of the flood and their perceived risks; including more detailed information was helpful for drivers to make informed decisions regarding a flooded roadway. We suggest that designers include flood depth information to help drivers accurately perceive the depth and risk regarding a flooded roadway.

Flood risk communication is imperative to aiding people’s decision-making in flood situations (Covello et al., 1986). Risk communication can be at different levels of granularity or detail to meet the users’ needs (Chen et al., 2018; Jorgensen et al., 2015). Flood risk communication conveys the hazard of a flood and helps people make informed decisions regarding it (Mileti, 1995). These warnings can be communicated through applications on mobile devices, such as weather or navigation applications (Cumiskey et al, 2015). However, there is a lack of research on how flood warnings should be effectively presented to drivers from a mobile navigation application (see Hamilton et al., 2020). Our previous study examined how the varied amount of information presented in flood warnings affected drivers’ decisions in a simulated driving study (Garcia & Chen, 2023). However, the range of flood depths used in that study did not exceed a typical vehicle’s ground clearance; if flood water reaches or exceeds the ground clearance, it can damage the vehicle (Gerhardt, 2019).

To fill this gap, the current study included a wider range of flood depths and further investigated factors that affect drivers’ decisions given flood warnings from a mobile navigation application in a simulated driving scenario. A total of 85 participants were recruited. This study manipulated the type of flood warning (flood, no flood, flood of 2 inches, flood of 5 inches, flood of 5 inches maximum, flood of 8 inches; within-subjects) presented to the participants in the driving scenarios. The dependent variable was drivers’ decisions given a potentially flooded roadway. The drivers could continue straight and drive through a potential flood or accept the alternate route provided by the GPS and turn before encountering the flood. Participants first took the Motion Sickness Susceptibility Questionnaire (MSSQ; Golding, 2006) to determine if they had a high likelihood of motion sickness. Eligible participants started with a practice drive on a driving simulator equipped with the STISIM Drive. Afterward, they read the first scenario regarding the vehicle they were driving, answered comprehension questions, and began their first experimental drive in the simulator. Toward the end of the drive, participants received one of the six types of flood warnings and acted before answering questions about perceived risk. Each participant experienced all six flood-warning conditions, with the order of the conditions counterbalanced among participants. After the last cycle of the simulated drive and questions, participants answered demographic questions before being granted research credits and dismissed.

Our results indicated that participants were significantly more risk-avoidant when the warning indicated the presence of any flood on the roadway. Similar results were found for the perceived flood depth and the perceived risks. Additionally, the percentage of risk-avoidant actions increased as the flood depth in the warning increased; participants were the most risk-avoidant in the 8-inch flood condition, then the 5-inch flood condition, and lastly, the 2-inch flood condition. Interestingly, participants acted similarly when given a general flood warning and when they received the flood-of-8 inches warning, accepting the alternate route more often than in the other three flood warnings providing the depth of the flood. However, there were significant differences in the perceived flood depth and the perceived risk of driving through the flood between these two conditions. This may indicate a ceiling effect of how perceived risk affects behavior, where their actions are always the same once perceived risks reach a certain level.

This study provides insights into drivers’ mental models regarding flood warnings and how they perceive a flood when either the depth information is provided or not, and when the maximum depth information is provided. This study can inform mobile navigation application designers about how much information to include in flood warnings. We suggest that designers include flood depth information to help drivers accurately perceive the depth and risk regarding a flooded roadway.