Rip currents are among the deadliest meteorological hazards in the United States. In order to demonstrate the proper actions to take when caught in a rip current in a memorable and engaging way, a virtual reality video game is developed.
Beachgoers in the United States face many different hazards, but rip currents are annually the deadliest for ocean swimmers. Despite the risk presented by rip currents, it is apparent that the public has a limited understanding of their danger and the proper mitigating actions to take when caught in one. A virtual reality (VR) video game placing participants in a simulated rip current was developed to help ameliorate this issue. The VR game was used to survey beachgoers on the Atlantic Coast of Long Island, New York during July and August 2019. The actions participants took when confronted with the rip current were recorded, along with whether they escaped it or drowned. An interview with each player was also conducted after they partook in the game to determine the realism of the rip current simulation and its effectiveness in demonstrating proper actions to take when impacted by one. Analysis of those results indicates that VR has potential to communicate rip current risk and ways to minimize it in a unique and engaging manner. However, further work is needed to improve ease of use of the VR simulation and better understand how factors such as demographics influence perceived rip current risk and behavioral response.
Rip currents are “strong, narrow flows of water that extend away from the beach1.” Rip currents can commonly occur on any beach with breaking waves and can transport swimmers quickly away from shore. Hazardous rip currents can occur on seemingly ‘safe’ beach days with wave heights of only 2 to 3 feet2, and thus can surprise swimmers as they are carried a considerable distance from shore. This puts swimmers at risk of panic, exhaustion, and even drowning. As a result, rip currents are one of the leading causes of weather fatalities in the United States. For example, in 2018, 71 deaths were attributed to rip currents, and for the 10-year period 2009-2018, an average of 58 individuals perished each year3. Rip currents are the leading hazard for beachgoers; in 2018, rip current deaths represented 65% of all ‘surf zone’ fatalities in the United States. There seems to be some demographic control on rip current vulnerability, as one study found that men are over six times more likely than females to drown from rip currents than females4. Moreover, additional research found that infrequent beach users are more likely to make poorer beach safety choices5 and that non-locals are considerably more likely than locals to sustain injury in the surf zone6,7.
Nevertheless, despite their place among the deadliest weather hazards in the United States, rip currents are poorly understood by the public. A survey of 392 public beach users in Texas determined that only 13% could correctly identify a rip current from photographs presented to them8, while similar results were found in studies conducted in Pensacola Beach, Florida9 (15%) and Miami Beach, Florida10 (27%). More broadly, Houser et al (2017)5 performed an Internet-based survey with 1622 respondents across 49 of the 50 US states and found that 54% of the participants correctly reported an action to take when caught in a rip current. However, the self-selected nature of the survey sample dictated that only 10% of the sample was infrequent beach users, who are most vulnerable to rip currents and were shown in the survey to possess less knowledge of what to do in one.
It is clear that rip currents present a unique challenge, given that they are poorly understood by the public, can occur suddenly over small scales with minimal or no prior warning, and may result in death. Thus, new approaches are needed to address this public safety challenge. Immersive technology such as virtual reality (VR) provides an innovative approach to increase rip current literacy and encourage positive behavior upon impact. Prior research has indicated that VR and similar types of immersive media are highly effective at communicating information. VR is generally defined as an interactive experience taking place within a simulated environment that incorporates auditory and visual feedback, usually with the assistance of a headset. One recent study11 asserted that VR is a mature technology, well-suited to assist in the scientific inquiry process. Moreover, other recent research12 showed that when individuals read a New York Times story with a VR supplement, they were more likely to perceive the source as being credible, recall the information presented, share it with others, and feel an emotional connection, than those who read the article in traditional media, with just text and graphics. Additional studies13,14 concluded that immersive media promotes education by increasing engagement and real-world applicability of a topic. Most recently, researchers15 leveraged VR to simulate a Category 3 hurricane landfall and determined that survey respondents viewing the VR were significantly more likely to consider evacuation than those only viewing traditional text and graphic products. Despite its clear utility, no studies or initiatives have shown comprehensively how VR can be effectively applied to the unique challenge of training beach users to better locate and react to rip currents. The present work fills that research gap by first teaching individuals how to swim and wave in a virtual ocean environment and then evaluating how they react to the sudden and unwarned onset of a rip current. Participants were trained in both swimming and waving for help because each of those actions are regarded as valid responses when caught in a rip current16,17, with conditions particular to an individual rip often dictating which action might be most effective in facilitating escape18. We hypothesize that the realistic and memorable nature of a VR rip current simulation will allow participants to successfully take evasive action in the virtual game and then report that the experience enhanced their knowledge of rip current risk and mitigation.
All methods used were approved by the Hofstra University Institutional Research Board (IRB). The VR video game developed was used to survey 64 individuals.
NOTE: Scripts were written in the C# language, and are available for download at: https://github.com/Jasebern/HofstraVR.
1. Creation of VR rip current video game: Virtual environment and user input/output
2. Survey individuals with the VR rip current video game
The VR rip current video game survey was conducted in July and August 2019 on Long Island at the Town of Hempstead Beach in Point Lookout, New York (detailed results can be found in Supplementary Table 1-3). 64 individuals played the game and responded to the survey, with 60 escaping the rip current and 4 drowning (i.e., stamina reached zero). Among the 60 who escaped, 51 exited by waving for help, and 9 did so by physically swimming out of the rip. Given that most participants waved for help as a means of escape, the amount of time taken in the simulation was skewed towards higher values, with a mean of 11.1 s, median of 9.5 s, and standard deviation of 6.2 s (Figure 6a). Conversely, the ending stamina was skewed towards lower values, with a mean of 36.8, median of 41.3, and standard deviation of 15.3 (Figure 6b). Most participants were able to accurately assess the situation and determine an appropriate course of action to escape the rip current relatively quickly. However, there was a slight uptick in ending stamina closer to zero (i.e., between 0 and 12). That finding may have been caused by the heavy breathing (which began playing when stamina fell below 20) helping individuals realize they were in more imminent danger, and as a result, they changed their strategy and were able to escape before stamina reached zero.
After completing the video game, participants were asked a series of dichotomous, Likert, and open-ended questions regarding the VR and its effectiveness. There were 51 respondents to the Likert-scale question (scale of 1 to 5 with 5 being highest) asking if they felt better prepared for a rip current after interacting with the VR. The average response was 3.81, with a minimum of 1, maximum of 5, and standard deviation of 1.01. Moreover, 61 individuals responded to a similar Likert-scale question asking about how immersive the VR experience was, with an average of 3.96, minimum of 2, maximum of 5, and standard deviation of 0.79. Participants were also asked if they had been caught in a rip current prior to playing the game, and if so, how the VR compared to real life. 17 individuals responded to the latter question, with 7 stating that the simulation bore at least some resemblance to real life. 7 of the respondents found that the VR was not as realistic or scary as real life, while 4 asserted that it was not at all similar.
In addition, participants were provided with a set of six brief statements meant to capture their opinion of the VR experience and asked which they most agreed with (Table 2). Of the 58 respondents to that question, 53 selected statements stating that the VR helped them feel better prepared for a rip current, with only 5 selecting ones saying it did not help. 30 of the 58 chose the statement that said the VR helped them feel better prepared because it was realistic, and 19 chose the one stating it helped because it was scary or made them felt worried. Last, users were asked to identify the most and least useful aspects of the VR, along with any suggestions for improvement. 19 individuals provided useful aspects of the simulation, with the most common being its realism (6), the inclusion of VR (3), instruction provided (3), and ability to wave (3). Conversely, 6 respondents reported the least useful aspects, with 3 of the 6 mentioning the short duration of the game as a negative. Accordingly, regarding the improvements, there were 19 responses, with 13 suggesting an expansion of the simulation such as more scenarios, additional training, or more options.
Figure 1. Main Menu scene. Opening scene of the VR experience. Please click here to view a larger version of this figure.
Figure 2. Buoy test scene. First training scene in the VR experience. Please click here to view a larger version of this figure.
Figure 3. Wave test scene. Second training scene in the VR experience. Please click here to view a larger version of this figure.
Figure 4. Rip current scene. User evaluation scene in the VR experience. Please click here to view a larger version of this figure.
Figure 5. VR hardware setup screen. Demonstrates proper configuration for connecting virtual reality equipment to a computer. Please click here to view a larger version of this figure.
Figure 6. Results of rip current virtual reality video game (A) Bar chart showing ending stamina of all participants (B) Bar chart showing time taken for all participants. Please click here to view a larger version of this figure.
Script Name | Script Function | Scenes Used |
MainMenu | Controls menu buttons | Main Menu |
MouseHover | Controls highlighting of menu buttons | Main Menu |
PlayerController | 1) Stores user input from Oculus controller joystick | Main Menu, Buoy Test, Wave Test, Rip Current |
2) Stores user input from head movements in Oculus headset | ||
PlayerMotor | 1) Physically move the player in the environment (i.e., swim) based on input from PlayerController script | Buoy Test, Wave Test, Rip Current |
2) Rotates the camera view based on input from PlayerController script | ||
PlayerMotor2 | 1) Inherits and extends functionality of PlayerMotor | Buoy Test, Wave Test, Rip Current |
2) If player is swimming, plays a sound of arms splashing in water | ||
For Rip Current Scene Only: | ||
3) Applies a constant drifting motion to player away from the beach to simulate being pulled away from shore in a rip current | ||
4) Creates and tracks ‘Stamina’ variable based on a timer and user input; Stamina begins at 60 and decreases by 1*second if player is stationary and by 3*second if player is swimming | ||
5) Creates a timer variable which tracks time elapsed in Rip Current Scene | ||
6) Assigns each user a unique player number based on date and sequential player of that day | ||
7) If player waves, print player number, current Stamina, time elapsed, and player condition (‘Waved’) in a text document; transition to Main Menu (also transitions from Wave Test scene to Rip Current scene) | ||
8) If player escapes rip current, print player number, current Stamina, time elapsed, and player condition (‘Escaped’) in a text document; transition to Main Menu | ||
9) If player stamina reaches zero, print player number, current Stamina [0], time elapsed, and player condition (‘Drowned’) in a text document; transition to Main Menu | ||
Checkmark | If player swims between buoys, entering the collider box, transition to next training scene (Wave Test) | Buoy Test |
FloatObject | The water has no collider, meaning the player should fall straight through the water due to gravity. This script simulates floating to keep the player at the level of the water. | Buoy Test, Wave Test, Rip Current |
FemaleAnimate | If player presses the ‘A’ or ‘X’ button on the Oculus controller, initiates a hand waving animation in the left player arm, and plays an audio clip of a hand splashing water | Wave Test, Rip Current |
Buyoancy2 | Records whether or not the hand is waved in a scene and if in Rip Current scene, record in PlayerMotor2 script that player waved | Wave Test, Rip Current |
RipExit | 1) If player exits rip current collider box, record in PlayerMotor2 script that player escaped the rip current | Rip Current |
2) If stamina is below 20, start playing heavy breathing audio emanating from the player |
Table 1. Scripts developed for project. Scripts were written in the C# language.
Positive Statements |
i) The VR experience helped me feel be better prepared because it was realistic. |
ii) The VR experience helped me feel be better prepared because it was scary/or made me feel worried. |
iii) The VR experience helped me feel be better prepared because it taught me what to do. |
Negative Statements |
i) The VR experience did not help me feel be better prepared because it was not realistic. |
ii) The VR experience did not help me feel be better prepared because it was scary and/or made me feel worried. |
iii) The VR experience did not help me feel be better prepared because it did not teach me what to do. |
Table 2. Brief statements regarding rip current virtual reality experience. Participants were asked to select which one they most agreed with.
Supplementary Table 1. Individual VR simulation results. Please click here to download this file.
Supplementary Table 2. Aggregated survey demographic results. Please click here to download this file.
Supplementary Table 3. Selected post-VR interview results. Please click here to download this file.
Preliminary analysis of the results from the follow-up survey demonstrates the VR rip current video game was generally effective in accurately portraying risk and demonstrating proper actions to take in an engaging and memorable way. Respondents to the Likert-scale questions indicated the VR simulation resulted in them feeling more prepared than not for a rip current and also that it was fairly immersive. Moreover, the results of choosing one of six brief statements clearly showed the video game was helpful given that over 90% of the selections were positive. Similarly, in free response questions, numerous participants lauded the VR for characteristics such as its realism and interactivity. The overall player outcomes of the video game also underscored the efficacy of the experience at conveying proper actions to take in a rip current. 60 of the 64 participants successfully escaped the rip, most of them by waving for help, and the majority also took an evasive action quickly.
Some feedback also indicates that improvements can be made to this and future VR simulations developed. Indeed, more instruction might be necessary, especially for individuals with less experiencing playing video games and using VR. Additional optional training scenes are one possibility for allaying those concerns. Further, realism can always be improved to make the VR experience more relatable and meaningful for participants. To do so, enhancements such as physically moving one’s arms to swim (instead of using a joystick controller) and further distinguishing the rip current from the surrounding ocean water can be incorporated.
The results of the survey also provide unique insight into individual behavioral responses to a rip current scenario. For example, 51 of the 64 participants were able to escape the rip current by waving for help. However, in the follow-up survey, only 20 of those participants stated that waving or calling for help was the preferred action to take in the rip current. It is possible some of the inconsistency in knowledge versus action can be explained by the survey order, as the instructions for waving always occurred just before the rip current simulation, which may have predisposed some individuals to wave for help in escaping the rip. Thus, randomizing the order of the training scenes could allow for even more realistic results in the future. It is also possible, however, that the intensity (i.e., feeling of being quickly pulled from shore) and rapid onset of the simulated rip current caused individuals to either forget, or be deterred from attempting, a more complicated, riskier evasive action: swimming parallel to shore. This is further confirmed by the fact that 20 respondents mentioned swimming parallel (or ‘sideways’) to shore was a proper action to take, but only 9 participants escaped the rip current in that manner.
Moreover, the gap between rip current knowledge and action, and resultant personal risk, was demonstrated by individuals believing they knew the proper response, but then performing an incorrect one. Four participants drowned (i.e., stamina reached zero) in the simulation, despite all four of them stating afterwards that they knew want to do in a rip current. Three of the four, however, reported an incorrect evasive action, with the fourth also showing a limited understanding of what to do, mentioning that they should swim ‘away out of it,’ but not specifying a direction to swim. Similarly, 45 of the 64 survey respondents affirmed they would know if they were caught in a rip current. Of those 45, however, 10 clearly demonstrated in their responses that they did not actually know what a rip current was, confusing it with a phenomenon such as an ‘undertow’ that pulls individuals underwater and may involve large waves. Thus, the combined results of the VR simulation and survey indicate two primary obstacles in rip current risk communication: 1) Some individuals do not know what a rip current is, or have incorrect knowledge of a rip current, and therefore may not take the proper mitigating action, and 2) When caught suddenly in a rip current, even individuals who know what to do in one may forget or disregard those actions, potentially exposing themselves to risk.
Future research can expand upon this work in order to better understand how sociodemographic factors influence personal mitigation of rip current risk. For instance, 57 of the 64 participants in the present survey reported living within 30 minutes of the beach, while 54 stated they visited at least ‘occasionally.’ However, many rip current fatalities involve individuals residing far from the beach who may only visit once a year or less for vacation. Future surveys can be conducted at more neutral locations or online to gain a broader sample and understand behavioral differences in rip current reaction between those who visit the beach more and less frequently.
Undoubtedly, VR has the unique ability to allow users to conceptualize risk and learn proper mitigating actions in a memorable way. Improved understanding of its current shortcomings, especially as they pertain to certain demographic characteristics, will allow researchers and emergency managers to capitalize on immersive technology in a nuanced way and develop the next generation of effective warning products.
The authors have nothing to disclose.
This publication is a product resulting from NYSG project R/CHD-14 funded under award NA18OAR4170096 from the National Sea Grant College Program of the U.S. Department of Commerce’s National Oceanic and Atmospheric Administration, to the Research Foundation for State University of New York on behalf of New York Sea Grant. The statements, findings, conclusions, views and recommendations are those of the author(s) and do not necessarily reflect the views of any of those organizations.
Dell 17.3" Alienware 17 R5 Laptop | Dell | PC for virtual reality development | |
Oculus Rift S | Oculus | Virtual reality headset |