Impact-based severe weather warnings in New Zealand: Survey data

(Inc. GST)
(Ex. GST)
Write a Review

Potter, S.H.; Kreft, P.V.; Milojev, P.; Noble, C.; Montz, B.; Dhellemmes, A. 2017 Impact-based severe weather warnings in New Zealand: Survey data. Lower Hutt, N.Z.: GNS Science. GNS Science report 2017/01 36 p.; doi: 10.21420/G2001G

Abstract: From September to October 2015 we conducted an online survey of members of the public to investigate their perceptions of a hypothetical severe weather event involving strong winds and their stated intended protective actions. Impact-based severe weather warnings (that is, warnings based on the potential impacts of the weather event) are increasingly being used by meteorological agencies instead of more traditional phenomenon-based severe weather warnings (which are warnings based on wind speed, for example, regardless of the potential effects of the event). The survey data were tested to see whether impact-based severe weather warnings differed significantly from phenomenon-based severe weather warnings in terms of prompting protective actions, and to check for demographic differences in survey responses. The survey had 1364 responses from self-selected participants. The survey results indicate that impact-based warnings may be more effective than phenomenon-based warnings in influencing the recipients’ perception of the hazardous event. In particular, recipients of impact-based warnings (a) reported finding it easier to understand the possible effects of the weather event, (b) believed that the wind would be more threatening, and (c) were more concerned about the wind event, in comparison to those who received the phenomenon-based warning. There was no significant difference in perceptions about the credibility of the message. In terms of intending to act on the information, significantly fewer respondents reported that they were less likely to do nothing different, and more likely to intent to seek further information, when they received the impact based warning. However, there was no significant difference in participants stating that they intended to take many of the other protective actions as a result of receiving the warning message. Perceptions of threat, concern, and credibility of the message influence taking protective actions, as do demographic variables and the local context. This report presents the results of the survey. (auth)