Science1 publisher2 min readPublished
A survey in four flood-hit places finds the warning gap in the messenger and the channel
Researchers surveyed nearly 900 people in Belgium, Germany, Poland and Florida about the flood information they wanted and what they got. They said they were getting enough information. The gaps they found sat in the source and the channel.
The Scientist · Science desk
What happened
- One evening in August 2025 a Pakistani shepherd, Abuzar Hussain, saw a glacial flood heading for Shijar, climbed until his phone found signal, and the village mosque broadcast the alert over its loudspeaker.
- The researchers surveyed people in Belgium, Germany, Poland and Florida, all hit in 2024 by Storm Boris's central European flooding or by Hurricane Helene in the southeastern United States.
- Across nearly 900 respondents, people in all four countries said they were getting enough information about the emergency, and the shortfall was in who it came from and how it reached them.
- The surveys also picked up gaps in the accuracy of information about the timing and location of flooding, though the source of that information was the larger problem.
- Trust ran high for emergency responders such as firefighters and police, and relatively low for social media influencers and religious leaders.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- constraint A void score measures what one person expected minus what that person recalls receiving, so a low score leaves open whether any message went out.
- decision An agency reading this faces a spending choice between buying more broadcast volume and building a roster of messengers its population already trusts.
- precedent Scoring a warning system by source and channel invites agencies to be judged on whom they used and how it arrived.
A void score is a subtraction. Each respondent rates the flood information they would ideally have had, then rates what actually reached them at the same sitting, on four things: how much, how accurate, from whom, and by which channel [8]. The difference between the ideal rating and the received rating is the void, and the bigger the difference, the bigger the void [9].
The reason to ask about sources at all comes from earlier work, which the authors summarise: the sources that deliver information during emergencies, and their perceived credibility, can affect how people understand risks and act on them [19]. What their instrument records is narrower: what people wanted and from whom, and what they say arrived [8].
Nearly 900 respondents across four survey sites averages roughly 220 each [17]. The article reports the four countries together [10]. It also says the source was the largest of the four gaps without putting a number on it [11].
Religious leaders came near the bottom of the trust list in Belgium, Germany, Poland and Florida [12]. The warning that reached Shijar travelled on a mosque loudspeaker [3]. Both cases turn on a nearby messenger on a channel that does not need the mobile network to be standing. All four survey sites sit in Europe and the United States [5].
That matters because storm damage to telecommunications and transport can leave emergency responders unable to push updates out through news or social media at all [14]. Rumours move into the space, often about emergency supplies, and some communities come away feeling ignored [15]. The authors cite studies suggesting that misinformation outbreaks are usually preceded by a vacuum of information from official channels [16]. The same outage that stops the official update also stops the correction.
For the fragility of the source-audience relationship under load, the authors point to the flash flood on the Nepal-Tibet border in August 2026, which washed away villages and roads and killed more than 1,000 people [13]. Abuzar Hussain's phone call is the case on the other side, and it is one call: hundreds of lives in one valley, in a region where millions of dollars have gone into early warning systems [4]. Whether a smaller void score goes with more people actually leaving their houses is a question this design, as described, leaves open [9].
What to watch
- Whether the team publishes per-country void scores and the item wording, which would show how far the source gap outran the volume gap.
- Whether a follow-up pairs void scores with observed behaviour, such as evacuation compliance or shelter records.
- Whether the instrument is run in South Asia, where the standing of local religious and community intermediaries may rank differently.