SOME people in Fiji just stubbornly refuse to evacuate when disaster warnings are issued.
Well, that problem is not unique to any one country.
From New York, in the US, to Tokyo, Japan, national weather forecasters and television commentators revealed they were looking at ways to address the disconnect between forecast data and human behaviour.
While weather officials in New York have found, through experiences with Hurricane Katrina and Sandy, that personalised announcements were a more effective means of communicating with the people, Japan's National Meteorological Agency, after the Tohoku Earthquake of March 11, 2011, has decided to "overestimate" the magnitude of earthquakes measuring close to eight on the Richter scale. And the nation's only public broadcaster, NHK, after the death of more than 20,000 in that 3/11 quake, has also decided to make changes and "shout" when earthquake and tsunami alerts were announced — hoping the strategy would jolt the normally calm Japanese into action.
Speaking in New York to journalists who are visiting disaster struck areas in the US, Japan, and China, National Weather Service director eastern region Jason Tuell said they were working with social scientists and the media to learn how to communicate weather forecasts more effectively.
He said forecasters needed to learn how to translate technical information that the public found too complex and difficult to navigate.
A self assessment by the National Oceanic and Atmospheric Administration found that during Sandy, there were serious problems communicating the risk to those who needed to know.
The consensus now is to emphasise what a storm does, more than what to call it.
"People want to know when and how long before it arrives, right now where I am standing, when is the water going to hit my toes, how deep it will get and when my feet will get dry again," Mr Tuell explained.
He referred to a personalised message by New Jersey meteorologist Gary Szatkowski, which helped the Sandy evacuation process.
Mr Szatkowski asked people to evacuate and think about loved ones, and the stress placed on emergency responders and rescue and recovery teams when loss of life was involved.
In another incident, Mr Tuell said an emergency responder got people to take warnings seriously when he distributed toe tags at evacuation zones, asking them to put it on so they could be identified if they did not survive the ordeal.
He said the assessment found that while the NWS, as the federal forecaster, had produced an accurate tracking of Sandy five days before landfall, the "division of responsibilities" when the characteristics of the storm changed and who should make announcements caused a lot of "confusion".
In total, the storm was responsible for 147 direct deaths, the destruction of about 650,000 homes, at least $50billion (F$92.52b) in damage and left about 8.5 million without power.
Mr Tuell said the assessment also found that the National Hurricane Centre's homepage, which did not contain storm bulletins, was highly trafficked.
Warning policies has since been tweaked, officially making the centre a one-stop-shop for online information about storms.
"Our information might have been too technical. We need to understand how people perceive information and we are looking to social scientists to advise on how we can communicate information.
"Weather forecasting is not an exact science but there is enough suggestion and data that as a country we have to be better prepared.
"We also need to work closely with the media, we rely on them to get the information out, they are skilled at communicating."
Takeshi Koizumi, from JMA's earthquake and tsunami division said while they worked with advanced technology, the 20,000 death toll in the Tohoku quake "stood out" and revealed the importance of evacuation during a tsunami.
Therefore, the content of the first warning was crucial, he said. The first 3/11 warning announced a quake of 7.9 but the strength was revised to 8.8 three minutes later when rupture was completed.
Mr Koizumi said it was not until 25 minutes later that a sudden rise of sea level about 30 kilometres offshore was noted.
"The first warning must not mislead the people," he said.
"The power went out after the earthquake in many places so even if we updated, they would not have gotten the information.
"Some people took the coastal barriers for granted and we found there were very few casualties where people evacuated.
"We have decided to overestimate earthquakes whose reading is around eight or more in the hope that people would have already evacuated by the time the real data is updated. And we don't want to give tsunami heights, we will just say 'big tsunami coming'.
"We need to do this because even though we have intense public awareness, 20,000 were killed."
Japan has since 3/11 updated its equipment, enhanced seismic stations and installed offshore sensors.
Mr Koizumi said ultimately, effective disaster mitigation in any country required public and individual participation.
"One must take all measures possible when warnings are issued."