New modelling being launched this week is trying to improve how the Pacific's cyclone season is predicted.
Cyclone outlooks have been a thing for years - they try and predict how many cyclones there may be, and how strong they could be - based on factors such as ocean temperatures and whether there's an El Nino or not.
But their accuracy has been hit-and-miss, and they're only released about a month before cyclone season starts in November.
A team of researchers - including New Zealand's National Institute for Water and Atmospheric Research - have been working on a way to generate predictions to a country level months in advance.
The lead researcher is Dr Andrew Magee, from the University of Newcastle in Australia.
He told Jamie Tahana that the modelling could help countries improve early warnings and disaster planning.
An earlier forecast of 2020 cyclones Photo: SPREP
