
Developers of a new weather forecasting app are confident the powers of artificial intelligence (AI) will help deliver the 'holy grail' of weather information to growers, providing hyper-local forecasts.
Jane Bunn, a prominent weather forecaster with a lengthy history in television, primarily in Victoria, is launching the Jane's Weather app in conjunction with the AgriFutures growAG platform designed to help developers of novel agricultural products link up with investors.
AgriFutures is a research and development corporation that has unique overarching responsibilities for big picture research, including start-up businesses.
Jane's Weather will operate by providing growers aggregated information from a number of the top global weather models into one platform, eliminating the need to compare across different weather apps and sites, with a long-term view to provide accurate information down to a kilometre by kilometre grid level.
Ms Bunn said bringing together all available data under one umbrella is just the first step, with a view to bridging existing gaps in the information landscape.
"Relying solely on traditional weather models is no longer sufficient to power precision agriculture," she said.
"Most farms have some sort of weather network setup, but they aren't located near a bureau station, so traditional weather models don't incorporate their data into forecasts."
"We're working on bridging that gap by using data direct from a user's farm to supplement those holes in the current weather modelling and create a more accurate picture.
"The latest update to Jane's Weather is incorporating machine learning and artificial intelligence to customise forecasts to each site's unique terrain, giving farmers the best data to inform things like spraying windows, crop growth predictions, frost mitigation and irrigation scheduling."
To unlock this capability, Jane's Weather is seeking investment of up to $1.5 million from the investment community via growAG.
AgriFutures growAG engagement manager, David Lord, said the app presented an exciting opportunity for investors looking for scaleable technology that aims to solve a fundamental problem facing every agriculture industry.
"Farmers constantly have to make high value decisions based on the weather, so they need the most accurate information," he said.
The funding will be used to scale up machine learning from local observation networks to tailor the forecast to individual properties, and bring the forecast range down to a 1 x 1km grid.
Ms Bunn said the end goal was to be able to map the wildly variable thunderstorms that can see dramatic differences in rainfall in very short distances.
"Showers and thunderstorms are a huge part of our weather here in Australia, but these beasts are not that big - less than a kilometre wide," Ms Bunn said.
"It's amazing when you really pinpoint weather data and see that one area is getting drenched and another close by is completely dry.
She said the 1 x 1 grids would be at least nine times more detailed than what is currently widely available.
"The top weather models currently forecast on a 9 x 9km or 12 x 12km grid, with some models even up to 50 x 50km, and that's just not accurate enough to know what's happening across your farm."
"We need funding to access the remaining data which will enable high resolution forecasts on a 1 x 1km grid."
Ms Bunn said Jane's Weather could drive productivity gains operationally.
"With hyperlocal forecasting and AI-driven predictions the app can say 'you won't be able to do that job until 3pm with today's weather, so you can go and do something else in the meantime."