Traditionally weather forecasts have always been broad-casted by the government agencies and departments like Met office or in the USA NOAA. They used supercomputers to crack satellite pictures, powerful radars and data collected in a more physical manner.
However, in recent times a hoard of new tech startups has started broadcasting weather prediction which in most cases are more accurate than the forecast predicted by the traditional agencies. These startups are using cheaper components, cloud computing power to forecast weather more accurately.
Components or Hardware much cheaper than the traditional agencies use
According to CEO of Spire, the weather observation systems are under tremendous pressure and inadequate to predict the accurate weather forecast in a time when global warming has caused the weather to go haywire and more accuracy is needed in predicting the weather.
Spire, which started its operation in 2012, has its own arrangement of 58 small CubeSat satellites in every position of globe monitoring the moisture level of the air along with air pressure and temperatures. It sells data accumulated through its system to big weather agencies like NOAA.
People are actually starving for accurate and detail weather information but due to inadequate input, the traditional system cannot cope up with the demand. The reason for that is those traditional systems are still using technologies which are as old as their components.
Combining meteorological data with radar and satellite data
CimaCell which was founded in 2015, uses meteorological data gathered from communication networks like mobile phone networks and put these together with data collected from radar and satellite to produce a more accurate result.
CimaCell is trying to sense the environment in multiple ways, their sensors can mark rain and snow even though radar might not. After collecting and collating all the data they broadcast a weather forecast for six hours.
Other startups like Earth Networks, Jupiter, Understory and Riskpulse are also combining and collating data gathered from multiple sources and using analytics to produce new kind of weather reports.
There are a number of startups in the business of weather forecast
The reason for this sudden boom is first, the cheaper but more capable components which are a result of the revolution in mobile phone technology initiated by iPhone in 2007.
The ground-based sensors were developed along with satellites much smaller than their traditional size. For example, CubeSat with a standard size of 10cm3 modules developed in 1999.
Secondly, giants like Google or Amazon supplies abundant computing power in the cloud which makes the small startups to handle a complex task of predicting weather much easier.
