The Weather Observations Website (http://wow.metoffice.gov.uk/) was created in 2011 with the objective of increasing the number of weather observations shared amongst citizen scientists. The website has far surpassed expectations and has seen over half a billion observations submitted by around 10,000 amateur observers globally. The success of WOW has encouraged the Bureau of Meteorology in Australia, MetService in New Zealand and KNMI in the Netherlands to implement their own portals into the website and they have reported great success in extending their reach and creating partnerships with citizens and other agencies. A new version of WOW, the WOW Engine, is currently in development which aims to take this successful formula and extend it to a more flexible, agile and adaptable data management platform. Using API technology it will be possible to quickly and easily ingest new sources of observational data which will be managed, stored and visualised through a variety of channels. The WOW Engine will incorporate the ability to capture complex metadata as default and will comply with the WMO WIGOS Metadata principles, thus allowing users to benefit from the potential of WIGOS to create a ‘network of networks', a truly Integrated Observing System.
As agencies are expected to extend their warning capabilities away from traditional threshold based weather warnings and towards impact based warnings, there is also a need to gather evidence of impacts. WOW contains this capability and will in the future be capable of mining data from social media and other live sources.
Harnessing the power of citizen scientists is potentially a game changer for meteorology as the increasing resolution of NWP models is not matched by a corresponding increase in the density of traditional observing networks. The citizen scientist with an app on their phone, or in their car or home, can provide supplementary observations that will provide useful additional detail to the modellers and forecasters. A key challenge is how to manage the balance between quantity and quality of these observations and to identify the most effective ways to use this kind of data.