Description:

The proliferation of connected weather sensors continues at an ever increasing rate. This includes the addition of IoT sensors, and the connection of traditional sensor networks to the internet. Together, the body of collected data is often referred to as Community Weather, and is comprised of sensor owners that include individuals, private companies and national weather services who contribute their data to centralized repositories.

On the surface, the build-out of the Community Weather Network might appear to be ad hoc, but clear strategies are evident in how the Weather Enterprise is incentivising targeted IoT sensor deployment. In addition, these IoT sensors are being connected in a diverse manner as cellular, satellite and other data networks expand, become cheaper and more accessible. This means that IoT sensors are spreading into remote regions without the need to connect to the wired internet. As such, the Community Weather Network is not only becoming more dense, but is spreading into untapped remote regions. This has large implications on how the collective datasets may be used by the Weather Enterprise.

The volume of observations contributes to the big data challenge, but opens the door to using AI and machine learning (ML) techniques. The data collected from these sensors is dense, but varies wildly in quality and consistency. Despite this, these sensors are finding diverse applications in the Weather Enterprise. Organizations are finding ways of identifying the valuable data using AI/ML. However, the contribution of high-quality metadata from the Community Weather Network remains a significant challenge.