The weather of the future will be predicted not only by meteorologists - but also by citizens

In Croatia, citizens are set to play a key role in shaping future weather forecasts, moving beyond the traditional domain of meteorologists. This shift is driven by a growing network of real-time observations made by ordinary people, supported by advanced technology and data analysis tools. The concept, known as crowd-based forecasting or “crowdnowcasting,” has evolved from a theoretical idea into a practical approach with the potential to significantly enhance local weather monitoring and early warning systems. The initiative began over a decade ago, when a Croatian researcher developed the idea of using public reports as part of a broader meteorological observation system. Today, thanks to artificial intelligence, large data sets, and increased connectivity, this vision is becoming a reality. With nearly every citizen carrying a smartphone equipped with GPS, cameras, and constant internet access, there is now a vast potential for real-time weather reporting. These contributions can provide valuable insights into localized weather phenomena that conventional systems might miss. Traditional meteorological networks rely on automated stations, radar, satellites, and numerical models to generate accurate forecasts. While these systems offer high-quality data, they are limited in their ability to capture every local event. For example, a summer downpour could affect one village while remaining dry just a few kilometers away. Similarly, thunderstorms may devastate one orchard but leave neighboring fields untouched. In such cases, human observers become crucial. When multiple individuals report similar conditions, such as the start of rain, lightning, or strong winds, these observations can create highly detailed maps of atmospheric changes. Sometimes, these reports arrive faster than official systems can confirm the same phenomenon. This capability allows for more precise and timely updates, especially in areas where infrastructure is sparse or where extreme weather events occur unexpectedly. The future of meteorology will be two-way, with citizens contributing data and receiving refined forecasts and warnings in return. However, individual reports alone are insufficient. They must be validated against radar data, satellite imagery, station measurements, and predictive models. Artificial intelligence plays a vital role here, helping to identify unreliable or false reports and distinguishing those that align with actual atmospheric conditions. Croatia’s national meteorological hub would serve as the central node of this evolving system. It would integrate diverse sources of information, including citizen reports, radar data, satellite feeds, automatic weather stations, storm detection systems, numerical models, IoT sensors connected to the internet, social media analyses, and AI-driven algorithms. These IoT devices, small, automated instruments that measure temperature, humidity, pressure, rainfall, wind speed, or soil moisture, are already used in agriculture, smart cities, vineyards, orchards, private weather stations, and industrial settings. Integrating them into a unified meteorological network would effectively add thousands of virtual weather stations to Croatia's existing infrastructure. This comprehensive system would go beyond current capabilities, offering dynamic visualizations of atmospheric changes minute by minute. It would particularly improve early warnings for severe weather events such as thunderstorms, heavy rains, flash floods, hail, dense fog, and other localized hazards. The National Hydrometeorological Service (DHMZ) is considered the most logical entity to lead the development of such a hub, given its existing resources and expertise in operating radar and automated weather systems.