We spoke with researchers from the Meteorology and Climate department to explore the state of the art in this field.
Meteorological research has been central to CIMA Research Foundation since its inception: accurate precipitation forecasting is key to effective flood prevention, a phenomenon that has always been a core focus of our research. Like any scientific discipline, however, meteorology evolves over time. What challenges does it face today?
We discussed this with Elena Oberto and Massimo Milelli, a researcher and the head of Meteorology and Climate department, respectively.
Local forecasts, global forecasts: how to integrate them?
“One of the most characteristic aspects of today’s meteorological forecasting models is their global scale,” they explain. “Increasingly, efforts are directed towards models capable of analyzing the dynamics of the entire planet. These models rely on coupling different components, integrating data related to the atmosphere, land, and ocean. However, for our work—focused on protecting communities from the most severe events—large-scale models are valuable only when combined with models providing much more localized forecasts.”
Localized models provide detailed information on where and how intense rainfall might affect an area. Civil protection efforts require two elements: a large-scale forecasting model (not necessarily global, but, for example, European) to estimate the arrival and potential development of a weather system. Then, a second step identifies exactly where the most rainfall will occur, and the related risks for the area. Which part of a city will be affected? Is there a river nearby? Is flooding a concern?
In addition to their spatial resolution, the models used by CIMA Research Foundation are also faster than large-scale models. “Speed is another critical element of civil protection forecasting,” Oberto and Milelli explain. “The sooner we have information on the evolution of a meteorological event, the sooner we can implement measures to address it. The goal of a good Early Warning System is exactly that: to rely on fast, accurate, and updated forecasts.”
Nowcasting: Ultra-Short-Term Forecasting
For this reason, much of CIMA Research Foundation’s research focuses on nowcasting, a term that refers to forecasting techniques for a very short time horizon, between 0 and 6 hours. “While longer-term forecasts would ideally provide more time to deploy risk mitigation strategies, we must consider that the uncertainty of model outputs inevitably increases as the forecast period lengthens,” the researchers note.
Models consist of equations that, starting from initial conditions (the current state of the atmosphere), use the physical laws governing atmospheric motion to predict future states with a degree of uncertainty. Since civil protection requires the most precise data possible, this field focuses on models offering short-term forecasts.
“Much of our work focuses on making nowcasting predictions increasingly detailed and reliable. A key tool in this effort is data assimilation techniques—systems to ‘update’ the data fed into the model as a meteorological phenomenon evolves,” Oberto and Milelli explain. “We are particularly focused on an algorithm we developed a few years ago, called SWING. It allows post-processing of data by combining the results of the forecast model (based on an initial atmospheric conditions dataset, wind patterns, etc.) with recent observational data, such as radar rainfall estimates and lightning activity. The result is highly reliable six-hour forecasts.”
To further improve reliability, researchers are working to integrate the SWING algorithm with different forecasting models. “Using results from multiple models could create an even more precise and reliable ‘multi-model’ system.”
New data, new forecasts
New data sources, especially satellite data, are also enhancing the reliability of meteorological models. Innovative tools that provide precise observations of atmospheric conditions from space represent a valuable resource. “In November, we began a project called Hydroterra+, for which we are the scientific coordinators. The project aims to test data that could potentially be acquired during a new ESA satellite mission,” Oberto and Milelli add.
Exploring new data sources is not the only focus of the Meteorology and Climate department at CIMA Research Foundation. While the importance of short-term forecasts for civil protection is clear, long-term forecasts also have their place. “Floods aren’t the only meteorological phenomena involving water. Drought is becoming an increasingly significant risk due to the climate crisis,” the researchers conclude. “In water resource management, seasonal forecasts—spanning several months—are becoming particularly important. For this reason, in some international projects, we are also conducting studies in this field in collaboration with other research departments at CIMA Research Foundation.” We will share the results of these efforts in the coming months.