Upper-level troughs and ridges drive much of midlatitude weather, from storms to heatwaves, but tracking them automatically has been difficult. We trained an artificial intelligence model on examples created by expert meteorologists, producing a tool that detects these features more accurately than earlier methods. We used it to build a long-term record of these features, improving understanding of Mediterranean weather.

The North Atlantic region serves as a source of moisture and energy for Mediterranean storms. Its impact over the Levant region remains an open question due to its smaller weather systems and their longer distance from the ocean. We find an optimal circulation pattern which allows North Atlantic influence to reach farther into the eastern Mediterranean, thus making storms stronger and rainier. This may be relevant for future Mediterranean climate, which is projected to become much drier.

Cyclone detection and tracking methods (CDTMs) have different approaches in defining and tracking cyclone centers. This leads to disagreements on extratropical cyclone climatologies. We present a new approach that combines tracks from individual CDTMs to produce new composite tracks. These new tracks are shown to correspond to physically meaningful systems with distinctive life stages.