New method to detect and quantify weather persistence associated with hydro-climatic extremes

Changes in weather persistence are of particular concern in the context of climate change as periods of longer persistence can reinforce weather extremes. In our study we apply structural image recognition methods to global ERA5 reanalysis data to identify when, where and how long isolines of atmospheric geopotential height fields run in similar tracks. We identify regions and episodes around the world in which, retrospectively, unusually long-lasting weather patterns repeatedly occurred. Concerning the temperature and precipitation meteorological fields, we derive a connection between the occurrence of weather persistence and hydro-climatic extreme events.

Based on our new method we find that weather persistence has been particularly increasing in Northern Hemisphere mid-latitudes in summer confirming earlier studies. Here, highly populated regions like Central Europe have experienced long-term increases in persistent weather conditions of up to 4-5% between 1981 and 2019 amplifying the risk of prolonged heat waves and droughts. Further, we show that climate models tend to have difficulties in capturing the dynamics of weather persistence and thus may severely underestimate the frequency and magnitude of future extremes events in their climate projections.