Extraordinary 2021 snowstorm in Spain reveals critical threshold response to anthropogenic climate change

The attribution of extreme weather events to anthropogenic climate change (ACC) is a growing area of research. However, the impact of ACC on heavy snowstorms remains largely unexplored. This study focuses on the exceptional January 2021 snowfall event in Spain, known as Filomena, which caused widespread chaos and significant economic damage. Public discourse following the event suggested a link between Filomena and global warming, prompting this research to investigate the validity of this hypothesis. Two primary arguments underpin this hypothesis: (1) Arctic amplification leading to a wavier jet stream and more frequent cold air outbreaks; and (2) a warmer atmosphere holding more moisture, potentially intensifying snowfall. However, existing literature presents conflicting views on ACC's influence on the jet stream and snowfall. This study aims to unravel the relationship between global warming and the Filomena snowstorm by considering both dynamic and thermodynamic drivers, using a novel approach combining circulation analogs and a storyline attribution tool in a common framework. Unlike probabilistic attribution methods, this approach is suitable for examining snow events due to the limitations of climate models in accurately simulating snow extremes.

Previous research has explored the attribution of extreme rainfall events to ACC, but studies on extreme snowfall are far less numerous. Some studies suggest Arctic amplification and a wavier jet stream could increase the likelihood of Filomena-like events. Others focus on the thermodynamic aspect, highlighting the capacity of a warmer atmosphere to hold more moisture, potentially intensifying snowfall if temperatures remain low enough. Conversely, other studies argue against a significant impact of ACC on the jet stream and propose that temperature increases have led to reduced snow. The conflicting findings underscore the need for further investigation into the relationship between global warming and extreme snowfall events.

This study uses a novel approach combining circulation analogs and a storyline attribution tool. First, the flow analog technique was used to identify historical circulation patterns similar to that of Filomena, using daily mean sea level pressure (MSLP) and geopotential height at 500 hPa (Z500) data from the 20th Century Reanalysis (20th-CR.V3) and ERA5 reanalysis. This helped determine if the frequency of Filomena-like atmospheric patterns has increased with warming. Secondly, a storyline attribution approach, using Pseudo Global Warming (PGW) simulations with the Weather Research and Forecasting (WRF) model, was employed to assess the influence of ACC on snowfall intensity. Counterfactual simulations for pre-industrial and future (SSP5-8.5 scenario) climates were generated by perturbing the WRF model's initial and boundary conditions with anthropogenic climate signals extracted from five CMIP6 climate models. These simulations perturbed thermodynamic variables such as temperature, dew point, skin temperature, sea surface temperature, and specific humidity. An ensemble of 40 simulations (8 WRF physical configurations × 5 CMIP6 models) for past and future climates, and 8 simulations for the present climate, allowed for an assessment of uncertainty. Spectral nudging was employed to prevent large-scale dynamic drift from ERA5 reanalysis. The simulated snowfall was validated against AEMET observations and reanalysis data.

The analysis of Filomena's flow analogs revealed no significant long-term trend in the frequency of similar atmospheric patterns, suggesting ACC didn't increase the likelihood of such events. However, the PGW simulations showed a significant and uneven impact of ACC on snowfall intensity. In northwestern highlands, ACC intensified snowfall by up to +40% compared to pre-industrial conditions and more than +60% in future projections, whereas in southern lowlands, it weakened snowfall by up to -80%, and by -100% in some areas for future climate. This contrasting pattern is strongly linked to a critical temperature threshold: above the threshold, warming leads to reduced snowfall due to the phase change of precipitation to liquid; below the threshold, increased atmospheric moisture intensifies snowfall. Analysis of three specific cities (Soria, Madrid, and Ciudad Real) showcased this pattern. Soria (north) showed increasing snowfall, while Ciudad Real (south) experienced drastically reduced snowfall. Madrid, situated near the threshold, displayed a slight increase in current snowfall compared to pre-industrial conditions, but future simulations indicate a decrease. Overall, while the total amount of snow in the study area decreased with warming, the heterogeneity in the impact is noteworthy. The study also found that the critical temperature threshold that separates the regions of snowfall intensification and weakening, applies at both the climatic (decades) and event (days) scales. Using ERA5 data, the study estimated the threshold for the northern hemisphere to be approximately -5°C at the climatic scale and -1°C at the event scale for Filomena.

The findings demonstrate a complex and spatially heterogeneous response of the Filomena snowstorm to ACC, highlighting the significance of thermodynamic influences. While the frequency of the triggering atmospheric pattern remained stable, ACC strongly altered the snowfall intensity and distribution. The contrasting effects of warming on snowfall are explained by the combined impact of increased temperature and humidity, exceeding a crucial temperature threshold. The results confirm that climate models accurately simulate this threshold effect at both climatic and event scales. The study provides a general rule for assessing ACC's thermodynamic impact on snowfall, based on the relationship between current temperature and the freezing point. This includes scenarios where ACC may have little or no effect, decrease or increase snowfall. The study’s findings are potentially extrapolatable to other extreme snowfall events due to their basis on fundamental thermodynamic principles.

This study reveals the complex and uneven impact of ACC on extreme snowstorms, demonstrating significant intensification in high-elevation and high-latitude areas, while simultaneously causing weakening or elimination of snowfall in lower elevations and latitudes. The critical temperature threshold defining this contrasting response is a key insight. Future research could focus on broader applications of this threshold-based framework to analyze other extreme snow events globally and refine the accuracy of climate models in simulating snow extremes.

The study focused on a single extreme snowfall event (Filomena). While the findings' grounding in fundamental thermodynamic principles suggests broader applicability, further research is needed to confirm the generalizability of these conclusions across various geographical locations and event types. The uncertainties associated with climate model projections and the reliance on a specific regional climate model (WRF) also need to be considered.