Genesis Conditions and Extratropical Cyclone Intensity in the North Atlantic and Europe: A Statistical and Machine Learning Analysis

Extratropical cyclones (ETCs) produce most of the day-to-day weather variability in Europe and can cause potentially damaging winds and precipitation. Therefore, it is of special societal interest to understand the atmospheric conditions which lead to impactful ETCs. In this work, we build on well-established theories of ETC intensification by using statistical and machine learning methods to quantify relationships between atmospheric conditions at time of ETC genesis, here called precursors, such as baroclinicity and potential vorticity, and ETC intensity measures, such as relative vorticity and wind speed. To capture full variability in ETC intensity, we analyse multiple measures to quantify the intensity from both dynamical and impact-relevant perspectives. The aim of the study is to identify the precursors which control the intensity of ETCs and quantify how changes in their pattern and magnitude relate to responses in the intensity. A further aim is to investigate if impactful ETCs are associated with specific patterns in their precursors.

ERA5 reanalysis data from 1979 to 2022 between October and March was used to track ETCs in the North Atlantic and Europe and to analyse five intensity measures for each ETC. The response of these measures, namely 850-hPa relative vorticity, 850-hPa wind speed, wind footprint, precipitation, and a storm severity index, to changes in ETC precursors analysed from ERA5 was quantified using ensemble sensitivity analysis (ESA). Sensitivity patterns were calculated for an ensemble of all tracked ETCs, and the precursors which were found to have the most control on ETC intensity were identified. These precursor fields were then classified with a self-organizing map (SOM), and correspondence between the precursor classes and ETC intensity was investigated to identify precursor states potentially leading to impactful ETCs.

The ESA indicated that large values in wind-based intensity measures were associated mostly with a stronger jet stream, especially downstream of the ETC and tilting northeastward, as well as a larger meridional temperature gradient, with an emphasis on warming on the equatorward side at upper levels and cooling on the poleward side at lower levels. Precipitation response was controlled by one-sided changes in temperature and moisture, with warming and more moisture especially in the warm sector of an ETC being associated with more precipitation. Potential vorticity anomalies at both upper and lower levels were found to have little control on ETC intensity. The SOM classification was able to detect precursor states which were associated with impactful ETCs at a statistically significant level, and these states were consistent with signals from the ESA. These results can be used to study the occurrence of intense ETCs in future climate projections and to better forecast potentially impactful ETCs.