Abstract
Compound cold extreme weather events—co‐occurring multivariate events—have been defined as either cold‐dry (CD) or cold‐wet (CW) depending on the absence or presence of heavy precipitation. Both event types induce varying levels of social and economic impacts across multiple sectors such as health, transport and energy depending on which type of event is experienced. In this study, we characterise these CD and CW events in the United Kingdom (UK) using a location‐specific percentile approach and assess their relationship with a set of 30 UK‐specific weather patterns to determine the event drivers. The results show that there are up to 14 CD days per winter season in the west of the study region compared to 4–8 CD days in the east. The inverse is shown for CW with 0–1 days per winter season in the west and 2–3 days in the east. CD events are predominantly driven by anticyclonic weather patterns (which are classified in the negative North Atlantic Oscillation regime), and CW days are driven by cyclonic weather patterns. This study provides evidence that a location‐specific approach alongside weather pattern analysis could be adopted as a tool for impact‐based forecasting at a medium‐range lead time to forecast CD and CW events.