Sequences of large-scale meteorological patterns associated with wildfire spread events in the Northwestern United States

Background Large-scale meteorological patterns (LSMP) have been linked to variation in fire weather, wildfire ignitions, and burned area. However, the influence of multi-day sequences of these patterns on wildfire spread events remains unclear. Aims This study classifies multi-day LSMP sequences and assesses their influence on the likelihood of wildfire spread events in the Northwestern United States. Methods We employ a two-stage approach using self-organizing maps to classify multi-day LSMP sequences from 500 hPa geopotential heights. We then estimated the impact of these sequences on wildfire spread likelihood using regression analysis on 2003–2020 regional fire data. Key results Certain LSMP sequences, including persistent ridging, characterized by positive height anomalies over the region, are preferentially associated with wildfire spread events. Conclusions Classifying multi-day LSMP sequences using self-organizing maps can potentially aid in forecasting wildfire spread events. Implications Fire weather meteorologists can apply this approach to identify and forecast conditions associated with potential wildfire spread events, enabling better coordination.