Wildfire spread is influenced significantly by weather fluctuations. Specifically, wind speed and direction (wind vector) change can drastically alter fire intensity and spread. Wind vector change can result from synoptic or mesoscale weather systems and small-scale meteorological processes, such as thermal circulations and low-level wind shear in complex terrain. These small-scale processes are usually underrepresented in numerical weather forecasting models, usually needing to be resolved by more expensive sub-kilometre grid resolution simulations. Recent New Zealand wildfires, such as 2017 Port Hills and 2019 Pigeon Valley wildfires, experienced wind change due to local sea breezes and strong nocturnal downslopes flows, exacerbating fire behaviour. The aim of this research is to investigate if vector wind change (VWC) can be a metric to better represent the effect of mesoscale and microscale weather and the subsequent impact in extreme fire behavior. This was achieved utilizing hourly VWC, or the difference in magnitude of the hourly wind vectors, from the Fire and Emergency New Zealand (FENZ) national network of more than 200 weather monitoring stations. An additional variable, wind direction change (WDC), was also calculated to include the degree change of wind direction. To identify high-risk stations, the top 20% of stations for VWC and WDC were calculated and investigated spatially across the New Zealand landscape. The high-risk stations are located primarily on the South Island, inland and in areas of complex terrain. There is little to no variation of these stations when mapped in each synoptic type, suggesting that the main factor in determining high VWC and WDC is meso and microscale terrain driven meteorology as opposed to larger synoptic regimes. Critically, the current fire risk metric, the Fire Weather Index, does not include wind direction changes at high windspeeds. Therefore, the inclusion of VWC and WDC as additional metrics in fire risk calculations could increase operational understanding of high-risk locations and terrain impacts on extreme and unpredictable fire behavior.

Background. Wildfire spread is influenced significantly by the weather variability. Wind speed and direction changes, resulting from synoptic weather systems and small-scale meteorological processes in complex terrain, can drastically alter fire intensity and spread. Aims. To investigate the use of vector wind change (VWC) integrated with the Fire Weather Index (FWI) as a new metric in fire danger. Methods. A 20-year FWI and modified FWI was calculated from weather station and gridded numerical weather simulation data. Key results. High VWC is found primarily on the South Island, inland and in areas of complex terrain. After incorporating VWC into the FWI, data from the modified FWI show spatiotemporal patterns that highlight the impact of wind variability in the fire danger. Conclusions. High VWC station data mapped with synoptic type suggest the primary factor in determining high VWC is meso-and micro-scale terrain-driven meteorology, not larger synoptic regimes. Implications. The current fire danger metric, the Fire Weather Index (FWI), does not include wind direction changes for high wind speeds. Therefore, the inclusion of VWC as an additional metric in fire danger calculations in a modified FWI could increase operational understanding of high-danger locations and terrain impacts on extreme and unpredictable fire behaviour.