Where lightning strikes: New study reveals how terrain affects NSW's lightning
We know that lightning is a pervasive natural hazard in NSW, with wide-ranging risk implications, including the potential to damage critical infrastructure, disrupt power grids, and ignite wildfires. But have you ever wondered whether lightning is more likely to strike in some parts of NSW than others?
While meteorologists rely on dynamic, rapidly changing weather data to forecast storms, a study led by Sima Rahmani, Prof Jason Evans, and Prof Jason Sharples at NSW Bushfire and Natural Hazards Centre and UNSW asks a different question: Can location, terrain and landscape tell us where lightning is most likely to strike?
Published in Environmental Research: Climate, this new study analysed nearly 10 million lightning strikes across NSW between 2016 and 2021. Using advanced machine learning, the research team demonstrated that static geographical and terrain characteristics alone can meaningfully capture spatial patterns of lightning occurrence at 250m resolution.
“This research gives us another piece of the puzzle for understanding how land–atmosphere interactions are reflected in the spatial patterns of lightning across NSW," Sima said.
"By looking at static geographical characteristics and terrain, we can start to build a clearer picture of the patterns associated with where lightning occurs. This could provide a useful baseline for improving future modelling, forecasting and risk assessment across sectors exposed to lightning, particularly in a changing climate.”
Highlights in the findings
The "sweet spot" for lightning: lightning does not hit the NSW landscape evenly. The highest concentration of strikes occurs in an elevation band between 350m and 650m, particularly along the western slopes and transitional foothills of the Great Dividing Range.
Coastal proximity and latitude are key: When modelling lightning occurrence, distance to the coast and latitude emerged as significant predictors. These features likely act as proxies for large-scale atmospheric circulation patterns and moisture availability; two large-scale environmental controls associated with lightning occurrence.
Elevation variability matters: local relief (the variation between high and low points within a 10km area) can be as important as elevation itself in capturing the relationship between terrain and lightning occurrence.
Dry lightning is more strongly modulated by terrain: The study took a special look at dry lightning strikes that occur with little rain, making them a potentially hazardous subset of strikes for wildfire ignitions. For dry lightning, topographic features play an even stronger role. Here, latitudinal position overrides coastal distance and reveals a more localised spatial pattern.
Why this research matters
While weather forecasting and nowcasting can tell us when a storm is likely to occur or what might happen in the short-term, this study highlights the importance of looking at the mostly unchanging terrain, location and landscape features when considering lightning risk. By using machine learning to identify these patterns, this approach can help map lightning risk over longer periods and support planning and risk management.
"Knowing where in the landscape lightning is more likely to occur allows agencies to locate suppression assets strategically and deal with lightning started fires quickly," said Prof Jason Evans.
This research can also contribute to a baseline measure for:
bushfire management
infrastructure and renewable energy
regional risk planning
environmental modelling
Read the full open-access paper: Rahmani, S., Evans, J. P., & Sharples, J. (2026). Machine learning-based spatial modelling of lightning occurrence informed by terrain characteristics. Environmental Research: Climate, 5, 035025, https://doi.org/10.1088/2752-5295/ae81d9
Comments