Wildfire risk as a dynamic constraint shaping energetic landscapes under climate change

Energetic landscapes emerge from the spatial coupling of energy resources, land use, and environmental risk. In fire-prone regions, wildfires increasingly act as a territorial constraint that reshapes where land-based energy production can be realized. However, most assessments of solar energy potential rely on static representations of land availability and historical climate conditions, limiting their relevance under climate change.

This study develops an interpretable machine learning framework to predict daily human-caused wildfire occurrence and extends it by integrating climate change scenarios to explore how future wildfire risk interacts with solar energy potential across degraded and previously disturbed lands. A stacking ensemble model is trained using daily meteorological variables, environmental characteristics, and indicators of human accessibility. SHAP-based interpretation is applied to identify key drivers of wildfire occurrence under present-day conditions. Climate scenario data are subsequently introduced to project future wildfire susceptibility, which is spatially overlaid with estimates of solar energy potential to characterize shifts in energetic landscapes.

The results show that short-term meteorological extremes dominate present-day wildfire occurrence, while accessibility-related factors reflect the spatial imprint of human activity. Under future climate scenarios, wildfire susceptibility intensifies and expands spatially, intersecting with areas currently identified as having high solar potential. As a result, both the magnitude and spatial configuration of realizable energy potential are dynamically reshaped when wildfire risk is treated as an integral component of the energy landscape rather than an external disturbance.

By framing wildfire risk as a constitutive element of energetic landscapes, this study provides action-relevant spatial insights into how climate-driven hazards may redefine land-based climate mitigation potential under increasing climate uncertainty.