GPU-accelerated visualization of historical wildfire events with temporal geospatial data integration

Wildfires severely affect ecosystems and populations. Reconstructing documented fires helps emergency services study hazards and train responders. Many current tools are difficult to integrate and lack real-time performance across scenarios. This paper presents a GPU-accelerated Unity system for replaying and rendering documented wildfire events with temporal precision. The proposed method is a visualization pipeline for historical fire records-not a fire propagation model-using recorded rather than computed ignition and extinction times. It integrates geospatial processing, particle rendering, and temporal event modeling for fire behavior analysis. The system processes historical data chronologically, transforms geographic coordinates to maintain spatial fidelity, and renders fire stages with terrain-aligned particles. It incorporates wind data and applies performance optimizations including object pooling and GPU-accelerated particle rendering. Results illustrate reconstruction of documented fire patterns with stable, real-time frame rates in high-density scenarios, highlighting applicability for scientific analysis and training. • A geodetic transformation pipeline converting historical fire coordinates to a spatially consistent metric frame using ellipsoidal Earth geometry. • A deterministic temporal scheduler enabling timestamp-faithful replay of fire sequences with GPU-accelerated particle rendering and meteorological integration. • Performance validation demonstrating real-time interactive frame rates with up to 407 simultaneous fires on dedicated and integrated GPUs.