Objective Meteorological conditions are among the dominant factors controlling wildfire spread, and the interaction between wind and fire behavior constitutes a key issue in forest fire research. The coupled wildfire–atmosphere model Meso-NH/ForeFire is capable of representing fire–atmosphere interactions through a two-way feedback mechanism; however, its applicability under the complex topographic conditions of China has not yet been fully evaluated.

Method The “April 13” forest fire in Anze County, Shanxi Province, was selected as the case study. Driven by ERA5 reanalysis data, the Meso-NH/ForeFire coupled model was applied to simulate the wildfire spread process. Simulated near-surface wind fields and fire spread results were evaluated using meteorological station observations and the observed burned area, and comparative analyses were conducted against simulations without fire–atmosphere coupling.

Result (1) Meso-NH effectively captures the trends in near-surface wind variations, and the simulated wind fields show good overall consistency with observations from meteorological stations. (2) The coupled model simulations indicate that the fire spread is predominantly northwestward under the influence of southeast winds, which is consistent with field observations. (3) Compared to uncoupled simulations, the coupled model provides a more accurate depiction of the dominant fire spread direction and fire head migration paths. Accuracy evaluation results show that the coupled model achieves a Sorensen's Coefficient (SC) of 0.59 and a Mean Hausdorff Distance (MHD) of 306.3 m. Although some localized overestimation is observed, the model generally captures the primary controlling mechanisms and spatial patterns of forest fire spread.

Conclusion The Meso-NH/ForeFire wildfire–atmosphere coupled model is capable of effectively representing the overall fire spread pattern and its stage-dependent characteristics under complex mountainous terrain. The model shows strong potential for regional-scale wildfire behavior analysis and scenario simulation in China.