<p>Airblast is a common safety hazard in block caving. It occurs when the roof of a goaf collapses suddenly over a large area, causing rapid compression and release of air, resulting in high-speed airflow that can cause injury and equipment damage. To effectively assess and prevent airblast hazards, it is necessary to study the catastrophic behavior of airblast parameters. The air gap, caving scale, and muckpile height are airblast key parameters. This research employs the LBM-DEM (Coupled Lattice Boltzmann Method and Discrete Element Method) to model the airblast formation process. Combined simulation experiments were conducted to examine the effects of air gap, muckpile height, and caving scale, with sensitivity analysis performed to determine airblast response. We propose a multivariate regression model that, using three arguments, expresses the maximum airblast velocity at the drawpoints. Results show that maximum air velocity during roof collapse is positively correlated with air gap and caving scale, and negatively correlated with muckpile height. At a caving scale of 10³ m³, the air gap is the most critical parameter affecting maximum air velocity. Moreover, the interactions among these parameters exhibit distinct coupling characteristics. The findings provide a theoretical basis and reference for assessing and preventing airblast hazards.</p>