Three-Dimensional Response-Surface Modeling and Its Application In Quantifying Canopy Interception and Ground Deposition During Aerial Firefighting

During the aerial application of fire-extinguishing agents, operational effectiveness is fundamentally determined by the balance between canopy interception and ground deposition, yet the non-linear coupled mechanisms driven by vegetation structure and droplet transport parameters remain insufficiently quantified. To address this issue, this study constructs a response-surface modeling framework to investigate the interception-deposition partitioning during canopy penetration. First, the Lindeman-Merenda-Gold (LMG) analysis method is employed to isolate and rank the contributions of droplet parameters under multicollinearity. Subsequently, a robust three-dimensional response-surface model incorporating ridge regression and a Huber-loss-based weighted least squares estimator is constructed. This modeling approach establishes continuous mappings from operational parameters, namely flow rate and droplet size, to the penetration responses of interception and deposition rates. The results from the experimental observations, integrating with the response-surface analysis, show that increasing LAI enhances canopy interception while reducing ground deposition. Specifically, mean interception rises from 28.26% to 35.10% for Schima superba, and from 30.07% to 38.93% for Pinus massoniana, with proportional decreases in ground deposition. Overall, canopy densification strengthens droplet capture and weakens penetration. Additionally, conifer canopies demonstrate higher interception and lower deposition compared to broadleaf canopies under comparable LAI conditions.