The study of fuel load models provides a theoretical foundation for the scientific assessment and prevention of forest fire risks in Sichuan Province,while also contributing to sustainable forest management and fire prevention strategies.Using data from 6 848 standard arbor sample plots collected during the Sichuan Province Forest Fire Risk Survey,arbor forests were categorized into five forest types based on vegetation zones and types.The fuel load characteristics of different vegetation types were analyzed,and a Pearson correlation analysis was conducted to identify the key factors influencing fuel load.Utilizing various curve-fitting techniques,optimal relationships between these pivotal factors and fuel load were established,which were subsequently combined to construct a regression model based on composite variables.This composite variable model was compared to a multiple stepwise regression model to identify the most effective fuel load estimation model.1)Significant differences were observed in the fuel load characteristics among vegetation types.The arbor layer accounted for the largest proportion of the fuel load,while the litter layer contributed the least.2)Key factors influencing fuel load varied across vegetation types.Average tree height,average diameter at breast height,and age group emerged as major influencing factors across multiple vegetation types,whereas altitude,canopy density,and stand density had significant impacts in specific vegetation types.3)Both the composite variable regression model and the multiple stepwise regression model exhibited strong predictive performance.However,the optimal model depended on the forest type.The composite model performed best for cold coniferous forests,while the multiple stepwise regression model achieved higher predictive accuracy for cold broadleaf forests,subtropical coniferous forests,subtropical broadleaf forests,and subtropical mixed forests.The models developed in this study are methodologically robust and generate reliable predictions.These models are recommended for estimating the fuel loads of various vegetation types in Sichuan Province,providing critical support for future fire risk assessments and prevention strategies.