Predicting time-to-harvest in mixed-species forests using a random survival forest algorithm

Survival analysis is composed of a group of analytical approaches that can be used to predict the occurrence of harvest activities, which provides insightful information about the dynamics of natural resources utilization in a region. Recently, random survival forest (RSF) has been proposed in survival analysis to capture the complex relationships among variables. The main objective of this study was to employ the RSF algorithm to examine the temporal evolution of tree harvest, accounting for stand and environmental variables. Specifically, the predictability of the RSF model was compared with the Cox proportional hazard (Cox) model, a popular model in survival analysis. Important variables in explaining the variation of harvest time were identified. Data collected by the USDA Forest Service, Forest Inventory and Analysis (FIA) program from permanent plots in the southern Appalachian region were utilized in the analysis. Results showed that the RSF model consistently outperformed the Cox model based on prediction accuracy. Among 14 variables examined, ownership, forest type, elevation, state, and slope emerged as most important. Utilizing only these five variables in a reduced model produced satisfactory prediction accuracy compared to the full model (i.e., the models with all variables included). The findings of this work provide insights for forest managers and policy makers to utilize survival analysis methods in understanding harvest activities at the regional scale.