A Comparison of Models of Stand Volume in Spruce-Fir Mixed Forest in Northeast China

Based on a multiple linear regression model, random forest algorithm and generalized additive stand volume model was constructed to provide theoretical basis for sustainable management. A total of 224 fixed plots in the Jingouling farm, Wangqing County, Jilin Province, were used as data sources. Specifically, 157 training data, 77 test data. The effects structure variables, topography cutting diversity variables climate analyzed. explained 95.51% volume, 95.45% volume. Stand had more influence spruce-fir than other variables. Among evenness index, Shannon index Simpson relatively greater impact times intensity first direct relationship with small study area.