“…In this study, the distance correlation coefficient was used to measure the correlation between the variables and the stock volume, and independent variables more suitable for estimating the stock volume were screened out. The distance correlation coefficient not only reflects the linear relationship between variables but also expresses the nonlinear relationship between variables; additionally, it does not require any model assumptions or parameter conditions [27]. The 39 factors selected as the characteristics were as follows: land type, area, landform, elevation (m), slope (degree), slope aspect, slope position, soil name, soil texture, soil-layer thickness (cm), humus-layer thickness, site-quality grade, land-management type, forest-land protection grade, traffic area, forest land-quality grade, undergrowth vegetation type, undergrowth vegetation height (m), undergrowth vegetation total coverage (%), forest category, forest belt length (m), forest belt width (m), number of forest belt rows, age, average diameter at breast height (cm), average height (m), canopy density or coverage, density, number of trees (plants) per mu, number of trees (plants), afforestation seedling age, Band2, Band3, Band4, Band5, Band6, Band7, Band10, and Band11.…”