Increasing numbers of explanatory variables tend to result in information redundancy and “dimensional disaster” in the quantitative remote sensing of forest aboveground biomass (AGB). Feature selection of model factors is an effective method for improving the accuracy of AGB estimates. Machine learning algorithms are also widely used in AGB estimation, although little research has addressed the use of the categorical boosting algorithm (CatBoost) for AGB estimation. Both feature selection and regression for AGB estimation models are typically performed with the same machine learning algorithm, but there is no evidence to suggest that this is the best method. Therefore, the present study focuses on evaluating the performance of the CatBoost algorithm for AGB estimation and comparing the performance of different combinations of feature selection methods and machine learning algorithms. AGB estimation models of four forest types were developed based on Landsat OLI data using three feature selection methods (recursive feature elimination (RFE), variable selection using random forests (VSURF), and least absolute shrinkage and selection operator (LASSO)) and three machine learning algorithms (random forest regression (RFR), extreme gradient boosting (XGBoost), and categorical boosting (CatBoost)). Feature selection had a significant influence on AGB estimation. RFE preserved the most informative features for AGB estimation and was superior to VSURF and LASSO. In addition, CatBoost improved the accuracy of the AGB estimation models compared with RFR and XGBoost. AGB estimation models using RFE for feature selection and CatBoost as the regression algorithm achieved the highest accuracy, with root mean square errors (RMSEs) of 26.54 Mg/ha for coniferous forest, 24.67 Mg/ha for broad-leaved forest, 22.62 Mg/ha for mixed forests, and 25.77 Mg/ha for all forests. The combination of RFE and CatBoost had better performance than the VSURF–RFR combination in which random forests were used for both feature selection and regression, indicating that feature selection and regression performed by a single machine learning algorithm may not always ensure optimal AGB estimation. It is promising to extending the application of new machine learning algorithms and feature selection methods to improve the accuracy of AGB estimates.

The importance of the height-diameter (H-D) relationship in forest productivity is well known. The general nonlinear regression model, based on the mean regression technical, is not able to give a complete description of the H-D relationship. This study aims to evaluate the H-D relationship among relative competition levels and develop a quantile regression (QR) model to fully describe the H-D relationship. The dominance index was applied to determine the relative competition levels of trees for the Chinese fir. Based on the basic Weibull growth model, the mean regression for five relative competition levels and 11 QR models was constructed with 10-fold cross-validation. We have demonstrated that the H-D relationship for the Chinese fir strongly correlated with relative competition states, but the five curves from mean regression models did not show a notable difference between the trends of H-D relationship under different competition levels. Similar regression results were found in QR models of the specific quantiles; the average tree height of five competition levels varied between 5.78% and 17.65% (i.e., about 0.06 and 0.18 quantiles). In addition, some special curves of the H-D relationship such as the QR models of the 0.01 and 0.99 quantile showed the H-D relationship under certain conditions. These findings indicate that the QR models not only evaluated the rates of change of the H-D relationships in various competition levels, but also described their characteristics with more information, like the upper and lower boundary of the conditional distribution of responses. Although the flexible QR curves followed the distribution of the data and showed more information about the H-D relationships, the H-D curves may not intersect with each other, even when the trees reached their maximum height. Hence, the QR model requires further practice in assessing the growth trajectory of the tree’s diameter or tree height to gain better results.