Improving the Accuracy of Estimating Forest Carbon Density Using the Tree Species Classification Method

Abstract

The accurate and effective estimation of forest carbon density is an essential basis for effectively responding to climate change and achieving the goal of carbon neutrality. Aiming at the problem of the significant differences in the forest carbon model parameters of different tree species, this study used the tree forest in Yueyang City, Hunan Province, China, as the study object and used the random forest classification algorithm through the Google Earth Engine platform to classify the dominant tree species within the forested range of the study area based on the image elements. The overall accuracy in the forest/non-forest classification (primary classification) was 93.79% with a Kappa of 0.9145. The overall accuracy in the dominant species classification (secondary classification) was 87.30% with a Kappa of 0.7747. Based on the classification, a multiple linear regression (MLR), support vector machine (SVM), and random forest (RF) were constructed for different dominant tree species by combining some Forest Resource Inventory data and remote sensing data. The results showed that the RF model had a significantly higher coefficient of determination (R2 = 0.4054–0.7602) than the MLR (R2 = 0.0900–0.4070) and SVM (R2 = 0.1650–0.4450) as well as a substantially lower RMSE and MAE; its spatial distribution of forest carbon density ranged from 3.06 to 62.80 t·hm−2. Compared with the spatial distribution of the forest carbon density (4.64 to 31.96 t·hm−2) without the classification of dominant species, the method eliminated the problems of severe overfitting and significant underestimation of peak values when estimating under unclassified conditions. The method provides a reference for the remote sensing inversion of forest carbon density on a large scale.

The Sensitivity Feature Analysis for Tree Species Based on Image Statistical Properties

“…The SVM is a widely utilized classification algorithm in the field of forestry [39]. The performance of the SVM model in representing the tree species was assessed by calculating various evaluation metrics.…”

A Novel Framework for Forest Above-Ground Biomass Inversion Using Multi-Source Remote Sensing and Deep Learning

“…The unit of the ground dataset was small class, which consisted of diverse types of information, including the geographical coordinates of the sample plots, dominant tree species, tree diameter at breast height (DBH), and tree height. According to the sampling design methodology described in the Committee for Earth Observing Satellites (CEOS) AGB validation protocol, a total of 300 forested plots were randomly set up within the timberland boundary, and one plot was taken from each small class [19,27,48]. Based on the available data, the experiment was conducted using two distinct tree species.…”

“…The former often includes a limited number of parameters, such as multivariate linear regression (MLR) and generalized linear models (GLM), which require the addition of restrictive hypothesis functions between features and AGB [25,26]. However, due to the inherent complexity in the relationship between AGB and RS data, parametric models often exhibit limited accuracy [27]. Contrary to the parametric models, non-parametric machine learning models process multidimension complex data by adopting more flexible mappings, such as the classical RF and support vector regression (SVR) models [28].…”