Estimation of Aboveground Carbon Density of Forests Using Deep Learning and Multisource Remote Sensing

Abstract

Forests are crucial in carbon sequestration and oxygen release. An accurate assessment of forest carbon storage is meaningful for Chinese cities to achieve carbon peak and carbon neutrality. For an accurate estimation of regional-scale forest aboveground carbon density, this study applied a Sentinel-2 multispectral instrument (MSI), Advanced Land Observing Satellite 2 (ALOS-2) L-band, and Sentinel-1 C-band synthetic aperture radar (SAR) to estimate and map the forest carbon density. Considering the forest field-inventory data of eastern China from 2018 as an experimental sample, we explored the potential of the deep-learning algorithms convolutional neural network (CNN) and Keras. The results showed that vegetation indices from Sentinel-2, backscatter and texture characters from ALOS-2, and coherence from Sentinel-1 were principal contributors to the forest carbon-density estimation. Furthermore, the CNN model was found to perform better than traditional models. Results of forest carbon-density estimation validated the improvements effectively by combining the optical and radar data. Compared with traditional regression methods, deep learning has a higher potential for accurately estimating forest carbon density using multisource remote-sensing data.

Effects of Cadmium Stress on Carbon Sequestration and Oxygen Release Characteristics in A Landscaping Hyperaccumulator—Lonicera japonica Thunb.

“…The present study is in accordance with the report by [ 63 , 64 ], which proposed that landscape plants with functional carbon sequestration and oxygen release can exchange more CO 2 and O 2 with the external environment, then transform more solar energy into organic matter stored in plants. It is known that all the principal biotic components of carbon sequestration and oxygen release stem from the photosynthetic process of plants [ 21 , 65 ]. With the increase in Cd treatment concentration, P , W CO 2 , W O 2 , P CO 2 and P O 2 in L. japonica began to decrease; however, when Cd treatment concentration was up to 125 mg kg −1 , P , W CO 2 , W O 2 , P CO 2 and P O 2 in L. japonica had no significant decrease compared with the control, which indicated a good photosynthetic ability of the plant even under high concentrations of Cd treatment.…”

A Landscape Approach to Understanding Carbon Sequestration Assets at a State-Wide Scale for Sustainable Urban Planning

“…In the context of current advancements, our research distinguishes itself by offering a holistic evaluation of carbon stocks, encompassing both below-and above-ground reserves. This approach is in contrast to studies that typically emphasize aboveground biomass [35]. Our work advances the field by employing precise species-specific identification, thereby enhancing the granularity of carbon dynamic analysis.…”

Machine learning-based grassland aboveground biomass estimation and its response to climate variation in Southwest China

“…Tang et al (2021) also showed the RF model was superior to SVM, PLSR and back-propagation artificial neural network (BP-ANN) in AGB monitoring of the headwater of the Yellow River grasslands. Some recent comparison studies on AGB estimation performance of DL and ML models suggested CNN algorithm was found to perform better than classical ML models (i.e., RF, SVM), due to CNN was more sensitive to changes in features (Du et al, 2021;Zhang et al, 2022). While another case study showed RF would obtain slightly better accuracy than CNN if the data was elaborately designed (Dong et al, 2020).…”