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

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