Exploring parameter selection for carbon monitoring based on Landsat-8 imagery of the aboveground forest biomass on Mount Tai

A novel feature selection method using support vector machine recursive feature elimination (SVM-RFE) significantly enhances the accuracy of estimating aboveground forest biomass on Mount Taishan, outperforming traditional stepwise regression techniques. This approach leverages Landsat-8 imagery to identify optimal parameter subsets for biomass estimation, addressing critical challenges in global carbon monitoring under the United Nations' REDD+ mechanism. The study demonstrates that integrating single-band information, vegetation indices, texture features, and terrain factors can improve AGB estimation precision, offering a robust tool for environmental protection and climate stabilization efforts.

Qiu, A., Yang, Y., Wang, D., Xu, S., & Wang, X. (2019). Exploring parameter selection for carbon monitoring based on Landsat-8 imagery of the aboveground forest biomass on Mount Tai. European Journal of Remote Sensing, 53(sup1), 4–15. https://doi.org/10.1080/22797254.2019.1686717