Estimating mangrove above-ground biomass at Maowei Sea, Beibu Gulf of China using machine learning algorithm with Sentinel-1 and Sentinel-2 data
- 22 July 2022
- journal article
- research article
- Published by Taylor & Francis in Geocarto International
- Vol. 37 (27) , 15778-15805
Abstract
Blue carbon ecosystems such as mangroves are natural barriers to resisting and alleviating the impact of storm surges and extreme catastrophic weather. Accurate and efficient determination of the aboveground biomass of mangroves is of great importance for the protection and restoration of blue carbon ecosystems and their response to climate change. This study proposes a light gradient boosting model (LGBM) based on particle swarm optimization (PSO) algorithm for feature selection. We constructed and verified the proposed model using 227 quadrat datasets from a field survey and Sentinel-1 and Sentinel-2 data. The determination coefficient (R2) and root-mean-square error (RMSE) were used to evaluate the performance of the model. Compared with random forest(RF), K-nearest neighbourhood regression(KNNR), extreme gradient boosting(XGBR), LGBM, and other machine learning algorithms, the LGBM-PSO model achieves better results (R2 = 0.7807, RMSE = 24.6864 Mg·ha−1), The predicted range of mangrove biomass is 4.623–206.975 Mg·ha−1. Therefore, the use of multisource remote sensing data combined with the LGBM-PSO model can provide better prediction results of aboveground biomass of mangroves, thereby providing a new method for estimating the aboveground biomass of large-scale mangroves.Keywords
Funding Information
- National Natural Science Foundation of China
- Natural Science Foundation of Guangxi Zhuang Autonomous Region
- Guangxi Key Research and Development Program
- Guangxi Forestry Science and Technology Promotion Demonstration Project
- Key Research Base of Humanities and Social Sciences in Guangxi Universities ‘Beibu Gulf Ocean Development Research Center’
- High level talent introduction project of Beibu Gulf University
- Guangxi Autonomous Region college students innovation and entrepreneurship training program
This publication has 81 references indexed in Scilit:
- Remote Sensing of Mangrove Ecosystems: A ReviewRemote Sensing, 2011
- Managing Forests for Climate Change MitigationScience, 2008
- Top 10 algorithms in data miningKnowledge and Information Systems, 2007
- The potential of L‐band SAR for quantifying mangrove characteristics and change: case studies from the tropicsAquatic Conservation: Marine and Freshwater Ecosystems, 2007
- Relating Radar Remote Sensing of Biomass to Modelling of Forest Carbon BudgetsClimatic Change, 2004
- Overview of the radiometric and biophysical performance of the MODIS vegetation indicesRemote Sensing of Environment, 2002
- Present state and future of the world's mangrove forestsEnvironmental Conservation, 2002
- A Decision-Theoretic Generalization of On-Line Learning and an Application to BoostingJournal of Computer and System Sciences, 1997
- Mapping biomass of a northern forest using multifrequency SAR dataIEEE Transactions on Geoscience and Remote Sensing, 1994
- An Introduction to Kernel and Nearest-Neighbor Nonparametric RegressionThe American Statistician, 1992