Gross Primary Production (GPP) for China from 2001–2020 Estimated by Machine Learning Methods
Description
Based on flux tower observations, this study comprehensively evaluated five mainstream GPP products, including MODIS (Moderate Resolution Imaging Spectroradiometer), PML-V2 (Penman-Monteith-Leuning Version 2), GOSIF (Global Orbiting Carbon Observatory-2 based Solar Induced chlorophyll Fluorescence), CEDAR (sCaling Ecosystem Dynamics with ARtifical intelligence), and TL-LUE, and quantified their uncertainties using the Bayesian Three-Cornered Hat method. On this basis, by integrating multi-source data, a high-fidelity GPP dataset was generated using five machine learning methods: Categorical Boosting, Support Vector Machine Regression, Light Gradient Boosting, Extreme Gradient Boosting, and Random Forest. Validated against data from 15 flux tower sites in mainland China, CatBoost exhibited the best performance, with the lowest RMSE and MAE and the highest R². This dataset was calculated using CatBoost.The spatial resolution is 0.05 degrees.
Files
Steps to reproduce
Using five common remote sensing products (including MODIS (Moderate Resolution Imaging Spectroradiometer), PML-V2 (Penman-Monteith-Leuning Version 2), GOSIF (Global Orbiting Carbon Observatory-2 based Solar Induced chlorophyll Fluorescence), CEDAR (sCaling Ecosystem Dynamics with ARtifical intelligence), and TL-LUE), combined with flux tower data from 15 Chinese sites (Changbaishan (CBS), Qianyanzhou (QYZ), Yanshan (YS), Maoershan (MES), Huzhong (HZ), Dangxiong (DX), Haibei (HB), Inner Mongolia (IMG), Ruoergai (REG), Naqu (NQ), Damao (DM), Yucheng (YC), Jinzhou (JZ), Changlin (CL), Panjin (PJ)), five machine learning methods were trained and used for spatial extrapolation. Finally, the results from the five machine learning methods (including Categorical Boosting, Support Vector Machine Regression, Light Gradient Boosting, Extreme Gradient Boosting, and Random Forest) were compared to select the best-performing method and data product.
Institutions
- Xinjiang UniversityXinjiang, Ürümqi