您好，欢迎访问甘肃省农业科学院 机构知识库！
甘肃省农业科学院机构知识库

全部

  • 全部
  • 标题
  • 学者
  • 机构
  • 关键词

Can Machine Learning Algorithms Successfully Predict Grassland Aboveground Biomass?

文献类型： 外文期刊

作者： Wang, Yue 1 ; Qin, Rongzhu 1 ; Cheng, Huzi 2 ; Liang, Tiangang 3 ; Zhang, Kaiping 1 ; Chai, Ning 1 ; Gao, Jinlong 3 ; Feng, Qisheng 3 ; Hou, Mengjing 3 ; Liu, Jie 3 ; Liu, Chenli 3 ; Zhang, Wenjuan 4 ; Fang, Yanjie 5 ; Huang, Jie 6 ; Zhang, Feng 1 ;

作者机构： 1.Lanzhou Univ, Coll Ecol, Lanzhou 730000, Peoples R China

2.Indiana Univ, Dept Psychol & Brain Sci, Lab Cognit Control, Bloomington, IN 47401 USA

3.Lanzhou Univ, Coll Pastoral Agr Sci & Technol, State Key Lab Grassland Agroecosyst, Lanzhou 730020, Peoples R China

4.Inst Qinghai Prov Nat Resources Survey & Monitori, Xining 810000, Peoples R China

5.Gansu Acad Agr Sci, Inst Dryland Farming, Key Lab High Water Utilizat Dryland Gansu Prov, Lanzhou 730070, Peoples R China

6.Gansu Acad Agr Sci, Anim Husb Pasture & Green Agr Inst, Lanzhou 730000, Peoples R China

7.Nanjing Agr Univ, NAU MSU Asia Hub, Nanjing 210095, Peoples R China

关键词： MODIS; Google Earth Engine; biomass inversion; spatio-temporal scalability; model building

期刊名称：REMOTE SENSING （ 影响因子：5.349； 五年影响因子：5.786 ）

ISSN：

年卷期： 2022 年 14 卷 16 期

页码：

收录情况： SCI

摘要： The timely and accurate estimation of grassland aboveground biomass (AGB) is important. Machine learning (ML) has been widely used in the past few decades to deal with complex relationships. In this study, based on an 11-year period (2005-2015) of AGB data (1620 valid AGB measurements) on the Three-River Headwaters Region (TRHR), combined with remote sensing data, weather data, terrain data, and soil data, we compared the predictive performance of a linear statistical method, machine learning (ML) methods, and evaluated their temporal and spatial scalability. The results show that machine learning can predict grassland biomass well, and the existence of an independent validation set can help us better understand the prediction performance of the model. Our findings show the following: (1) The random forest (RF) based on variables obtained through stepwise regression analysis (SRA) was the best model (R-vad(2) = 0.60, RMSEvad = 1245.85 kg DW (dry matter weight)/ha, AIC = 5583.51, and BIC = 5631.10). It also had the best predictive capability of years with unknown areas (R-indep(2) = 0.50, RMSEindep = 1332.59 kg DW/ha). (2) Variable screening improved the accuracy of all of the models. (3) All models' predictive accuracy varied between 0.45 and 0.60, and the RMSE values were lower than 1457.26 kg DW/ha, indicating that the results were reliably accurate.

  • 相关文献
作者其他论文 更多>>
置顶
购物车
反馈

© 京ICP备09089781号-29 主办单位：甘肃省农业科学院

学术资源： 中国农业科学院 农业专业知识服务系统 农科联盟资源共建共享平台 中国农业科技文献与信息服务平台 中国农业期刊集成服务平台