Adaptation of Different Machine Learning Algorithms for Grassland Biomass Estimation

doi:10.11733/j.issn.1007-0435.2022.11.034

Acta Agrestia Sinica ›› 2022, Vol. 30 ›› Issue (11): 3156-3164.DOI: 10.11733/j.issn.1007-0435.2022.11.034

Adaptation of Different Machine Learning Algorithms for Grassland Biomass Estimation

BU Ling-xin¹, LAI Quan^1,2, LIU Xin-yi¹

1. College of Geographical Sciences, Inner Mongolia Normal University, Hohhot, Inner Mongolia 010022, China;
2. Key Laboratory of Remote Sensing and GIS in Inner Mongolia Autonomous Region, Hohhot, Inner Mongolia 010022, China

Received:2022-04-29 Revised:2022-07-07 Published:2022-12-02

不同机器学习算法在草原草地生物量估算上的适应性研究

卜灵心¹, 来全^1,2, 刘心怡¹

1. 内蒙古师范大学地理科学学院, 内蒙古呼和浩特 010022;
2. 内蒙古自治区遥感与地理信息系统重点实验室, 内蒙古呼和浩特 010022

通讯作者: 来全,E-mail:laiquan@imnu.edu.cn
作者简介:卜灵心(1998-)，男，内蒙古通辽人，硕士研究生，主要从事机器学习及生物量估算研究，E-mail：blx_1998@163.com
基金资助:
引进高层次人才科研启动金项目(2022JBYJ030)；内蒙古自然科学基金(2022MS04006)；国家自然科学基金国际(地区)合作与交流项目(4191101037)；内蒙古自然科学基金(2021MS04016)资助

Abstract

Abstract: Accurate estimating of grassland aboveground biomass (AGB) is important for scientific adjustment of the grass-livestock relationship,ecological environment protection and sustainable development of grassland resources. Based on remote sensing,meteorological and digital elevation model data,we paper used three machine learning algorithms,including the Support vector machines (SVM),BP Neural Networks (BP) and Random forest (RF) to establish a grassland AGB estimation model in Xilin Gol League and evaluate the estimation potential of the three models. The accuracy validation results showed that the RF algorithm had the highest regression accuracy in the study area (R=0.88,RMSE=0.10,MSE=0.01,MAE=0.07). The model established by the SVM algorithm had higher regression accuracy in the meadow steppe and desert steppe,while the regression ability of the RF algorithm had a relative advantage in the typical steppe. The analyses results of the contribution of different characteristic variables to the estimated AGB showed that the fractional vegetation cover (FVC),Normalized Difference Vegetation Index (NDVI),Enhanced vegetation index (EVI) and Precipitation (PRCP) greatly influence the AGB. The results of this study provide scientific suggestions to improve the accuracy of above-ground biomass estimation and selecting methods in arid/semi-arid grassland areas.

Key words: Grassland above-ground biomass, Machine learning, Accuracy assessment, Xilingole Grassland

摘要： 准确估算草地地上生物量(Aboveground biomass，AGB)对于科学调整草畜关系、保护生态环境和实现草地资源的可持续发展具有重要意义。本文以锡林郭勒盟不同草地类型为研究对象，基于遥感数据、气象数据和数字高程模型数据，利用支持向量机(Support vector machines,SVM)、BP神经网络(BP neural networks,BP)和随机森林(Random forest,RF)三种机器学习算法建立AGB估算模型，评估三种机器学习算法模型估算AGB的潜力。精度验证结果表明，在研究区内不区分草地类型整体建立估算模型时RF算法的回归精度最高(R=0.88，RMSE=0.10，MSE=0.01，MAE=0.07)。SVM算法建立的模型在草甸草原和荒漠草原回归精度较高，而RF算法回归能力在典型草原具有相对优势。不同特征变量对估算AGB的贡献分析结果表明，植被覆盖度(Fractional vegetation cover,FVC)、归一化植被指数(Normalized difference vegetation Index,NDVI)、增强植被指数(Enhanced vegetation index,EVI)和降水量(Precipitation,PRCP)四个变量对AGB估算结果的影响较大。本文研究结果为干旱/半干旱区草地地上生物量估算精度的提高和方法的选择提供科学建议。

关键词: 草地地上生物量, 机器学习, 精度评估, 锡林郭勒草原

CLC Number:

S812

BU Ling-xin, LAI Quan, LIU Xin-yi. Adaptation of Different Machine Learning Algorithms for Grassland Biomass Estimation[J]. Acta Agrestia Sinica, 2022, 30(11): 3156-3164.

卜灵心, 来全, 刘心怡. 不同机器学习算法在草原草地生物量估算上的适应性研究[J]. 草地学报, 2022, 30(11): 3156-3164.

0
/ / Recommend

Add to citation manager EndNote|Ris|BibTeX

URL: https://manu40.magtech.com.cn/Jweb_cdxb/EN/10.11733/j.issn.1007-0435.2022.11.034

https://manu40.magtech.com.cn/Jweb_cdxb/EN/Y2022/V30/I11/3156

References

[1] ADAMS J M,FAURE H,FAURE-DENARD L,et al. Increases in Terrestrial Carbon Storage from the Last Glacial Maximum to the Present[J]. Nature,1990,348(6303):711-714
[2] 王盼. 草原补贴对牲畜数量的影响[D].重庆:西南大学,2019:8-9
[3] LANENROTH W K,HUNT H W,SWIFT D M,et al. Estimating Aboveground net Primary Production in Grasslands:A simulation Approach[J]. Ecological Modelling,1986,33(2-4):297-314
[4] 史展. 基于RS的藏北草原生物量时空变化及驱动力研究[D].北京:中国科学院大学,2014:12-15
[5] 来全. 基于日光诱导叶绿素荧光的蒙古高原GPP模拟及其对干旱的响应研究[D].长春:东北师范大学,2021:4-5
[6] 杨淑霞. 三江源地区高寒草地生物量和草畜平衡的时空变化动态及其影响因素研究[D].兰州:兰州大学,2017:8-9
[7] 李想,于红博,刘月璇,等. 锡林郭勒不同草原类型群落生物量及多样性研究[J]. 草地学报,2022,30(1):196-204
[8] LU D. The Potential and Challenge of Remote Sensing-based Biomass Estimation[J]. International Journal of Remote Sensing,2006,27(7):1297-1328
[9] 张雨欣,黄健熙,金云翔,等. 草地地上生物量估算模型研究进展[J]. 草地学报,2022,30(4):850-858
[10] ZHU F,WANG H,LI M,et al. Characterizing the Effects of Climate Change on Short-term Post-disturbance Forest Recovery in Southern China from Landsat Time-series Observations (1988-2016)[J]. Frontiers of Earth Science,2020,14(4):816-827
[11] COLIN J,JUNGHO I. Forest Biomass Estimation from Airborne LiDAR Data Using Machine Learning Approaches[J]. Remote Sensing of Environment,2012(125):80-91
[12] 陈鹏飞,王卷乐,廖秀英,等. 基于环境减灾卫星遥感数据的呼伦贝尔草地地上生物量反演研究[J]. 自然资源学报,2010,25(7):1122-1131
[13] PRASAD R,PANDEY A,SINGH K et al. Retrieval of Spinach Crop Parameters by Microwave Remote Sensing with Back Propagation Artificial Neural Networks:A Comparison of Different Transfer Functions[J]. Advances in Space Research,2012,50(3):363-370
[14] JIN X,DIAO W,XIAO C,et al. Estimation of Wheat Agronomic Parameters Using new Spectral Indices[J]. PLoS One,2013,8(8):e72736
[15] LI C,WUIF H,SCHMID B,et al. Estimating Plant Traits of Alpine Grasslands on the Qinghai-Tibetan Plateau Using Remote Sensing[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2018,11(7):2263-2275
[16] QU Y,ZHAO Y,DING G,et al. Spatiotemporal Patterns of the Forage-livestock Balance in the Xilin Gol steppe,China:Implications for Sustainably Utilizing Grassland-ecosystem Services[J]. Journal of Arid Land,2021,13(2):135-151
[17] LIU Q,TONG Y. The Effects of Land Use on the Eco-environmental Evolution of Farming-Pastoral Region in North China:with an Emphasis on Duolun County in Inner Mongolia[J]. Acta Ecol Sin,2003(23):1025-1030
[18] 张泽飞. 锡林郭勒盟生态灾害风险评价[D].呼和浩特:内蒙古师范大学,2010:12-14
[19] JIANG R,WANG P,XU Y,et al. Assessing the Operation Parameters of a Low-altitude UAV for the Collection of NDVI Values over a Paddy Rice Field[J]. Remote Sensing,2020,12(11):1850
[20] GUPTA R K,PRASAD S,NADHAM T,et al. Relative Sensitivity of District Mean RVI and NDVI over an Agrometeorological Zone[J]. Advances in Space Research,1993,13(5):261-264
[21] GUNATHILAKA M. Modelling the Behavior of DVI and IPVI Vegetation Indices Using Multi-Temporal Remotely Sensed Data[J]. International Journal of Environment,Engineering and Education,2021,3(1):9-16
[22] HUETE A,JUSTICE C,LIU A. Development of Vegetation and Soil Indices for MODIS-EOS[J]. Remote Sensing of Environment,1994,49(3):224-234
[23] MOUNTRAKIS G,IM J,OGOLE C. Support Vector Machines in Remote Sensing:A review[J]. ISPRS Journal of Photogrammetry and Remote Sensing,2011,66(3):247-259
[24] CORINNA C,VLADIMIR V. Support-Vector Networks[J]. Machine Learning,1995,20(3):273-297
[25] RODRIGUEZ V,SANCHEZ M,CHICA M,et al. Machine Learning Predictive Models for Mineral Prospectivity:An Evaluation of Neural Networks,Random Forest,Regression Trees and Support Vector Machines[J]. Ore Geology Reviews,2015(71):804-818
[26] 曹书博,李嘉豪,周世玉,等. 基于模糊最小绝对非线性回归的木材热物性参数预测模型[J]. 林业科学,2021,57(11):142-151
[27] CHEN L,WANG Y. Research on the Material Fire Risk Index Based on the GA-BP Neural network[J]. Journal of Physics:Conference Series,2021,1865(4):042125
[28] WANG L,ZHOU X,ZHU X,et al. Estimation of Biomass in Wheat Using Random Forest Regression Algorithm and Remote Sensing Data[J]. The Crop Journal,2016,4(3):212-219
[29] BREIMAN L. Random Forests[J]. MACH LEARN,2001,45(1):5-32
[30] 王云飞,庞勇,舒清态,等. 基于随机森林算法的橡胶林地上生物量遥感反演研究——以景洪市为例[J]. 西南林业大学学报,2013,33(6):38-45
[31] CHAN C W,PAELINCKX D. Evaluation of Random Forest and Adaboost Tree-based Ensemble Classification and Spectral Band Selection for Ecotope Mapping Using Airborne Hyperspectral Imagery[J]. Remote Sensing of Environment,2008,112(6):2999-3011
[32] IFTIKHAR A,FIONA C,EDWARD D,et al. Satellite Remote Sensing of Grasslands:from Observation to Management-a Review[J]. Journal of Plant Ecology,2016,9(6):649-671
[33] DEO R C,SAHIN M. Forecasting Long-term Global Solar Radiation with an ANN Algorithm Coupled with Satellite-derived (MODIS) Land Surface Temperature (LST) for Regional Locations in Queensland[J]. Renewable and Sustainable Energy Reviews,2017(72):828-848
[34] WANG Z,YU Q,GUO L. Quantifying the Impact of the Grain-for-Green Program on Ecosystem Health in the Typical Agro-Pastoral Ecotone:A Case Study in the Xilin Gol League,Inner Mongolia[J]. International Journal of Environmental Research and Public Health,2020,17(16):5631
[35] CHI D,WANG H,LI X,et al. Assessing the Effects of Grazing on Variations of Vegetation NPP in the Xilingol Grassland,China,Using a Grazing Pressure Index[J]. Ecological Indicators Integrating Monitoring Assessment & Management,2018(88):372-383

Adaptation of Different Machine Learning Algorithms for Grassland Biomass Estimation

不同机器学习算法在草原草地生物量估算上的适应性研究

PDF

Knowledge

Abstract

Cite this article

share this article

References

Related Articles 1

Recommended Articles

Metrics

Comments