Constructing a biomass model of Larix olgensis primary branches based on TLS

doi:10.12302/j.issn.1000-2006.202204037

JOURNAL OF NANJING FORESTRY UNIVERSITY ›› 2023, Vol. 47 ›› Issue (2): 130-140.doi: 10.12302/j.issn.1000-2006.202204037

Previous Articles Next Articles

Constructing a biomass model of Larix olgensis primary branches based on TLS

TANG Yiren¹(), JIA Weiwei¹^,²^,^*(), WANG Fan¹, SUN Yuman¹, ZHANG Ying¹

1. School of Forestry, Northeast Forestry University, Harbin 150040, China
2. Key Laboratory of Sustainable Forest Ecosystem Management, Ministry of Education, Northeast Forestry University, Harbin 150040, China

Received:2022-04-15 Revised:2022-06-13 Online:2023-03-30 Published:2023-03-28

Abstract

Abstract:

【Objective】 To explore the feasibility of using terrestrial laser scanning (TLS) point cloud data to estimate branch biomass and find the optimum model to predict the branch biomass of Larix olgensis. 【Method】 The characteristic factors (branch length (L_BL), branch chord length (L_BCL), branch diameter (d_B), branch angle (A_B), branch arch height (H_BAH), basal area of branches (S_BAB), and relative depth within the crown (d_RDINC)) of 733 first-order branches extracted using point cloud data of 26 artificial larch plants and the corresponding measured data from Mengjiagang Forestry Farm were used as data sources to establish first-level branch biomass base models at the branch level. Additionally, the feasibility of using TLS data to establish branch biomass models by comparing the differences between base models were analyzed. Finally, the prediction effects of the base model, mixed-effects model, and random forest model were compared separately using TLS data. 【Result】 The independent variables selected in the base model were S_BAB and L_BCL. Using TLS data to build the branch biomass base model had a better prediction accuracy. Comparing the results of the three models showed that the random forest model showed the best results both on the training and test sets, in the following order: random forest model > mixed-effects model > base model. The R² of the random forest model improved by 1.32% and 4.89%, and the RMSE decreased by 11.23% and 13.60%, respectively, compared to the mixed model and the base model. 【Conclusion】 The results of this paper proved that the branch biomass can be estimated accurately based on TLS using the random forest algorithm, which not only laid a practical foundation for the application of the random forest algorithm on stand growth models, but also provided an important reference value for the application of TLS in the study of crown structure.

Key words: Larix olgensis, point cloud data, branch characterization factor, branch biomass, mixed-effects model, random forest

CLC Number:

S758

TANG Yiren, JIA Weiwei, WANG Fan, SUN Yuman, ZHANG Ying. Constructing a biomass model of Larix olgensis primary branches based on TLS[J]. JOURNAL OF NANJING FORESTRY UNIVERSITY, 2023, 47(2): 130-140.

Figures/Tables 11

Fig.1

Table 1

Fig.2

Table 2

Table 3

Table 4

Fig.3

Fig.4

Fig.5

Fig.6

Table 5

References 32

[1]	董利虎, 李凤日, 贾炜玮. 基于线性混合效应的红松人工林枝条生物量模型[J]. 应用生态学报, 2013, 24(12): 3391-3398.
	DONG L H, LI F R, JIA W W. Linear mixed modeling of branch biomass for Korean pine plantation[J]. Chin J Appl Ecol, 2013, 24(12): 3391-3398.DOI:10.13287/j.1001-9332.2013.0571.
[2]	DELUCIA E H M H C E V. Climate-driven changes in biomass allocation in pines[J]. Glob Change Bio, 2000, 6(5):587-593.
[3]	HELMISAARI H S, MAKKONEN K, KELLOMÄKI S, et al. Below-and above-ground biomass, production and nitrogen use in eastern Finland[J]. For Ecol Manag, 2002, 165(1/2/3):317-326.DOI:10.1016/s0378-1127(01)00648-x.
[4]	MÄKINEN H, MÄKELÄ A. Predicting basal area of Scots pine branches[J]. For Ecol Manag, 2003, 179(1/2/3): 351-362. DOI:10.1016/S0378-1127(02)00551-0.
[5]	GONZALEZ DE TANAGO J, LAU A, BARTHOLOMEUS H, et al. Estimation of above-ground biomass of large tropical trees with terrestrial LiDAR[J]. Methods Ecol Evol, 2018, 9(2): 223-234. DOI: 10.1111/2041-210x.12904.
[6]	STOVALL A E L, VORSTER A G, ANDERSON R S, et al. Non-destructive aboveground biomass estimation of coniferous trees using terrestrial LiDAR[J]. Remote Sens Environ, 2017, 200: 31-42.DOI:10.1016/j.rse.2017.08.013.
[7]	ATKINS J W, BOHRER G, FAHEY R T, et al. Quantifying vegetation and canopy structural complexity from terrestrial Li DAR data using the forest package[J]. Methods Ecol Evol, 2018, 9(10): 2057-2066. DOI:10.1111/2041-210X.13061.
[8]	卢军, 刘宪钊, 孟维亮, 等. 基于地面激光点云数据的单木三维重建方法[J]. 南京林业大学学报(自然科学版), 2021, 45(6): 193-199.
	LU J, LIU X Z, MENG W L, et al. Methodology of individual tree 3D reconstruction based on terrestrial laser scanning point cloud data[J]. J Nanjing For Univ (Nat Sci Ed), 2021, 45(6): 193-199.
[9]	张颖, 贾炜玮. 基于地基激光雷达的落叶松人工林枝条因子提取和建模[J]. 应用生态学报, 2021, 32(7): 2505-2513.
	ZHANG Y, JIA W W. Extraction of branch factors and model construction for Larix plantation using terrestrial laser scanning (TLS)[J]. Chin J Appl Ecol, 2021, 32(7): 2505-2513. DOI:10.13287/j.1001-9332.202107.034.
[10]	范伟伟, 刘浩然, 徐永胜, 等. 基于地基激光雷达和手持式移动激光雷达的单木结构参数提取精度对比[J]. 中南林业科技大学学报, 2020, 40(8): 63-74.
	FAN W W, LIU H R, XU Y S, et al. Comparison of extraction precision of individual tree structure parameters based on terrestrial laser scanning and hand-held mobile laser scanning[J]. J Central South Univ For & Technol, 2020, 40(8): 63-74. DOI:10.14067/j.cnki.1673-923x.2020.08.008.
[11]	MÄKELÄ A. A carbon balance model of growth and self-pruning in trees based on structural relationships[J]. For Sc, 1997, 43(1): 7-24. DOI:10.1093/FORESTSCIENCE/43.1.7.
[12]	MEDHURST J L, BATTAGLIA M, CHERRY M L, et al. Allometric relationships for Eucalyptus nitens (Deane and Maiden) Maiden plantations[J]. Trees, 1999, 14(2): 91-101. DOI: 10.1007/PL00009756.
[13]	LEHTONEN A, MÄKIPÄÄ R, HEIKKINEN J, et al. Biomass expansion factors (BEFs) for Scots pine, Norway spruce and birch according to stand age for boreal forests[J]. For Ecol Manag, 2004, 188(1/2/3): 211-224. DOI: 10.1016/j.foreco.2003.07.008.
[14]	REPOLA J. Biomass equations for Scots pine and Norway spruce in Finland[J]. Silva Fenn, 2009, 43(4): 4-184.DOI: 10.14214/sf.184.
[15]	贾炜玮, 李凤日, 董利虎, 等. 基于相容性生物量模型的樟子松林碳密度与碳储量研究[J]. 北京林业大学学报, 2012, 34(1): 6-13.
	JIA W W, LI F R, DONG L H, et al. Carbon density and storage for Pinus sylvestris var. mongolica plantation based on compatible biomass models[J]. J Beijing For Univ, 2012, 34(1): 6-13. DOI:10.13332/j.1000-1522.2012.01.018.
[16]	曾伟生, 唐守正. 利用度量误差模型方法建立相容性立木生物量方程系统[J]. 林业科学研究, 2010, 23(6): 797-803.
	ZENG W S, TANG S Z. Using measurement error modeling method to establish compatible single-tree biomass equations system[J]. For Res, 2010, 23(6): 797-803. DOI:10.13275/j.cnki.lykxyj.2010.06.004.
[17]	许昊, 孙玉军, 王新杰, 等. 利用线性混合效应模型模拟杉木人工林枝条生物量[J]. 应用生态学报, 2015, 26(10): 2969-2977.
	XU H, SUN Y J, WANG X J, et al. Simulation of the branch biomass for Chinese fir plantation using the linear mixed effects model[J]. Chin J Appl Ecol, 2015, 26(10): 2969-2977. DOI:10.13287/j.1001-9332.20150921.026.
[18]	董灵波, 刘兆刚, 李凤日, 等. 基于线性混合模型的红松人工林一级枝条大小预测模拟[J]. 应用生态学报, 2013, 24(9): 2447-2456.
	DONG L B, LIU Z G, LI F R, et al. Primary branch size of Pinus koraiensis plantation: a prediction based on linear mixed effect model[J]. Chin J Appl Ecol, 2013, 24(9): 2447-2456. DOI:10.13287/j.1001-9332.2013.0490.
[19]	姜立春, 张锐, 李凤日. 基于线性混合模型的落叶松枝条长度和角度模型[J]. 林业科学, 2012, 48(5): 53-60.
	JIANG L C, ZHANG R, LI F R. Modeling branch length and branch angle with linear mixed effects for Dahurian larch[J]. Sci Silvae Sin, 2012, 48(5): 53-60. DOI:10.11707/j.1001-7488.20120508.
[20]	姜立春, 李凤日, 张锐. 基于线性混合模型的落叶松枝条基径模型[J]. 林业科学研究, 2012, 25(4): 464-469.
	JIANG L C, LI F R, ZHANG R. Modeling branch diameter with linear mixed effects for Dahurian larch[J]. For Res, 2012, 25(4): 464-469. DOI:10.13275/j.cnki.lykxyj.2012.04.013.
[21]	HAUGLIN M, ASTRUP R, GOBAKKEN T, et al. Estimating single-tree branch biomass of Norway spruce with terrestrial laser scanning using voxel-based and crown dimension features[J]. Scand J For Res, 2013, 28(5): 456-469. DOI:10.1080/02827581.2013.777772.
[22]	GONZALEZ DE TANAGO J, LAU A, BARTHOLOMEUS H, et al. Estimation of above-ground biomass of large tropical trees with terrestrial LiDAR[J]. Methods Ecol Evol, 2018, 9(2): 223-234. DOI:10.1111/2041-210x.12904.
[23]	LAU A, BENTLEY L P, MARTIUS C, et al. Quantifying branch architecture of tropical trees using terrestrial LiDAR and 3D modelling[J]. Trees, 2018, 32(5): 1219-1231. DOI:10.1007/s00468-018-1704-1.
[24]	陈世林. 激光雷达单木参数提取与生物量估算研究[D]. 北京: 北京林业大学, 2020.
	CHEN S L. Research on extraction of single tree parameters and biomass estimation based on LiDAR[D]. Beijing: Beijing Forestry University, 2020.
[25]	胡靖扬, 贾宝军, 林宽, 等. 辽东山区长白落叶松枝叶生物量模型建立与评估[J]. 西南林业大学学报, 2016, 36(5): 52-57,117.
	HU J Y, JIA B J, LIN K, et al. Establishment and evaluation of biomass models of branches and leaves for Larix olgensis plantation in montane region of eastern Liaoning Province[J]. J Southwest For Univ, 2016, 36(5):52-57,117.
[26]	ASHRAF M I, ZHAO Z Y, BOURQUE C P A, et al. Integrating biophysical controls in forest growth and yield predictions with artificial intelligence technology[J]. Can J For Res, 2013, 43(12): 1162-1171. DOI:10.1139/cjfr-2013-0090.
[27]	BREIMAN L. Random forests[J]. Mach Learn, 2001, 45(1):5-32.DOI:10.1023/A:1010933404324.
[28]	CUTLER D R, EDWARDS T C Jr, BEARD K H, et al. Random forests for classification in ecology[J]. Ecology, 2007, 88(11): 2783-2792. DOI: 10.1890/07-0539.1.
[29]	高若楠, 苏喜友, 谢阳生, 等. 基于随机森林的杉木适生性预测研究[J]. 北京林业大学学报, 2017, 39(12): 36-43.
	GAO R N, SU X Y, XIE Y S, et al. Prediction of adaptability of Cunninghamia lanceolata based on random forest[J]. J Beijing For Univ, 2017, 39(12): 36-43. DOI:10.13332/j.1000-1522.20170260.
[30]	欧强新, 李海奎, 雷相东, 等. 基于清查数据的福建省马尾松生物量转换和扩展因子估算差异解析:3种集成学习决策树模型的比较[J]. 应用生态学报, 2018, 29(6): 2007-2016.
	OU Q X, LI H K, LEI X D, et al. Difference analysis in estimating biomass conversion and expansion factors of Masson pine in Fujian Province, China based on national forest inventory data: a comparison of three decision tree models of ensemble learning[J]. Chin J Appl Ecol, 2018, 29(6): 2007-2016. DOI:10.13287/j.1001-9332.201806.019.
[31]	欧强新, 雷相东, 沈琛琛, 等. 基于随机森林算法的落叶松-云冷杉混交林单木胸径生长预测[J]. 北京林业大学学报, 2019, 41(9): 9-19.
	OU Q X, LEI X D, SHEN C C, et al. Individual tree DBH growth prediction of larch-spruce-fir mixed forests based on random forest algorithm[J]. J Beijing For Univ, 2019, 41(9): 9-19. DOI:10.13332/j.1000-1522.20180266.
[32]	卢婧, 冯仲科. 运用随机森林模型对北京市林分蓄积生长量的预测[J]. 东北林业大学学报, 2020, 48(5): 7-11.
	LU J, FENG Z K. Forecast of stand volume growth in Beijing by using random forest[J]. J Northeast For Univ, 2020, 48(5): 7-11. DOI:10.13759/j.cnki.dlxb.2020.05.002.

Related Articles 15

[1]	PENG Wenyue, JIA Weiwei, WANG Fan, LI Xin, LI Dandan. Extraction and construction of a QSM-based model of first-order branches of Larix gmelinii plantations [J]. JOURNAL OF NANJING FORESTRY UNIVERSITY, 2025, 49(2): 185-193.
[2]	YAN Yu, LI Qiujie, LI Weizheng. A single tree segmentation method for street trees facing side-looking MLS point clouds [J]. JOURNAL OF NANJING FORESTRY UNIVERSITY, 2024, 48(6): 166-174.
[3]	WANG Yue, MIAO Zheng, HAO Yuanshuo, LIU Xin, DONG Lihu. A single tree leaf area prediction model in the Larix olgensis and Fraxinus mandshurica mixed forest [J]. JOURNAL OF NANJING FORESTRY UNIVERSITY, 2024, 48(5): 235-245.
[4]	WANG Jiaxing, YAN Pingyu, SUN Baifei, LIU Jinhong, FENG Kele, ZHANG Hanguo. Growth variation and superior families early selection of Larix olgensis free-pollinated families [J]. JOURNAL OF NANJING FORESTRY UNIVERSITY, 2024, 48(5): 81-89.
[5]	ZHOU Youfeng, XIE Binglou, LI Mingshi. Mapping regional forest aboveground biomass from random forest Co-Kriging approach: a case study from north Guangdong [J]. JOURNAL OF NANJING FORESTRY UNIVERSITY, 2024, 48(1): 169-178.
[6]	GAO Xieyu, DONG Lihu, HAO Yuanshuo. Effects of thinning on Larix olgensis plantation stem form based on TLS [J]. JOURNAL OF NANJING FORESTRY UNIVERSITY, 2023, 47(6): 85-94.
[7]	SUN Yu, LI Fengri, XIE Longfei, DONG Lihu. Construction of the stand-level biomass model of Larix olgensis plantations based on stand and topographic factors [J]. JOURNAL OF NANJING FORESTRY UNIVERSITY, 2023, 47(3): 129-136.
[8]	SONG Lei, JIN Xingji, PUKKALA Timo, LI Fengri. Research on multi-objective management schedules of Larix olgensis plantations [J]. JOURNAL OF NANJING FORESTRY UNIVERSITY, 2023, 47(2): 150-158.
[9]	SUN Mingchen, JIANG Lichun. Standing volume prediction of Pinus sylvestris var. mongolica based on machine learning algorithm [J]. JOURNAL OF NANJING FORESTRY UNIVERSITY, 2023, 47(1): 31-37.
[10]	WANG Fan, JIA Weiwei, TANG Yiren, LI Dandan. Extraction of crown radius and development of crown profile model of Pinus koraiensis based on TLS [J]. JOURNAL OF NANJING FORESTRY UNIVERSITY, 2023, 47(1): 13-22.
[11]	GUO Changyou, GUO Hongxian, WANG Baohua. Study on increment model of individual-tree diameter of Cunninghamia lanceolata in consideration of climatic factors [J]. JOURNAL OF NANJING FORESTRY UNIVERSITY, 2023, 47(1): 47-56.
[12]	WANG Tian, WANG Xuefeng, LIU Jiazheng. The prediction model of moisture content of young Aquilaria sinensis leaves based on RFE_RF algorithm [J]. JOURNAL OF NANJING FORESTRY UNIVERSITY, 2022, 46(4): 177-184.
[13]	NIE Luyi, DONG Lihu, LI Fengri, MIAO Zheng, XIE Longfei. Construction of taper equation for Larix olgensis based on two-level nonlinear mixed effects model [J]. JOURNAL OF NANJING FORESTRY UNIVERSITY, 2022, 46(3): 194-202.
[14]	WU Jiong, JIANG Fugen, PENG Shaofeng, MA Kaisen, CHEN Song, SUN Hua. Estimating the tree height and yield of Camellia oleifera by combining crown volume [J]. JOURNAL OF NANJING FORESTRY UNIVERSITY, 2022, 46(2): 53-62.
[15]	LU Jun, LIU Xianzhao, MENG Weiliang, LI Hongjun. Methodology of individual tree 3D reconstruction based on terrestrial laser scanning point cloud data [J]. JOURNAL OF NANJING FORESTRY UNIVERSITY, 2021, 45(6): 193-199.

Metrics

Comments

www.nldxb.njfu.edu.cn

Edited ＆ Published by Editorial Department of Nanjing Forestry University(Natural Sciences Edition)
Address： No.159 Longpan Road,Nanjing 210037 Jiangsu,P.R.China
E-mail ：xuebao@njfu.edu.cn , xuebao@njfu.com.cn
Telephone +86-25-85428247,85427076
Distributed by China International Book Trading Corporation P.O. Box 399, Beijing ,P.R. China

数据 data	变量 variable		训练集train
数据 data	变量 variable		平均值 mean		最大值 max	最小值 min	标准差 SD	平均值 mean		最大值 max	最小值 min	标准差 SD
TLS数据 TLS data		单木因子tree factor		n=20					n=6
	树高/m(H)		18.69		25.82	13.12	3.85	19.74		23.05	14.05	3.56
	胸径/cm(D_BH)		19.22		31.10	9.06	5.81	18.31		26.35	12.32	5.53
		枝条因子branch factor		n=601					n=132
	枝长/cm(L_BL)		217.40		525.00	15.80	91.28	198.71		530.59	17.87	93.54
	弦长/cm(L_BCL)		201.20		493.00	14.80	84.21	186.19		500.30	15.03	85.15
	弓高/cm(H_BAH)		26.00		80.00	0.00	12.94	19.97		78.43	1.00	14.80
	基径/mm(d_B)		26.60		69.49	11.10	13.15	28.10		61.56	12.00	13.02
	枝条基部断面积/cm²(S_BAB)		6.81		37.93	1.04	8.05	7.52		29.76	1.14	7.14
	着枝角度/(°)(A_B)		60.00		150.00	15.00	16.02	53.00		104.00	19.00	15.00
	相对着枝深度(d_RDINC)		0.61		1.00	0.10	0.20	0.62		1.00	0.10	0.20
实测数据 real data		单木因子tree factor		n=20					n=6
	树高/m(H)		18.75		26.00	13.17	3.92	19.60		23.22	13.98	3.26
	胸径/cm(D_BH)		19.40		31.00	9.00	5.49	18.58		26.12	12.08	5.13
		枝条因子branch factor		n=601					n=132
	枝长/cm(L_BL)		210.00		532.00	10.00	99.51	182.00		515.00	13.00	92.18
	弦长/cm(L_BCL)		200.00		518.00	8	82.00	180.00		496.00	10.00	83.00
	弓高/cm(H_BAH)		23.00		70.00	0.00	12.00	19.00		81.00	1.00	15.00
	基径/mm(d_B)		27.32		70.12	6.28	14.55	29.48		65.74	8.51	13.74
	枝条基部断面积/cm²(S_BAB)		7.03		38.09	0.85	8.88	7.73		31.29	0.92	7.85
	着枝角度/(°)(A_B)		56.00		135.00	10.00	18.00	59.00		115.00	16.00	17.00
	相对着枝深度(d_RDINC)		0.62		1.00	0.14	0.21	0.62		1.00	0.11	0.20
	枝条生物量/g(W_b)		416.93		4 303.32	3.26	539.97	460.34		3 075.59	4.71	572.88

数据 data	参数 parameter	预测值 estimate	标准误 SE	t	P	拟合优度 goodness-of-fit statistic
数据 data	参数 parameter	预测值 estimate	标准误 SE	t	P	R²	RMSE/g	MAE/g	RMAE/%	F_p/%
实测数据 real data	a₀	0.069 77	0.025 77	9.951	<0.000 1
	a₁	0.720 14	0.034 74	15.184	<0.000 1	0.849 6	171.79	97.66	39.56	95.71
	a₂	1.350 26	0.084 06	12.975	<0.000 1
TLS数据 TLS data	a₀	0.049 56	0.017 01	12.914	<0.000 1
	a₁	0.788 71	0.033 27	23.708	<0.000 1	0.880 8	156.93	90.07	35.66	96.84
	a₂	1.342 01	0.069 31	19.363	<0.000 1

模型 model	随机效应参数 random effect parameter	参数个数 number of parameter	R²	赤池信息准则 AIC	贝叶斯准则 BIC	对数似然值 Log Likelihood	似然比 LRT	P
12	无	3	0.879	8 455.562	8 509.378	-4 217.56
12.1	a₀	5	0.887	8 433.588	8 455.825	-4 211.79
12.2	a₁	5	0.904	8 364.290	8 386.527	-4 177.15	66.044	<0.000 1
12.3	a₂	5	0.904	8 367.544	8 389.780	-4 178.77
12.4	a₀、a₁	7	0.905	8 437.588	8 468.719	-4 211.79
12.5	a₀、a₂	7	0.904	8 371.544	8 402.675	-4 178.77
12.6	a₁、a₂	7	0.913	8 338.741	8 369.872	-4 162.37	29.549	<0.000 1
12.7	a₀、a₁、a₂	10	不收敛

类型 type	矩阵 matrix	参数个数 parameters number	赤池信息准则 AIC	贝叶斯准则 BIC	对数似然值 Log Likelihood	似然比 LRT	P
方差-协方差 variance-covariance	复合对称矩阵CS	6	不收敛
	对角矩阵UN(1)	6	8 364.832	8 391.516	-4 176.42
	广义正定矩阵	7	8 338.741	8 369.872	-4 162.37	28.090 7	<0.000 1
自相关 autocorrelation	复合对称矩阵CS	8	不收敛
	一阶自回归矩阵AR(1)	8	8 340.194	8 375.773	-4 162.10
	一阶自回归与移动平均矩阵ARMA(1,1)	9	8 342.133	8 382.159	-4 162.07	0.060 7	0.805 4

模型model	数据集data	R²	RMSE/g	MAE/g	RMAE/%	F_p/%
基础模型 base model	训练集train	0.879	162.14	93.51	38.13	96.51
	测试集test	0.880	156.93	90.07	35.66	96.84
混合模型 mixed model	训练集train	0.913	150.49	83.65	31.58	97.02
	测试集test	0.911	152.75	86.86	32.59	96.99
随机森林模型 random forest model	训练集train	0.930	130.23	67.23	24.30	97.83
	测试集test	0.923	135.59	76.33	28.52	97.65

Constructing a biomass model of Larix olgensis primary branches based on TLS

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 11

References 32

Related Articles 15

Recommended Articles

Metrics

Comments

Author

Reader

Reviewer