Research on Inversion of Forest Volume Based on Domestic High-Resolution Data

doi:10.13466/j.cnki.lyzygl.2021.03.016

Abstract

Abstract:

Taking the Wangyedian Forest Farm in Inner Mongolia as the research area,combined with ground surveys,and based on the preprocessing of the GF-2 remote sensing data,48 remote sensing factors such as spectral information,vegetation index and texture information were extracted,and 8 remote sensing factors were selected for modeling by Pearson correlation coefficient method.Using multiple linear regression,multi-layer perceptron,K-nearest neighbor,support vector machine,and random forest model to estimate the forest volume,the forest volume inversion map in the study area was obtained.The results showed that:1) Among the remote sensing factors extracted from GF-2,mean of texture features based on the second-order matrix had a higher correlation with the forest volume;2) Random Forest had better estimation accuracy of forest volume than methods such as multiple linear regression,multi-layer perceptron,K-nearest neighbor and support vector machine,and its relative root mean square error (rRMSE) was 25.40%;3) The areas with high forest volume in the study area were mainly distributed in the west and southeast;the areas with low forest volume were mainly distributed in the northwest,central and northern parts,which were consistent with the actual investigation.The domestic GF-2 image and random forest algorithm had certain potential in the inversion of forest volume.

Key words: remote sensing, GF-2, random forest, forest stock volume, spatial distribution

CLC Number:

XIAO Yue, XU Xiaodong, LONG Jiangping, LIN Hui. Research on Inversion of Forest Volume Based on Domestic High-Resolution Data[J]. FOREST RESOURCES WANAGEMENT, 2021, 0(3): 101-107.

Figures/Tables 6

Fig.1

Tab.1

Formulas of the vegetation index

植被指数	公式
比值植被指数	$RVI = ρ NIR / ρ RED$
差值植被指数	$DVI = ρ NIR - ρ RED$
归一化植被指数	$NDVI = (ρ NIR - ρ RED) / (ρ NIR + ρ RED)$
土壤修正植被指数	$SAVI = 1 + L (ρ NIR - ρ RED) / ρ NIR + ρ RED + L) (L =$ 0,0.25,0.5,1)
大气抗阻植被指数	$ARVI = ρ NIR - ρ RB ρ NIR + ρ RB, 其中 ρ RB = ρ RED - γ (ρ BLUE - ρ RED)$
增强型植被指数	$EVI = [2.5 (ρ NIR - ρ RED)] / ρ NIR + 6 ρ RED - 7.5 ρ BLUE + 1)$
三角植被指数	$TVI = (ρ NIR - ρ RED) / (ρ NIR + ρ RED) + 0.5$
修正简单比值植被指数	$MSR = ρ NIR / ρ RED - 1 / (ρ NIR / ρ RED + 1)$
非线性指数	$NLI = (ρ NIR 2 - ρ RED) / (ρ NIR 2 + ρ RED)$

Tab.1

Tab.2

Tab.3

Fig.2

Fig.3

References 24

[1]	国庆喜, 张锋. 基于遥感信息估测森林的生物量[J]. 东北林业大学学报, 2003, 31(2):13-16.
[2]	王月婷. 基于多源遥感数据的森林蓄积量估算[D]. 北京:北京林业大学, 2015.
[3]	陈新云, 李利伟, 刘承芳, 等. 热带原始森林类型分类和蓄积量遥感反演研究[J]. 林业资源管理, 2019(2):39-46.
[4]	Ataee M S, Maghsoudi Y, Latifi H, et al. Improving Estimation Accuracy of Growing Stock by Multi-Frequency SAR and Multi-Spectral Data over Iran's Heterogeneously-Structured Broadleaf Hyrcanian Forests[J]. Forests, 2019, 10(8):641. doi: 10.3390/f10080641
[5]	李世波, 林辉, 王光明, 等. 基于GF-1的森林蓄积量遥感估测[J]. 中南林业科技大学学报, 2019, 39(8):70-75.
[6]	王佳, 宋珊芸, 刘霞, 等. 结合影像光谱与地形因子的森林蓄积量估测模型[J]. 农业机械学报, 2014, 45(5):216-220.
[7]	尤静妮. 基于高分遥感纹理信息的森林蓄积量估测研究[D]. 西安:西安科技大学, 2017.
[8]	王宗梅, 岳彩荣, 刘琦, 等. 基于光学和微波遥感数据的森林蓄积量估测模型研究[J]. 西南农业学报, 2018, 31(8):1722-1726.
[9]	曹林, 佘光辉, 代劲松, 等. 激光雷达技术估测森林生物量的研究现状及展望[J]. 南京林业大学学报:自然科学版, 2013, 37(3):163-169.
[10]	Zhang H, Zhu J, Wang C, et al. Forest Growing Stock Volume Estimation in Subtropical Mountain Areas Using PALSAR-2 L-Band PolSAR Data[J]. Forests, 2019, 10(3):276. doi: 10.3390/f10030276
[11]	黄丽艳, 闫巧玲, 高添, 等. 基于ALOSPALSAR雷达影像的人工林蓄积量估算——以塞罕坝机械林场华北落叶松人工林为例[J]. 生态学杂志, 2015, 34(9):2401-2409.
[12]	Gao T, Zhu J J, Yan Q L, et al. Mapping growing stock volume and biomass carbon storage of larch plantations in Northeast China with L-band ALOS PALSAR backscatter mosaics[J]. International journal of remote sensing, 2018, 39(22):7978-7997. doi: 10.1080/01431161.2018.1479793
[13]	Belgiu M, Drăgutt L. Random forest in remote sensing: A review of applications and future directions[J]. ISPRS Journal of Photogrammetry & Remote Sensing, 2016, 114:24-31.
[14]	郎晓雪, 许彦红, 舒清态, 等. 香格里拉市云冷杉林蓄积量遥感估测非参数模型研究[J]. 西南林业大学学报, 2019, 39(1):146-151.
[15]	刘帅. 基于BP神经网络的森林蓄积量估测[D]. 杭州:浙江农林大学, 2014.
[16]	汪康宁, 马婷, 吕杰. 基于随机森林算法的凉水自然保护区蓄积量反演研究[J]. 西南林业大学学报, 2016, 36(5):125-129.
[17]	王海宾, 彭道黎, 高秀会, 等. 基于GF-1 PMS影像和K-NN方法的延庆区森林蓄积量估测[J]. 浙江农林大学学报, 2018, 35(6):1070-1078.
[18]	张苏, 周小成, 黄洪宇, 等. 基于SVR的GF1号遥感影像森林蓄积量估测[J]. 贵州大学学报:自然科学版, 2019, 36(3):21-26.
[19]	刘唐, 江涛, 李昂, 等. 基于神经网络和不同立地质量的森林蓄积量遥感估测[J]. 山东科技大学学报:自然科学版, 2019, 38(2):25-35.
[20]	张凝, 冯仲科, 冯跃文, 等. 旺业甸实验林场针叶林蓄积量估测模型的研究[J]. 中南林业科技大学学报, 2013, 33(11):83-87.
[21]	蒋馥根, 孙华, ZHAO Feng, 等. 基于方差优化k最近邻法的森林蓄积量估测[J]. 森林与环境学报, 2019, 39(5):497-504.
[22]	刘兆华, 林辉, 龙江平, 等. 基于高分二号的旺业甸林场蓄积量估测模型研究[J]. 中南林业科技大学学报, 2020, 40(3):79-84.
[23]	Breiman L. Using iterated bagging to Debias regressions[J]. Machine Learning, 2001, 45(3):261-277. doi: 10.1023/A:1017934522171
[24]	蒋馥根, 孙华, 林辉, 等. 旺业甸林场人工林生物量遥感反演研究[J]. 中南林业科技大学学报, 2019, 39(10):88-94.

模型	RMSE/(m³/hm²)	RRMSE/%	R²
MLR	68.79	34.99	0.06
MLP	70.88	36.06	0.05
KNN	60.46	30.76	0.28
SVR	59.23	30.13	0.31
RF	49.94	25.40	0.51