Remote Sensing Estimation of Biomass of Pinus kesiya var.langbianensis by Geographically Weighted Regression Models

doi:10.13466/j.cnki.lyzygl.2017.01.015

Abstract

Abstract:

The Biomass model of Simao pine(Pinus kesiya var.langbianensis)was built based on the data collected from 120 Simao pine sampling trees,Landsat TM images in 2005 and the data of forest resource inventory in 2006 in Jinggu County,Yunnan Province.Then the remote sensing biomass estimation Model of Simao Pine were built by the ordinary least square(OLS)and geographically weighed regression(GWR).The results showed that:GWR model had a better fitting effect than OLS,in which coefficient of determination(R ²)was significantly bigger than the OLS model,Akaike information index(AIC)reduced by 7.832;It was obviously depicted from the sample test of independence that model prediction accuracy was improved from 72.70%(OLS)to 75.06%(GWR).The unit-area biomass was 49.02t / hm ² by inversion,and basically consistent with the measured data;it was lower than the measured data 1.229%,and less than the estimation value of OLS.The total biomass of Simao pine in Jinggu County was 2.101×10 ⁷ t based on GWR model.The study indicated that forest aboveground biomass estimation based on geographically weighed regression(GWR)model could improve effectively the fitting accuracy of forest biomass estimation model,and could be used to estimate the biomass of Simao pine forest by remote sensing.

Key words: Pinus kesiya var.langbianensis, biomass, remote sensing, ordinary least square(OLS), Geographically Weighted Regression(GWR)

CLC Number:

LÜ Yanyu, LI Chao, OU Guanglong, XIONG Hexian, WEI Anchao, ZHANG Bo, XU Hui. Remote Sensing Estimation of Biomass of Pinus kesiya var.langbianensis by Geographically Weighted Regression Models[J]. FOREST RESOURCES WANAGEMENT, 2017, (1): 82-90.

Figures/Tables 13

Fig.1

Tab.1

Tab.2

Fig.2

Tab.3

Tab.4

Fig.3

Tab.5

Tab.6

Fig.4

Tab.7

Tab.8

Fig.5

References 19

[1]	Brown S, Sathaye J, Cannell M, et al. Mitigation of carbon emissions to the atmosphere by forest management[C]// Proceedings of the IEICE General Conference.Japan:The Institute of Electronics,Information and Communication Engineers, 1996: 9735-9746.
[2]	岳彩荣 . 香格里拉县森林生物量遥感估测研究[D]. 北京林业大学, 2012.
[3]	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.
[4]	曾晶, 张晓丽 . 高分一号遥感影像下崂山林场林分生物量反演估算研究[J]. 中南林业科技大学学报, 2016(1):46-51.
[5]	李明诗, 谭莹, 潘洁 , 等. 结合光谱、纹理及地形特征的森林生物量建模研究[J]. 遥感信息, 2006(6):6-9.
[6]	孙雪莲, 舒清态, 欧光龙 , 等. 基于随机森林回归模型的思茅松人工林生物量遥感估测[J]. 林业资源管理, 2015(1):71-76.
[7]	玄海燕, 罗双华, 王大斌 . GWR模型中权函数的选取与窗宽参数的确定[J]. 甘肃联合大学学报:自然科学版, 2008,22(3):10-12.
[8]	玄海燕, 黎锁平, 刘树群 . 地理加权回归模型及其拟合[J]. 甘肃科学学报, 2007,19(1):51-52.
[9]	Zhang L, Shi H . Local modeling of tree growth by geographically weighted regression[J]. Forest Science, 2004,50(2):225-244.
[10]	Zhang L, Ma Z, Guo L . Spatially assessing model errors of four regression techniques for three types of forest stands[J]. Forestry:Oxford, 2008,81(2):209-225.
[11]	Fotheringham A S, Brunsdon C, Charlton M . Geographically Weighted Regression:the analysis of spatially varying relationships[J]. American Journal of Agricultural Economics, 2004,86(7):554-556. doi: 10.1111/ajae.v86.2
[12]	Brunsdon C, Fotheringham A S, Charlton M . Geographically Weighted Regression:a method for exploring spatial nonstationarity[J]. Stata Technical Bulletin, 1996,28(4):281-298.
[13]	刘畅 . 黑龙江省森林碳储量空间分布研究[D]. 哈尔滨:东北林业大学, 2014.
[14]	戚玉娇 . 大兴安岭森林地上碳储量遥感估算与分析[D]. 哈尔滨:东北林业大学, 2014.
[15]	郭含茹, 张茂震, 徐丽华 , 等. 基于地理加权回归的区域森林碳储量估计[J]. 浙江农林大学学报, 2015,32(4):497-508. doi: 10.11833/j.issn.2095-0756.2015.04.002
[16]	胥辉, 张会儒 . 林木生物量模型研究[M].云南科技出版社, 2002.
[17]	胥辉, 岳彩荣 , 基于遥感技术的香格里拉县森林景观变化与森林生物量估测研究[M].云南科技出版社, 2014.
[18]	玄海燕, 王静 . 地理加权回归模型的平稳性检验[J]. 甘肃联合大学学报:自然科学版, 2006,20(5):10-12.
[19]	张博, 欧光龙, 孙雪莲 , 等. 空间效应及其回归模型在林业中的应用[J]. 西南林业大学学报, 2016,36(3):144-152.

序号	模型	样本容量/个	模型参数			R²
序号	模型	样本容量/个	a	b	c	R²
1	W=a×D^b	120	0.059	2.541	0.467	0.869
2	W=a×D^b×H^c	120	0.058	2.120		0.900
3	W=a×(D²×H)^b	120	0.097	0.817		0.882

遥感因子	生物量	遥感因子	生物量	遥感因子	生物量	遥感因子	生物量
TM1	-0.228^*	PCA_2	-0.215^*	Tm2	-0.271^**	ME1_9	-0.208^*
TM7	-0.270^*	PCA_1	0.248^*	VIS123	-0.270^*	ME2_9	-0.271^**
TM5_45	-0.255^*	Third	0.258^*	SM4_5	0.210^*	ME3_9	-0.309^**
TM5	-0.271^**	Brightness	-0.262^*	ME1_5	-0.242^*	ME5_9	-0.325^**
ND57	0.253^*	ME1_7	-0.218^*	ME2_5	-0.276^**	ME7_9	-0.337^**
ND452	0.228^*	ME2_7	-0.281^**	ME3_5	-0.309^**	ME2_15	-0.232^*
ND32	-0.284^**	ME3_7	-0.323^**	ME5_5	-0.296^**	ME3_15	-0.272^**
TM3	-0.282^**	ME5_7	-0.332^**	EN5_5	-0.215^*	ME5_15	-0.278^**
Albedo	-0.263^*	ME7_7	-0.349^**	ME7_5	-0.306^**	ME7_15	-0.285^**

模型	样本容量	R²	RMSE	AIC
OLS	71	0.152	0.528	74.781
GWR	71	0.380	0.344	66.949

对比项	Sum Sq.	df		Mean Sq.	F
OLS模型残差	10.952	69.000		—	—
GWR模型改善	2.945	10.228		0.288	—
GWR模型残差	8.008	58.772	0.136	2.113

变量	参数β	标准误	(β-1)×SD	(β+1)×SD
截距	1.845	0.047	1.798	1.892
ME7_7	-0.164	0.047	-0.211	-0.117