结合冠层密度的森林净初级生产力遥感估测

doi:10.12302/j.issn.1000-2006.202008007

国家林草科技领军期刊
中国精品科技期刊
中国高校百佳科技期刊
江苏省新闻出版政府奖期刊奖
RCCSE林学权威期刊（A+）
CSCD核心期刊
Scopus数据库收录期刊
中文核心期刊
SCD核心期刊

作者加群：102861116

微信公众号：南京林业大学学报

高级检索

南京林业大学学报（自然科学版） ›› 2021, Vol. 45 ›› Issue (5): 153-160.doi: 10.12302/j.issn.1000-2006.202008007

结合冠层密度的森林净初级生产力遥感估测

李陶(), 李明阳^*(), 钱春花

南京林业大学林学院,江苏南京 210037

收稿日期:2020-08-04 接受日期:2020-10-12 出版日期:2021-09-30 发布日期:2021-09-30
通讯作者: 李明阳
基金资助:
国家自然科学基金项目(31770679)

Combining crown density to estimate forest net primary productivity by using remote sensing data

LI Tao(), LI Mingyang^*(), QIAN Chunhua

College of Forestry, Nanjing Forestry University, Nanjing 210037, China

Received:2020-08-04 Accepted:2020-10-12 Online:2021-09-30 Published:2021-09-30
Contact: LI Mingyang

摘要/Abstract

摘要：

【目的】 森林冠层密度与林分年龄、植被生长状况有关,在区域森林净初级生产力遥感估测中,结合森林冠层密度以期提高估测精度。【方法】 以广东省韶关市为研究对象,选用2017年Landsat-8 OLI影像、2017年357块森林资源连续清查固定样地数据为主要信息源,分别采用随机森林、多元线性回归、人工神经网络和K最近邻分类法等4种模型,结合森林冠层密度制图器(FCD)进行区域森林净初级生产力特征变量的选取、参数建模、模型精度评价和森林净初级生产力空间制图。【结果】 特征变量中,红光波段(B4)、归一化植被指数(NDVI)、比值植被指数(RVI)、叶面积指数(LAI)、缨帽变换土壤植被因子、纹理特征和地形特征在森林净初级生产力反演中有重要作用。将森林冠层密度因子加入反演模型后,4种遥感估测模型精度均有大幅度提高。对4种遥感估测模型进行性能比较,随机森林模型精度最高,其次是多元线性回归模型、人工神经网络模型,K-最近邻分类模型精度最低。研究区内森林净初级生产力平均值为10.689 t/(hm²·a),高森林净初级生产力 [≥18 t/(hm²·a)]林分面积仅占研究区的19.61%,主要分布在海拔较高的西北部。【结论】 结合冠层密度进行森林净初级生产力的建模,可有效提高模型估测精度。

关键词: 森林净初级生产力, 冠层密度, 遥感反演, 广东韶关市

Abstract:

【Objective】 Forest canopy density is related to stand age and vegetation growth status. In the remote sensing estimation of regional forest net primary productivity, combining forest canopy density may substantially improve the estimation accuracy. 【Method】 In this study, Shaoguan City in Guangdong Province was used as the case study area; the Landsat-8 OLI images and the fixed sample plot data of 357 forest resources continuous inventory in 2017 were collected as the main information sources. Using four models of random forest the multiple linear regression, artificial neural network and K-nearest neighbor classification methods, and forest canopy density, the selection of modeling variables, the evaluation of model accuracy, and the creation of a spatial map of net primary productivity of regional forest vegetation were performed. 【Result】 The results showed that among the characteristic variables, red band (B4), normalized vegetation index (NDVI), ratio vegetation index (RVI), leaf area index (LAI), soil vegetation factor of tassel cap transformation, texture characteristics, and terrain characteristics played important roles in the prediction of vegetation net primary productivity. After adding the forest canopy density factor into the inversion model, the accuracy of the four remote sensing estimation models was substantially improved. The performance comparison of the four remote sensing-based models indicated that the random forest model had the highest accuracy, followed by the multiple linear regression model and artificial neural network model; the K-nearest neighbor classification model had the lowest accuracy. The average net primary productivity of forests in the study area was 10.689 t/(hm²·a), and the area of high forest net primary productivity (≥18 t/(hm²·a)) accounted for 19.61% of the total area in the study area, which was mainly distributed in the northwest region at higher altitudes. 【Conclusion】 In this study, at the regional scale, the model prediction accuracy of forest net primary productivity can be effectively improved by combining canopy density.

Key words: forest net primary productivity, canopy density, remote sensing retrieval, Shaoguan City, Guangdong Province

中图分类号:

S757

李陶,李明阳,钱春花. 结合冠层密度的森林净初级生产力遥感估测[J]. 南京林业大学学报（自然科学版）, 2021, 45(5): 153-160.

LI Tao, LI Mingyang, QIAN Chunhua. Combining crown density to estimate forest net primary productivity by using remote sensing data[J].Journal of Nanjing Forestry University (Natural Science Edition), 2021, 45(5): 153-160.DOI: 10.12302/j.issn.1000-2006.202008007.

0
/ / 推荐

导出引用管理器 EndNote|Reference Manager|ProCite|BibTeX|RefWorks

链接本文: http://nldxb.njfu.edu.cn/CN/10.12302/j.issn.1000-2006.202008007

http://nldxb.njfu.edu.cn/CN/Y2021/V45/I5/153

图/表 7

表1

NPP遥感估测模型备选自变量"

变量类型 variable types	变量名(特征因子) variable name(characteristic facter)	变量个数 number of variables
单波段 single band	B₂, B₃, B₄, B₅, B₆, B₇	6
植被指数 vegetation index	I_NDVI,I_RVI,I_DVI,I_EVI,I_GVI,I_PVI	6
叶面积指数 leaf area index	$I L AI$	1
缨帽变换 Kauth-Thomas transformation	I_Bright,I_Green,I_Wet	3
地形因子 terraian factor	I_Slope	1
纹理特征 textural feature	B_2con, B_2dis, B_2mea, B_2hom, B_2asm, B_2ent, $B 2 ske$ , B_2cor B_3con, B_3dis, B_3mea, B_3hom, B_3asm, B_3ent, $B 3 ske$ , B_3cor B_4con, B_4dis, B_4mea, B_4hom, B_4asm, B_4ent, $B 4 ske$ , B_4cor B_5con, B_5dis, B_5mea, B_5hom, B_5asm, B_5ent, $B 5 ske$ , B_5cor B_6con, B_6dis, B_6mea, B_6hom, B_6asm, B_6ent, $B 6 ske$ , B_6cor B_7con, B_7dis, B_7mea, B_7hom, B_7asm, B_7ent, $B 7 ske$ , B_7cor	48

表1

表2

图1

图2

图3

表3

图4

参考文献 24

[1]	LIETH H, WJITTAKER R H. Primary productivity of the biosphere[M]. New York: Springer-Verland Press, 1975.
[2]	肖和善. 植被净初级生产力模型研究[J]. 亚热带水土保持, 2007, 19(4):24-28.
	XIAO H S. Research on the model of vegetation net primary productivity[J]. Subtrop Soil Water Conserv, 2007, 19(4):24-28.DOI: 10.3969/j.issn.1002-2651.2007.04.007. doi: 10.3969/j.issn.1002-2651.2007.04.007
[3]	吴文浩, 李明阳, 卜子汇. 基于开放式数据库的江苏省植被净生产力遥感估测方法研究[J]. 西北林学院学报, 2010, 25(5):146-151.
	WU W H, LI M Y, BU Z H. Estimation of net primary productivity of vegetation in Jiangsu Province based on open datasets[J]. Journal of Northwest Forestry University, 2010, 25(5):146-151.
[4]	徐新良, 曹明奎. 森林生物量遥感估算与应用分析[J]. 地球信息科学, 2006, 8(4):122-128.
	XU X L, CAO M K. An analysis of the applications of remote sensing method to the forest biomass estimation[J]. Geo-Infor-mation Science, 2006, 8(4):122-128.DOI: 10.3969/j.issn.1560-8999.2006.04.024. doi: 10.3969/j.issn.1560-8999.2006.04.024
[5]	朱教君, 康宏樟, 胡理乐. 应用全天空照片估计林分透光孔隙度(郁闭度)[J]. 生态学杂志, 2005, 24(10):1234-1240.
	ZHU J J, KANG H Z, HU L L. Estimation on optical porosity or canopy closure for a forest stand with hemispherical images[J]. Chinese Journal of Ecology, 2005, 24(10):1234-1240.
[6]	肖兴威. 中国森林生物量与生产力的研究[D]. 哈尔滨:东北林业大学, 2005.
	XIAO X W. Study on forest biomass and productivity in China[D]. Harbin:Northeast Forestry University, 2005.
[7]	LI Y C, LI C, LI M Y, et al. Influence of variable selection and forest type on forest aboveground biomass estimation using machine learning algorithms[J]. Forests, 2019, 10(12):1073.DOI: 10.3390/f10121073. doi: 10.3390/f10121073
[8]	余超, 王斌, 刘华, 等. 中国森林植被净生产量及平均生产力动态变化分析[J]. 林业科学研究, 2014, 27(4):542-550.
	YU C, WANG B, LIU H, et al. Dynamic change of net production and mean net primary productivity of China’s forests[J]. Forest Research, 2014, 27(4):542-550.DOI: 10.13275/j.cnki.lykxyj.2014.04.016. doi: 10.13275/j.cnki.lykxyj.2014.04.016
[9]	ZHAO M, ZHOU G S. Estimation of biomass and net primary productivity of major planted forests in China based on forest inventory data[J]. Forest Ecology & Management, 2005, 207(3):295-313. DOI: 10.1016/j.foreco.2004.10.049. doi: 10.1016/j.foreco.2004.10.049
[10]	JOSHI C, LEEUW J D, SKIDMORE A K, et al. Remotely sensed estimation of forest canopy density: a comparison of the performance of four methods[J]. International Journal of Applied Earth Observations & Geoinformation, 2006, 8(2):84-95.DOI: 10.1016/j.jag.2005.08.004. doi: 10.1016/j.jag.2005.08.004
[11]	RIKIMARU A. Landsat TM data processing guide for forest canopy density mapping and monitoring model[C]// ITTO workshopon utilization of sensing in site assessment and planning for rehabilitation of logged-over forest, Bankok, 1996: 1-8.
[12]	ROY P S, RIKIMARU A, MIYATAKE S. Tropical forest cover density mapping[J]. Tropical Ecology, 2002, 43(1):39-47.
[13]	姚登举, 杨静, 詹晓娟. 基于随机森林的特征选择算法[J]. 吉林大学学报(工学版), 2014, 44(1):137-141.
	YAO D J, YANG J, ZHAN X J. Feature selection algorithm based on random forest[J]. Journal of Jilin University (Engineering and Technology Edition), 2014, 44(1):137-141.DOI: 10.13229/j.cnki.jdxbgxb201401024. doi: 10.13229/j.cnki.jdxbgxb201401024
[14]	曹正凤. 随机森林算法优化研究[D]. 北京:首都经济贸易大学, 2014.
	CAO Z F. Study on optimization of random forests algorithm[D]. Beijing:Capital University of Economics and Business, 2014.
[15]	李欣海. 随机森林模型在分类与回归分析中的应用[J]. 应用昆虫学报, 2013, 50(4):1190-1197.
	LI X H. Using random forest for classification and regression[J]. Chinese Journal of Applied Entomology, 2013, 50(4):1190-1197.DOI: 10.7679/j.issn.2095-1353.2013.163. doi: 10.7679/j.issn.2095-1353.2013.163
[16]	张宝光. 人工神经网络在遥感数字图像分类处理中的应用[J]. 国土资源遥感, 1998, 10(1):3-5.
	ZHANG B G. The application of artificial neural network to classification processing of remote sensing digital images[J]. Remote Sensing Forest & Resources, 1998, 10(1):3-5.DOI: 10.6046/gtzyyg.1998.01.04. doi: 10.6046/gtzyyg.1998.01.04
[17]	滕敏, 卫文学, 滕宁. K-最近邻分类算法应用研究[J]. 软件导刊, 2015, 14(3):44-46.
	TENG M, WEI W X, TENG N. Application research of K-nearest neighbor classification algorithm[J]. Software Guide, 2015, 14(3):44-46.DOI: 10.11907/rjdk.143927. doi: 10.11907/rjdk.143927
[18]	陈尔学, 李增元, 武红敢, 等. 基于K-NN和Landsat数据的小面积统计单元森林蓄积估测方法[J]. 林业科学研究, 2008, 21(6):745-750.
	CHEN E X, LI Z Y, WU H G, et al. Forest volume estimation method for small areas based on K-NN and Landsat data[J]. Forest Research, 2008, 21(6):745-750.DOI: 10.3321/j.issn:1001-1498.2008.06.002. doi: 10.3321/j.issn:1001-1498.2008.06.002
[19]	ZHENG B, AGRESTI A. Summarizing the predictive power of a generalized linear model[J]. Statistics in Medicine, 2000, 19(13):1771-1781.DOI: 10.1002/1097-0258(20000715)19:13<1771:aid-sim485>3.0.co;2-p. doi: 10.1002/1097-0258(20000715)19:13<1771:aid-sim485>3.0.co
[20]	胡小飞, 唐宪, 胡月明, 等. 广州市城市森林净初级生产力遥感估算[J]. 中南林业科技大学学报, 2016, 36(5):19-25.
	HU X F, TANG X, HU M Y, et al. Estimation of net primary productivity of urban forest ecological system in Guangzhou City[J]. Journal of Central South University of Forestry & Technology, 2016, 36(5):19-25.DOI: 10.14067/j.cnki.1673-923x.2016.05.004. doi: 10.14067/j.cnki.1673-923x.2016.05.004
[21]	姜春. 广东省土地利用/覆盖变化对植被净初级生产力的影响研究[D]. 广州:中国科学院大学, 2015.
	JIANG C. Analyzing the effects of land cover change on vegetation net primary productivity in Guangdong Province[D]. Guangzhou: University of Chinese Academy of Sciences, 2015.
[22]	周爱萍, 向悟生, 姚月锋, 等. 广西植被净初级生产力(NPP)时空演变及主要影响因素分析[J]. 广西植物, 2014, 34(5):622-628,588.
	ZHOU A P, XIANG W S, YAO Y F, et al. Analyzing variation characteristics of vegetation net primary productivity (NPP) in Guangxi[J]. Guihaia, 2014, 34(5):622-628,588.DOI: 10.3969/j.issn.1000-3142.2014.05.008. doi: 10.3969/j.issn.1000-3142.2014.05.008
[23]	江洪. 基于森林碳汇信息的福建省低碳发展研究[D]. 福州:福州大学, 2011.
	JIANG H. Low carbon development study in Fujian Province,China based on forest carbon sinks information[D]. Fuzhou:Fuzhou University, 2011.
[24]	吴国训. 江西省森林植被净初级生产力及碳储量估算[D]. 南京:南京林业大学, 2015.
	WU G X. Net primary productivity and carbon storage estimation of forest in Jiangxi Province,China[D]. Nanjing:Nanjing Forestry University, 2015.

特征因子 characteristic factor	相关性 correlation	特征因子 characteristic factor	相关性 correlation
B₄	-0.267^**	B_5dis	0.121^*
I_NDVI	0.249^**	I_Bright	-0.185^**
I_RVI	0.279^**	I_Green	0.157^**
I_LAI	0.281^**	I_Slope	0.217^**
B_4mea	-0.288^**	I_den	-0.397^**

模型 model	不含冠层密度 without canopy density			含冠层密度 with canopy density
模型 model	R²	σ (RMSE)	σ (MAE)	R²	σ (RMSE)	σ (MAE)
随机森林 random forest	0.622	6.883	5.657	0.799	5.472	3.960
人工神经网络 artificial neural network	0.327	9.720	8.075	0.606	8.172	6.124
多元线性回归 multiple linear regression	0.129	12.154	9.656	0.498	9.319	7.117
K-最近邻分类法 K-nearest neighbor	0.142	13.057	9.427	0.424	10.700	6.985

结合冠层密度的森林净初级生产力遥感估测

Combining crown density to estimate forest net primary productivity by using remote sensing data

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 7

参考文献 24

相关文章 1

编辑推荐

Metrics

本文评价

作者

读者

审稿