基于Landsat-8 OLI数据的马尾松林蓄积量饱和点确定及估测

doi:10.13466/j.cnki.lyzygl.2020.06.022

摘要/Abstract

摘要：

以湖南省72块马尾松林样地(25 m×25 m)为研究对象,基于地统计学半方差函数的球状模型估测森林生物量饱和点的方法,利用Landsat-8 OLI数据估测湖南省马尾松林蓄积量的最大饱和点,依据该饱和点估测方法,提出一种新的蓄积量估测模型,将其与多元逐步回归模型进行对比分析。结果表明:基于波段信息,球状模型和二项式模型的最大饱和点估测结果分别为217.05,206.86 m³/hm²;基于植被指数信息,2种方案的最大饱和点估测结果分别为196.95,183.06 m³/hm²;基于新提出的二项式模型的蓄积量建模估测验证结果和多元逐步回归模型的建模估测验证结果的R²分别为0.49和0.29,MAE分别为53.76,63.35 m³/hm²,RMSE分别为58.71,69.53 m³/hm²。与球状模型相比,二项式模型的蓄积量饱和点估测方法更科学合理,是一种确定蓄积量遥感估测饱和点简便方法。与多元逐步回归模型相比,所提出的二项式模型估测森林蓄积量的方法效果更好。

关键词: 遥感影像, 森林蓄积量, 饱和点, 二项式估测模型, 马尾松

Abstract:

This paper takes 72 Pinus massoniana forest plots (25 m × 25 m) in Hunan Province as the research object based on the method of estimating the forest biomass saturation point proposed in previous studies (Scheme 1) and propose a simpler binomial saturation point estimation method (Scheme 2).According to the saturation point estimation method,the study proposes a new stock volume estimation model for a comparative analysis with multiple stepwise regressions.It finds that the maximum saturation point estimation results of Scheme 1 and Scheme 2 were 217.05 m³/hm² and 206.86 m³/hm²,respectively.Based on the vegetation index information,the maximum saturation point estimation results of Scheme 1 and Scheme 2 were 196.95 m³/hm² and 183.06 m³/hm²,respectively.In the modeling test phase,based on the newly-proposed binomial model and the multiple stepwise regression mode,R² was 0.49 and 0.29,MAE was 53.76 m³/hm² and 63.35 m³/hm²,and RMSE was 58.71 m³/hm² and 69.53 m³/hm²,respectively.Compared with Scheme 1,the method for estimating the saturation point of forest stock volume (FSV) in Scheme 2 was more scientific and reasonable,and it was a simple method for determining the saturation point of the FSV by remote sensing data.In addition,compared with the multiple stepwise regression model,the binomial model proposed in this study had a better effect in estimating FSV.

Key words: remote sensing image, forest stock volume, saturation point, binomial estimation model, Pinus massoniana

中图分类号:

S758.51

孙忠秋, 吴发云, 胡杨, 高显连, 高金萍. 基于Landsat-8 OLI数据的马尾松林蓄积量饱和点确定及估测[J]. 林业资源管理, 2020,(6): 135-142.

SUN Zhongqiu, WU Fayun, HU Yang, GAO Xianlian, Gao Jingping. Determination and Estimation of Pinus massoniana Stand Volume and Saturation Point Based on Landsat-8 OLI Data[J]. FOREST RESOURCES WANAGEMENT, 2020,(6): 135-142.

图/表 12

表1

表2

表3

表4

图1

表5

表6

图2

表7

图3

表8

图4

参考文献 24

[1]	毛学刚, 姚瑶, 范文义. 基于Landsat长时间序列的森林扰动参数提取与树高估算[J]. 林业科学, 2019,55(3):79-87.
[2]	周小成, 何艺, 黄洪宇, 等. 基于两期无人机影像的针叶林伐区蓄积量估算[J]. 林业科学, 2019,55(11):117-125.
[3]	韩宗涛, 江洪, 王威, 等. 基于多源遥感的森林地上生物量KNN-FIFS估测[J]. 林业科学, 2018,54(9):70-79.
[4]	庞勇, 蒙诗栎, 李增元. 机载高分辨率遥感影像的傅氏纹理因子估测温带森林地上生物量[J]. 林业科学, 2017,53(3):94-104.
[5]	严恩萍, 赵运林, 林辉, 等. 基于地统计学和多源遥感数据的森林碳密度估算[J]. 林业科学, 2017,53(7):72-84.
[6]	祖笑锋, 覃先林, 李增元, 等. 基于幂律分布的森林燃烧生物量卫星遥感估测方法[J]. 林业科学, 2017,53(10):90-99.
[7]	郭云, 李增元, 陈尔学, 等. 甘肃黑河流域上游森林地上生物量的多光谱遥感估测[J]. 林业科学, 2015,51(1):140-149. doi: 10.11707/j.1001-7488.20150117
[8]	戚玉娇, 李凤日. 基于KNN方法的大兴安岭地区森林地上碳储量遥感估算[J]. 林业科学, 2015,51(5):46-55. doi: 10.11707/j.1001-7488.20150506
[9]	李亦秋, 冯仲科, 邓欧, 等. 基于3S技术的山东省森林蓄积量估测[J]. 林业科学, 2009(9):85-93. doi: 10.11707/j.1001-7488.20090915
[10]	Chrysafis I, Mallinis G, Tsakiri M, et al. Evaluation of single-date and multi-seasonal spatial and spectral information of Sentinel-2 imagery to assess growing stock volume of a Mediterranean forest[J]. International Journal of Applied Earth Observation and Geoinformation, 2019,77:1. doi: 10.1016/j.jag.2018.12.004
[11]	Hu Y, Xu X, Wu F, et al. Estimating Forest Stock Volume in Hunan Province,China,by Integrating In Situ Plot Data,Sentinel-2 Images,and Linear and Machine Learning Regression Models[J]. Remote Sensing, 2020,12(1):186. doi: 10.3390/rs12010186
[12]	Mura M, Bottalico F, Giannetti F, et al. Exploiting the capabilities of the Sentinel-2 multi spectral instrument for predicting growing stock volume in forest ecosystems[J]. International Journal of Applied Earth Observation and Geoinformation, 2018,66:126. doi: 10.1016/j.jag.2017.11.013
[13]	Redowan M, Akter R, Islam M, et al. Estimating growing stock volume in a Bangladesh forest site using landsat TM and field-measured data[J]. International Journal of Geomatics and Geosciences, 2015,6(2):1607-1619.
[14]	Tokola T. The Influence of field sample data location on growing stock volume estimation in landsat TM-based forest inventory in Eastern finland[J]. Remote Sensing of Environment, 2000,74(3):422-431. doi: 10.1016/S0034-4257(00)00135-8
[15]	Zheng S, Cao C, Dang Y, et al. Retrieval of forest growing stock volume by two different methods using Landsat TM images[J]. International Journal of Remote Sensing, 2014,35(1):29-43. doi: 10.1080/01431161.2013.860567
[16]	王原. 湖南省森林资源保护策略研究[D]. 长沙:中南林业科技大学, 2019.
[17]	杨立岩. 多平台多源遥感测树因子提取技术与方法研究[D]. 北京:北京林业大学, 2018.
[18]	赵盼盼. 基于Landsat TM和ALOS PALSAR数据的森林地上生物量估测研究[D]. 杭州:浙江农林大学, 2016.
[19]	张彦林. Box-Cox变换在遥感数据建模中的应用[J]. 东北林业大学学报, 2010,38(8):120-122.
[20]	周洪泽, 曲智林. 连续森林清查应用地理信息估计蓄积量的方法[J]. 东北林业大学学报, 2000,28(5):50-53.
[21]	张秋江, 袁亚东. 用遥感及实测回归估计森林蓄积量[J]. 东北林业大学学报, 1997,25(2):18-23.
[22]	Huang W, Swatantran A, Duncanson L, et al. County-scale biomass map comparison:a case study for Sonoma,California[J]. Carbon Management, 2017,8(5-6):417-434. doi: 10.1080/17583004.2017.1396840
[23]	Wang J, Xiao X, Bajgain R, et al. Estimating leaf area index and aboveground biomass of grazing pastures using Sentinel-1,Sentinel-2 and Landsat images[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2019,154:189-201. doi: 10.1016/j.isprsjprs.2019.06.007
[24]	Xia H, Zhao W, Li A, et al. Subpixel Inundation Mapping Using Landsat-8 OLI and UAV Data for a Wetland Region on the Zoige Plateau,China[J]. Remote Sensing, 2017,9(1):31. doi: 10.3390/rs9010031

类别	最小值/(m³/hm²)	最大值/(m³/hm²)	平均/(m³/hm²)	标准误差/(m³/hm²)	方差/(m³/hm²)	观测数/个
总样本	4.25	318.47	125.36	9.31	6234.29	72
建模样本	5.66	318.47	117.01	10.73	5871.01	51
验证样本	4.25	309.98	145.65	18.05	6843.92	21

波段	波段名称	波长范围/μm	空间分辨率/m
B2	Blue	0.452~0.512	30
B3	Green	0.533~0.590	30
B4	Red	0.636~0.673	30
B5	NIR	0.851~0.879	30
B6	SWIR1	1.566~1.651	30
B7	SWIR2	2.107~2.294	30

	FSV	B2	B3	B4	B5	B6	B7
FSV	1
B2	-0.515^**	1
B3	-0.625^**	0.919^**	1
B4	-0.559^**	0.939^**	0.948^**	1
B5	-0.388^**	0.511^**	0.685^**	0.501^**	1
B6	-0.545^**	0.835^**	0.914^**	0.839^**	0.835^**	1
B7	-0.567^**	0.924^**	0.937^**	0.956^**	0.621^**	0.934^**	1

参数	球状模型		二项式模型
参数	估计	标准误	估计	标准误
a₀	723.105	25.051	756.247	27.216
a₁	-2.085	0.278	-3.149	0.430
a₂	1.475E-05	0.000	0.008	0.001

	FSV	DVI	RDVI	EVI	NDVI	GNDVI	LAI	RVI
FSV	1
DVI	-0.279^*	1
RDVI	-0.110	0.944^**	1
EVI	-0.224	0.984^**	0.966^**	1
NDVI	0.408^**	0.261^*	0.560^**	0.360^**	1
GNDVI	0.451^**	0.306^**	0.583^**	0.389^**	0.952^**	1
LAI	-0.224	0.984^**	0.966^**	1.000^**	0.360^**	0.389^**	1
RVI	0.365^**	0.290^*	0.577^**	0.382^**	0.974^**	0.946^**	0.382^**	1