Above-Ground Biomass Prediction of Arbor Forest in Altay Mountain Area Based on High-Resolution Remote Sensing Data

doi:10.13466/j.cnki.lyzygl.2023.02.014

Abstract

Abstract:

In order to accurately and conveniently estimate forest biomass at the regional scale,remote sensing characteristic variables such as vegetation index and texture were extracted based on high-resolution remote sensing data and field survey data,and the nearest neighbor algorithm (k-NN) was used to construct a forest aboveground biomass prediction model.The results showed that using k-NN to quantitatively estimate the biomass of tree forests at the regional scale,when k value was 2 and the characteristics were B1 (band 1),SR (simple vegetation index),NDVI (normalized vegetation index) and B4 (band 4),the forest biomass estimation results were optimal.The above ground biomass was 8 039 000 tons,and the average biomass per unit area was 82.15 t/hm².When the main age group of arbor forest wasmature forest,its area and biomass ratio werethe highest.The unit biomass of arbor forest was higher in the altitude range of 1 500 ~ 2 400m.

Key words: arbor forest, above-ground biomass, nearest neighbor algorithm (k-NN), remote sensing inversion, feature selection

CLC Number:

ZHANG Huifang, ZHU Yali, ZHANG Jinglu, GAO Jian, DILIXIATI·Baoerhan . Above-Ground Biomass Prediction of Arbor Forest in Altay Mountain Area Based on High-Resolution Remote Sensing Data[J]. FOREST RESOURCES WANAGEMENT, 2023, (2): 104-110.

Figures/Tables 12

Fig.1

Tab.1

Tab.2

Equations for 4vegetation indexes

植被指数	公式
归一化植被指数(NDVI)	NDVI= $N I R - R e d N I R + R e d$
简单植被指数(SR)	SR= $N I R R e d$
大气阻抗植被指数(ARVI)	ARVI= $N I R - R e d - r B l u e - R e d N I R + R e d - r B l u e - R e d$
增强型植被指数(EVI)	EVI=2.5× $N I R - R e d N I R + 6 × R e d - 7.5 × B l u e + 1$

Tab.2

Tab.3

Texture calculation method based on gray level co-occurrence matrix

纹理	公式
方差	V= $∑ i, j = 0 N - 1$ i×P_i_,_j(i-Me)²
协同性	H= $∑ i, j = 0 N - 1$ i× $P i, j 1 + (i - j) 2$
对比度	Co= $∑ i, j = 0 N - 1$ i×P_i,j(i-j)²
相异性	D= $∑ i, j = 0 N - 1$ i×P_i,j $i - j$
熵	E= $∑ i, j = 0 N - 1$ i×P_i,j(-lnP_i,j)
二阶矩	S= $∑ i, j = 0 N - 1$ i× $P i, j 2$
相关性	Cr= $∑ i, j = 0 N - 1$ i×P_i,j $(i - M e) (j - M e) V a i × V a j$

Tab.3

Tab.4

Precision index formula and description

精度指标	公式
确定系数(R²)	R²= $∑ i = 1 n (y i - y -) 2 - ∑ i = 1 n (y i - y^i) 2 ∑ i = 1 n (y i - y -) 2$
估计值的标准误差(SEE)	SEE= $∑ i = 1 n (y i - y^i) 2 / (n - p)$
平均系统误差(MSE)	MSE= $1 n$ ∑(y_i- $y^i$ )/ $y^i$ ×100%
平均预估误差(MPE)	MPE=t_a× $S E E y - i$ / $n$ ×100%
模型精度(P)	P=(1-MPE)×100%

Tab.4

Fig.2

Fig.3

Fig.4

Fig.5

Fig.6

Fig.7

Tab.5

References 31

[1]	于贵瑞, 李文华, 邵明安, 等. 生态系统科学研究与生态系统管理[J]. 地理学报, 2020, 75(12):2620-2635. doi: 10.11821/dlxb202012006
[2]	项文化, 黄志宏, 闫文德, 等. 森林生态系统碳氮循环功能耦合研究综述[J]. 生态学报, 2006(7):2365-2372.
[3]	戴巍, 赵科理, 高智群, 等. 典型亚热带森林生态系统碳密度及储量空间变异特征[J]. 生态学报, 2017, 37(22):7528-7538.
[4]	Thokchom A, Yadava P S. Biomass,carbon stock and sequestration potential of Schizostachyum pergracile bamboo forest of Manipur,north east India[J]. Tropical Ecology, 2017, 58(1):23-32.
[5]	Jhariya M K, Yadav D K. Biomass and carbon storage pattern in natural and plantation forest ecosystem of Chhattisgarh,India[J]. Journal of forest and environmental science, 2018, 34(1):1-11.
[6]	王振鹏, 陈金磊, 李尚益, 等. 湘中丘陵区不同恢复阶段森林生态系统的碳储量特征[J]. 林业科学, 2020, 56(5):19-28.
[7]	李路, 常亚鹏, 许仲林. 乌鲁木齐南山森林生物量和碳储量空间分布特征[J]. 水土保持研究, 2018, 25(5):72-77.
[8]	林立彬, 李铁华, 杨柳. 青冈栎混交林生物量及碳储量分布特征[J]. 水土保持学报, 2019, 33(1):139-144.
[9]	潘磊, 孙玉军, 王轶夫, 等. 基于Sentinel-1和Sentinel-2数据的杉木林地上生物量估算[J]. 南京林业大学学报:自然科学版, 2020, 44(3):149-156.
[10]	唐志明, 刘炳响, 屈宇. 河北太行山区典型水土保持林乔木层生物量及碳储量研究[J]. 林业资源管理, 2020(1):102-107.
[11]	Mitra A, Sengupta K, Banerjee K. Standing biomass and carbon storage of above-ground structures in dominant mangrove trees in the Sundarbans[J]. Forest Ecology and Management, 2011, 261(7):1325-1335. doi: 10.1016/j.foreco.2011.01.012
[12]	Gao Huilin, Dong Lihu, Li Fengri, et al. Evaluation of four methods for predicting carbon stocks of Korean pine plantations in Heilongjiang Province,China[J]. Plos One, 2015, 10(12):e0145017. doi: 10.1371/journal.pone.0145017
[13]	穆喜云, 张秋良, 刘清旺, 等. 基于激光雷达的大兴安岭典型森林生物量制图技术研究[J]. 遥感技术与应用, 2015, 30(2):220-225.
[14]	冯仲科, 黄晓东, 刘芳. 森林调查装备与信息化技术发展分析[J]. 农业机械学报, 2015, 46(9):257-265.
[15]	Gonzalez P, Battles J J, Collins B M, et al. Aboveground live carbon stock changes of California wildland ecosystems,2001—2010[J]. Forest Ecology and Management, 2015, 348:68-77. doi: 10.1016/j.foreco.2015.03.040
[16]	Bindu G, Rajan P, Jishnu E S, et al. Carbon stock assessment of mangroves using remote sensing and geographic information system[J]. The Egyptian Journal of Remote Sensing and Space Science, 2020, 23(1):1-9. doi: 10.1016/j.ejrs.2018.04.006
[17]	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, 36(1):46-51.
[18]	张景路, 张绘芳, 地力夏提·包尔汉, 等. 基于Landsat数字影像的阿尔泰山乔木林空间分布[J]. 南京林业大学学报:自然科学版, 2018, 42(4):153-158.
[19]	张绘芳, 朱雅丽, 地力夏提·包尔汉, 等. 阿尔泰山林区云杉和落叶松生物量分配格局研究[J]. 南京林业大学学报:自然科学版, 2017, 41(1):203-208.
[20]	张绘芳, 高亚琪, 李霞. 新疆西伯利亚云杉生物量模型研究[J]. 中南林业科技大学学报, 2015, 35(11):115-120.
[21]	蒙诗栎, 庞勇, 张钟军, 等. WorldView-2纹理的森林地上生物量反演[J]. 遥感学报, 2017, 21(5):812-824.
[22]	菅永峰, 韩泽民, 黄光体, 等. 基于高分辨率遥感影像的北亚热带森林生物量反演[J]. 生态学报, 2021, 41(6):2161-2169.
[23]	王思宇, 聂臣巍, 余汛, 等. 结合植被指数与纹理特征的玉米冠层FAPAR遥感估算研究[J]. 作物杂志, 2021(2):183-190.
[24]	Kaufman Y J, Tanre D. Atmospherically resistant vegetation index(ARVI)for EOS-MODIS[J]. IEEE transactions on Geoscience and Remote Sensing, 1992, 30(2):261-270. doi: 10.1109/36.134076
[25]	王正兴, 刘闯, Huete A. 植被指数研究进展:从AVHRR-NDVI到MODIS-EVI[J]. 生态学报, 2003(5):979-987.
[26]	Sun Qian, Sun Lin, Shu Meiyan, et al. Monitoring maize lodging grades via unmanned aerial vehicle multispectral image[J]. Plant Phenomics, 2019(2):1-16.
[27]	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.
[28]	Tomppo E, Halme M. Using coarse scale forest variables as ancillary information and weighting of variables in k-NN estimation:A genetic algorithm approach[J]. Remote Sensing of Environment, 2004, 92(1):1-20. doi: 10.1016/j.rse.2004.04.003
[29]	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.
[30]	孙珊珊, 田昕, 谷成燕, 等. 基于KNN-FIFS的内蒙古根河森林郁闭度遥感估测研究[J]. 遥感技术与应用, 2019, 34(5):959-969.
[31]	曾伟生, 唐守正. 立木生物量方程的优度评价和精度分析[J]. 林业科学, 2011, 47(11):106-113.

样地数/块	单位面积生物量/(t/hm²)
样地数/块	最大值	最小值	平均值	中位数	标准差
51	329.23	0.505	127.37	97.63	88.26

龄组	面积占比/%	单位面积平均生物量/hm²	总生物量/t	生物量占比/%
幼龄林	0.68	85.57	52253.66	0.65
中龄林	5.42	86.72	422852.69	5.26
近熟林	19.66	87.45	1546708.33	19.24
成熟林	61.47	89.06	4924706.46	61.26
过熟林	12.77	95.07	1092503.44	13.59
总计			8039024.585