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

doi:10.11707/j.1001-7488.20170311

摘要/Abstract

摘要： [目的] 从反映森林冠层大小的树冠纹理结构出发，利用高空间分辨率遥感影像中树冠纹理的周期性信息，提取基于傅里叶变换纹理序列的纹理指数（FOTO，Fourier-based textural ordination）估测森林地上生物量，探究FOTO纹理因子在温带森林生物量估测上的潜力，为提取新型纹理参数估算森林生物量提供新的参考途径。[方法] 以2009年9月获取的小兴安岭地区凉水国家自然保护区（47°11'N，128°53'E）高分辨率机载航空影像（空间分辨率0.5 m）为例，通过提取CCD影像的FOTO纹理参数，采用多元逐步回归方法对森林地上生物量进行参数反演，并对CCD 3个波段影像提取的9个FOTO纹理因子以及波段平均影像提取的3个FOTO纹理因子2种方法的生物量估测结果进行比较。同时，在研究中尝试采用5种不同尺寸（60 m×60 m，80 m×80 m，100 m×100 m，120 m×120 m和150 m×150 m）的窗口，产生不同尺寸的FOTO因子与生物量进行回归建模。最后，将FOTO纹理因子作为自变量与激光雷达反演的参考生物量进行拟合，利用多元逐步回归方法建立生物量模型，并采用十折交叉验证评估预测模型的泛化能力。[结果] FOTO纹理因子与森林生物量的相关性较高，CCD影像3个波段的9个FOTO纹理因子与生物量的R²均高于0.67，窗口60 m×60 m，80 m×80 m，100 m×100 m，120 m×120 m和150 m×150 m的估测精度分别为67.3%，73.4%，74.4%，78.3%和80.9%。CCD影像波段平均影像的3个FOTO纹理因子与生物量的R²均高于0.57，5种窗口尺寸的估测精度分别为58.2%，62.1%，64.3%，67.4%和70.9%。根据最优预测模型获得分辨率100 m的凉水试验区全覆盖生物量结果图，精度为74.41%，RMSE为50.55 t·hm^-2。[结论] 基于FOTO算法提取的纹理因子与森林地上生物量密切相关且无明显饱和现象，对我国北方温带混交林区的生物量反演有极大潜力。FOTO纹理因子与森林地上生物量的多元线性逐步回归模型R²达0.81，RMSE为46.78 t·hm^-2。

关键词: 温带森林, 高分辨率遥感影像, 森林地上生物量, FOTO算法, 纹理因子

Abstract: [Objective] The periodical information of forest areas in the high spatial resolution remote sensing image contains the structural and spatial distribution of forest canopy grains.We used the Fourier-based textural ordination (FOTO) indices for estimating forest aboveground biomass (AGB). This research explored FOTO textural indices as a potential technique to estimate forest AGB of temperate forest.[Method] The study area is located in the Liangshui Natural Reserve (47°11'N, 128°53'E), northeast of China. Based on the high spatial resolution airborne CCD data (0.5 m spatial resolution) acquired in September 2009, we derived the FOTO indices from CCD data and established the AGB regression model with multiple stepwise regression. The airborne LiDAR-derived AGB map was used as reference value. We compared the model performances of FOTO indices derived from three spectral reflection bands of CCD data with those from the average spectral reflection band of the CCD data. The window sizes for FOTO method were set as 60 m×60 m, 80 m×80 m, 100 m×100 m, 120 m×120 m and 150 m×150 m. Then the FOTO indices derived from different size windows were used as independent variables to build regression model with LiDAR-derived biomass. Ten-fold validation was performed to verify the generalization capability of the estimation model.[Result] The results showed that texture indices derived from FOTO method had a high correlation with forest AGB. The determination coefficient R² between the FOTO indices (9 FOTO indices from three spectral reflection bands) and LiDAR derived AGB were all above 0.67 for five windows sizes. The estimation accuracies were 67.3%, 73.4%, 74.4%, 78.3% and 80.9% for window size 60 m×60 m, 80 m×80 m, 100 m×100 m, 120 m×120 m and 150 m×150 m, respectively. The determination coefficient R² between the FOTO indices (3 FOTO indices from the average band) and LiDAR derived AGB were all above 0.57 for five windows sizes. The estimation accuracies were 58.2%, 62.1%, 64.3%, 67.4% and 70.9% for five windows sizes, respectively. We produced a wall-to-wall forest AGB map with the accuracy of 74.41% and the RMSE of 50.55 t·hm^-2.[Conclusion] This study results indicated that texture indices derived from FOTO method have great potential in estimating forest AGB without significant saturation phenomenon in temperate forests. FOTO indices have great potential for estimating AGB of temperate forests. The forest biomass derived from FOTO indices with the multiple stepwise regression showed a good relationship with the LiDAR-derived AGB, with R²=0.81, RMSE=46.78 t·hm^-2.

Key words: temperate forest, high spatial resolution data, forest AGB, FOTO, texture indices

中图分类号:

S757

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.

Pang Yong, Meng Shili, Li Zengyuan. Temperate Forest Aboveground Biomass Estimation Using Fourier-Based Textural Ordination (FOTO) Indices from High Resolution Aerial Optical Image[J]. Scientia Silvae Sinicae, 2017, 53(3): 94-104.

参考文献

陈尔学,李增元,庞勇,等.2007.基于极化合成孔径雷达干涉测量的平均树高提取技术.林业科学,43(4):66-70.
(Chen E X, Li Z Y, Pang Y, et al. 2007. Polarimetric SAR interferometry based mean tree height extraction technique. Scientia Silvae Sinicae, 43(4):66-70. [in Chinese])
郭庆华,刘瑾,陶胜利,等.2014.激光雷达在森林生态系统监测模拟中的应用现状与展望.科学通报,59(6):459-478.
(Guo Q H, Liu J, Tao S L, et al. 2014. Perspectives and prospects of LiDAR in forest ecosystem monitoring and modeling. Chinese Science Bulletin, 59(6):459-478. [in Chinese])
李德仁,王长委,胡月明,等.2012.遥感技术估算森林生物量的研究进展.武汉大学学报:信息科学版,37(6):631-635.
(Li D R, Wang C W, Hu Y M, et al. 2012. General review on remote sensing-based biomass estimation. Geomatics and Information Science of Wuhan University, 37(6):631-635. [in Chinese])
李文梅,陈尔学,李增元,等.2014.森林地上生物量的极化干涉SAR相干层析估测方法.林业科学,50(2):70-77.
(Li W M, Chen E X, Li Z Y, et al. 2014. Forest aboveground biomass estimation using polarimetric interferometry SAR coherence tomography. Scientia Silvae Sinicae, 50(2):70-77. [in Chinese])
林清泉.2009.计量经济学.2版.北京:中国人民大学出版社,170-190.
(Lin Q Q. 2009. Econometrics. 2nd ed. Beijing:China Renmin University Press, 170-190. [in Chinese])
刘丽娟,庞勇,范文义,等.2013.机载LiDAR和高光谱融合实现温带天然林树种识别.遥感学报,17(3):679-695.
(Liu L J, Pang Y, Fan W Y, et al. 2013. Fused airborne LiDAR and hyperspectral data for tree species identification in a natural temperate forest. Journal of Remote Sensing, 17(3):679-695. [in Chinese])
庞勇,李增元,陈尔学,等.2005.激光雷达技术及其在林业上的应用.林业科学,41(3):129-136.
(Pang Y, Li Z Y, Chen E X, et al. 2005. LiDAR remote sensing technology and its application in forestry. Scientia Silvae Sinicae, 41(3):129-136. [in Chinese])
庞勇,李增元.2012.基于机载激光雷达的小兴安岭温带森林组分生物量反演.植物生态学报,36(10):1095-1105.
(Pang Y, Li Z Y. 2012. Inversion of biomass components of the temperate forest using airborne LiDAR technology in Xiaoxing'an Mountains, Northeastern of China. Chinese Journal of Plant Ecology, 36(10):1095-1105. [in Chinese])
徐新良,曹明奎.2006.森林生物量遥感估算与应用分析.地球信息科学,8(4):122-128.
(Xu X L, Cao M K. 2006. An analysis of the applications of remote sensing method to the forest biomass estimation. Geo-information Science, 8(4):122-128. [in Chinese])
Barbier N, Couteron P, Proisy C, et al. 2010. The variation of apparent crown size and canopy heterogeneity across lowland Amazonian forests. Global Ecology and Biogeography, 19(1):72-84.
Couteron P. 2002. Quantifying change in patterned semi-arid vegetation by Fourier analysis of digitized aerial photographs. International Journal of Remote Sensing, 23(17):3407-3425.
Couteron P, Pelissier R, Nicolini E A, et al. 2005. Predicting tropical forest stand structure parameters from Fourier transform of very high resolution remotely sensed canopy images. Journal of Applied Ecology, 42(6):1121-1128.
Couteron P, Barbier N, Gautier D. 2006. Textural ordination based on Fourier spectral decomposition:a method to analyze and compare landscape patterns. Landscape Ecology, 21(4):555-567.
Eckert S. 2012. Improved forest biomass and carbon estimations using texture measures from WorldView-2 satellite data. Remote Sensing, 4(4):810-829.
Efron B, Gong G. 1983. A leisurely look at the bootstrap, the jackknife, and cross-validation. The American Statistician, 37(1):36-48.
Houghton R A. 2005. Aboveground forest biomass and the global carbon balance. Global Change Biology, 11(6):945-958.
Lu D, Batistella M. 2005. Exploring TM image texture and its relationships with biomass estimation in Rondônia, Brazilian Amazon. Acta Amazonica, 35(2):249-257.
Meng S, Pang Y, Zhang Z, et al. 2016. Mapping aboveground biomass using texture indices from aerial photos in a temperate forest of northeastern China. Remote Sensing, 8(3):230.
Picard R R, Cook R D. 1984. Cross-validation of regression models. Journal of the American Statistical Association, 79(387):575-583.
Ploton P, Pélissier R, Barbier N, et al. 2013. Canopy texture analysis for large-scale assessments of tropical forest stand structure and biomass. Springer New York, 237-245.
Ploton P, Pélissier R, Proisy C, et al. 2012. Assessing aboveground tropical forest biomass using google earth canopy images. Ecological Applications, 22(3):993-1003.
Proisy C, Couteron P, Fromard F. 2007a. Predicting and mapping mangrove biomass from canopy grain analysis using Fourier-based textural ordination of IKONOS images. Remote Sensing of Environment, 109(3):379-392.
Proisy C, Couteron P, Pélissier R, et al. 2007b. Monitoring canopy grain of tropical forest using Fourier-based textural ordination of veryhigh resolution images. IEEE Geoscience and Remote Sensing Symposium, 4324-4326.
Proisy C, Barbier N, Guéroult M, et al. 2011. Biomass prediction in tropical forests:the canopy grain approach. Remote Sensing of Biomass:Principles and Applications, 1-18.
Sarker M, Rahman L. 2011. Estimation of forest biomass using remote sensing. Doctor of Philosophy thesis, Hong Kong Polytechnic University.
Singh M. 2012. Forest structure and biomass in a mixed forest-oil palm landscape in Borneo. Master of Philosophy (MPhil), University of Oxford.

[1]	李兰, 陈尔学, 李增元, 任冲, 赵磊, 谷鑫志. 森林地上生物量的多基线InSAR层析估测方法[J]. 林业科学, 2017, 53(11): 85-93.
[2]	冯琦, 陈尔学, 李增元, 李兰, 赵磊. 基于机载P-波段全极化SAR数据的复杂地形森林地上生物量估测方法[J]. 林业科学, 2016, 52(3): 10-22.
[3]	郭云, 李增元, 陈尔学, 田昕, 凌飞龙. 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.
[4]	李文梅, 陈尔学, 李增元, 赵磊. 森林地上生物量的极化干涉SAR相干层析估测方法[J]. 林业科学, 2014, 50(2): 70-77.
[5]	贺鹏;张会儒;雷相东;徐广;高祥. 基于地统计学的森林地上生物量估计[J]. , 2013, 49(5): 101-109.
[6]	李晓娜;国庆喜;王兴昌;郑海富. 东北天然次生林下木树种生物量的相对生长[J]. 林业科学, 2010, 46(8): 22-32.
[7]	王兴昌王传宽张全智李世业李国江. 东北主要树种心材与边材的生长特征*[J]. 林业科学, 2008, 44(5): 102-108.