为探究GLAS波形数据在估测森林郁闭度方面的潜力,以吉林省汪清林业局经营区为研究区,利用高斯低通滤波器对GLAS波形数据进行平滑滤波,从平滑后的GLAS波形数据中提取比值能量参数(I)和差值能量参数(ec),针对不同森林类型分别建立森林郁闭度单变量模型和多变量模型。研究结果表明:利用参数I建立的单变量模型优于利用参数ec建立的单变量模型; 而利用参数I和ec建立的多变量模型明显优于单变量模型。对阔叶林来说,森林郁闭度模型的决定系数(R2adj)和均方根误差(RMSE)分别为0.72和0.07,模型验证的R2adj为0.74,RMSE为0.06; 而对于针叶林,模型的R2adj为0.80,RMSE为0.10,模型验证的R2adj为0.76,RMSE为0.11; 混交林模型的精度在阔叶林和针叶林之间,模型的R2adj为0.75,RMSE为0.09,模型验证的R2adj和RMSE分别为0.71和0.07。因此,GLAS波形数据在估测森林郁闭度方面具有一定的潜力,将参数I和ec联合能够提高GLAS波形数据估测森林郁闭度的精度。
Abstract
Using the Wangqing forestry area in Jilin Province as a study area, the potential ability of GLAS waveform for estimating forest canopy density was explored in this study. The GLAS waveform was smoothed and fitted by Gaussian low pass filters, and then waveform parameters(i.e. I and ec)were extracted from the smoothed GLAS waveform. The single-variable model and multi-variable model were developed with the two waveform parameters, respectively. The results showed that the single-variable model built with I were superior to that developed with ec. The accuracy of multi-variable models was better than that of the single-variable models. For broad-leaf forest, the R2adj and RMSE(σRMSE)were 0.72 and 0.07, respectively, and the validation results were R2adj=0.74 and σRMSE=0.06. The model accuracy of coniferous forest was the highest(R2adj=0.80, σRMSE=0.10)with validation results of R2adj=0.76 and σRMSE=0.11. The model accuracy of the mixed forest was higher than the broad-leaf forest, but lower than the coniferous forest. The values of R2adj and RMSE were 0.75 and 0.09, respectively, with the validation results of R2adj=0.71 and σRMSE=0.07. The results demonstrate that GLAS waveforms have the capability to estimate the forest canopy density, and the estimation accuracy can be improved by combining I and ec.
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 杜晓明,蔡体久,琚存勇.采用偏最小二乘回归方法估测森林郁闭度[J].应用生态学报,2008,19(2):273-277.
Du X M, Cai T J, Ju C Y. Estimation of forest canopy closure by using partial least square regression[J].Chinese Journal of Applied Ecology, 2008, 19(2): 273-277.
[2] 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.
[3] 李崇贵,蔡体久.森林郁闭度对蓄积量估测的影响规律[J].东北林业大学学报,2006,34(1):15-17.Doi:10.3969/j.issn.1000-5382.2006.01.006.
Li C G, Cai T J. Effect of forest canopy density on stock volume estimation[J]. Journal of Northeast Forestry University, 2006, 34(1): 15-17.
[4] Pu R, Gong P. Wavelet transform applied to EO-1 hyperspectral data for forest LAI and crown closure mapping[J]. Remote Sensing of Environment, 2004, 91(2): 212-224.Doi:10.1016/j.rse.2004.03.006.
[5] 杨存建,倪静,周其林,等.不同林分郁闭度与遥感数据的相关性[J].生态学报,2015,35(7):2119-2125.Doi:10.5846/stxb201306101626.
Yang C J, Ni J, Zhou Q L, et al. Correlation analysis of canopy density with remote sensing data for different forest stand[J]. Acta Ecologica Sinica, 2015, 35(7): 2119-2125.
[6] 谭炳香.高光谱遥感森林类型识别及其郁闭度定量估测研究[D].北京:中国林业科学研究院,2006.
Tan B X. extraction of forest types and estimation of forest canopy closure from hyperspectral remote sensing[D]. Beijing: Chinese Academy of Forestry, 2006.
[7] Lefsky M A, Hudak A T, Cohen W B, et al. Geographic variability in lidar predictions of forest stand structure in the Pacific Northwest[J]. Remote Sensing of Environment, 2005, 95(4): 532-548.Doi:10.1016/j.rse.2005.01.010.
[8] Koch B. Status and future of laser scanning, synthetic aperture radar and hyperspectral remote sensing data for forest biomass assessment[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2010, 65(6): 581-590.Doi:10.1016/j.isprsjprs.2010.09.001.
[9] 穆喜云,张秋良,刘清旺,等.基于机载 LiDAR 数据的林分平均高及郁闭度反演[J].东北林业大学学报,2015,43(9):84-89.Doi:10.3969/j.issn.1000-5382.2015.09.017.
Mu X Y, Zhang Q L, Liu Q W, et al. Inversion of forest height and canopy closure using airborne LiDAR data[J]. Journal of Northeast Forestry University, 2015, 43(9):84-89.
[10] Morsdorf F, Kötz B, Meier E, et al. Estimation of LAI and fractional cover from small footprint airborne laser scanning data based on gap fraction[J]. Remote Sensing of Environment, 2006, 104(1): 50-61.
[11] Schutz B E, Zwally H J, Shuman C A, et al. Overview of the ICESat Mission[J]. Geophysical Research Letters, 2005, 32(21): 97-116.Doi:10.1029/2005g1024009.
[12] García M, Popescu S, Riaño D, et al. Characterization of canopy fuels using ICESat/GLAS data[J]. Remote sensing of Environment, 2012, 123: 81-89.
[13] Xing Y, Gier A D, Zhang J, et al. An improved method for estimating forest canopy height using ICESat-GLAS full waveform data over sloping terrain: a case study in Changbai mountains, China[J]. International Journal of Applied Earth Observation, 2010, 12(5): 385-392.Doi:10.1016/j.jag.2010.04.010.
[14] 邢艳秋,王立海.基于ICESat—GLAS完整波形的坡地森林冠层高度反演研究——以吉林长白山林区为例[J].武汉大学学报(信息科学版),2009,34(6):696-700.
Xing Y Q, Wang L H.ICESat-GLAS full waveform based study on forest canopy height retrieval in slope area-a case study of forests in Changbai Mountains, Jilin[J].Journal of Wuhan University(Information Science Edition),2009, 34(6): 696-700.
[15] Lefsky M A, Harding D J, Michael K, et al. Estimates of forest canopy height and aboveground biomass using ICESat[J]. Geophysical Research Letters, 2005, 32(22), Doi:10.1029/2005GL025518.
[16] Sun G, Ranson K J, Kimes D S, et al. Forest vertical structure from GLAS: an evaluation using LVIS and SRTM data[J]. Remote Sensing of Environment, 2008, 112(1): 107-117.Doi:10.1016/j.rse.2006.09.036.
[17] Harding D J, Carabajal C C. ICESat waveform measurements of within-footprint topographic relief and vegetation vertical structure[J]. Geophysical Research Letters, 2005, 32(21), Doi: 10.1029/2005GL023471.
基金
收稿日期:2016-01-22 修回日期:2016-04-13
基金项目:国家林业公益性行业科研专项项目(201504319); 中央高校基本科研业务费专项资金项目(2572014AB08); 国家自然科学基金面上项目(41171274)
第一作者:邱赛(qiusai1128@163.com)。*通信作者:邢艳秋(yanqiuxing@nefu.edu.cn),教授。
引文格式:邱赛,邢艳秋,田静,等. 基于ICESat-GLAS波形数据估测森林郁闭度[J]. 南京林业大学学报(自然科学版),2016,40(5):99-106.
PDF(2605661 KB)
