Inversion of Masson pine forest LAI by multiple-perspective vegetation index
-
摘要: CHRIS/PROBA是目前具有最高空间分辨率（17 m×17 m）的星载多角度高光谱数据，该款数据在反演植被垂直结构参数，如树高、叶面积指数（leaf area index，LAI）等方面具有重要的应用前景。基于四尺度几何光学模型得到马尾松（Pinus massoniana Lamb.）冠层的归一化差分植被指数（normalized difference vegetation index，NDVI）各向异性分布规律，利用CHRIS红光特征波段和近红外特征波段构建一种新型多角度植被指数（normalized hotspot-dark-spot difference vegetation index，NHDVI），并将其应用于CHRIS数据对马尾松林的LAI遥感估算上。结果显示：（1）相比归一化差分植被指数（NDVI）与土壤调节植被指数（soil adjusted vegetation index，SAVI）而言，NHDVI能很好地融合光谱信息与角度信息，与地面实测LAI的决定系数达到0.7278；（2）利用NHDVI-LAI统计回归模型方法来反演LAI值，将得到的LAI值与地面实测值进行相关性分析，结果拟合优度达到0.8272，均方根误差RMSE为0.1232。与传统植被指数相比，包含角度信息的多角度植被指数对LAI的反演在精度上有较大提升，同时比基于辐射传输模型的反演方法更简易、实用。Abstract: CHRIS/PROBA is a multiple-angle sensor providing hyper-spectral data with 17 m×17 m spatial resolution that can be applied for data inversion of vegetation canopy structure parameters, such as tree height and leaf area index (LAI). We used a four-scale geometrical optics model to simulate anisotropy distribution regulation of the normalized difference vegetation index (NDVI) of a Pinus massoniana (Masson pine) forest canopy. By extracting the red and near infrared characteristic spectral bands from the 18 bands in CHRIS, a new multi-angle normalized hotspot-dark-spot difference vegetation index (NHDVI) was applied to the estimation of the LAI of Pinus massoniana forest using CHRIS data. The results showed that:(1) Compared with the NDVI and soil adjusted vegetation index (SAVI), NHDVI well integrated the spectral information and angle information with the ground measured LAI, and the coefficient of determination reached 0.7278; (2) The LAI was calculated by statistical regression of NHDVI-LAI. The correlation between the LAI and the measured values was 0.8272, significantly higher than that of SAVI, and the root mean square error (RMSE) was 0.1232. Thus, these findings indicate that angular information is important for improving the retrieval accuracy of LAI.
-
Keywords:
- Leaf area index /
- Multiple-perspective sensing /
- Vegetation index /
- Inversion /
- Masson pine forest
-
-
[1] 徐希孺. 物理遥感[M]. 北京:北京大学出版社, 2005:21-22. Xu YR. Physical Principle of Remote Sensing[M]. Beijing:Peking University Press, 2005:21-22.
[2] 盖利亚, 刘正军, 张继贤. CHRIS/PROBA高光谱数据的预处理[J]. 测绘工程, 2008,17(1):40-43. Gai LY, Liu ZJ, Zhang JX. Preprocessing of CHRIS/PROBA hyperspectral data[J]. Engineering of Surveying and Mapping, 2008, 17(1):40-43.
[3] 夏天, 吴文斌, 周清波, 周勇. 冬小麦叶面积指数高光谱遥感反演方法对比[J]. 农业工程学报, 2013, 33(3):139-147. Xia T, Wu WB, Zhou QB, Zhou Y. Comparison of two inversion methods for winter wheat leaf area index based on hyperspectral remote sensing[J]. Transactions of the Chinese Society of Agricultural Engineering, 2013, 33(3):139-147.
[4] 金晟业. 承德地区植被参数多角度遥感反演[D].长春:吉林大学, 2009:32. Jing CY. Inversion of vegetation parameters in Chengde using multi-angle satellite date[D]. Changchun:Jilin University, 2009:32. [5] Chopping M, Su L, Laliberte A, Rango A, Peters DPC, Kollikkathara N. Mapping shrub abundance in desert grasslands using geometric-optical modeling and multi-angle remote sensing with CHRIS/Proba[J]. Remote Sens Environ, 2006, 104(1):62-73.
[6] Quaife T, Barnsley M, Thackrahm TG,Lewis P. Mission status of CHRIS/PROBA and its utility for sampling the earth surface BRDF[C]//RSPS Proceedings. Rocheste:Bpydell & Brewer Publishers,2001:128-134.
[7] Miina R, Mait L, Matti M, Matti Mõttus, Andres Kuusk, Tiit Nilson, Joel Kuusk, Tõnu Lükk. Multi-angular reflectance properties of a hemiboreal forest. An analysis using CHRIS/PROBA data[J]. Remote Sens Environ, 2008, 112(5):2627-2642.
[8] 李新辉, 宋小宁, 冷佩. 利用CHRIS/PROBA数据定量反演草地LAI方法研究[J]. 国土资源遥感, 2011(3):61-66. Li XH, Song XN, Leng P. A Quantitative method for grassland LAI inversion based on CHRIS/PROBA data[J]. Remote Sensing for Land & Resources, 2011(3):61-66.
[9] 韦玮. 基于多角度高光谱CHRIS数据的湿地信息提取技术研究[D]. 北京:中国林业科学研究院, 2011. Wei W. Study on Wetland information extraction using multi-angle hyperspectral CHRIS image data[D]. Beijing:Chinese Academy of Forestry, 2011. [10] 李小梅. 多角度高光谱遥感森林类型分类方法研究[D]. 呼和浩特:内蒙古农业大学, 2010. Li XM. Study on classification of forest types using multi-angle and hyperspectral remote sening[D]. Hohhot:Inner Mongolia University, 2010. [11] 李晓敏, 张杰, 马毅, 任广波. 基于主成分分析和决策级融合的高光谱图像分类方法研究[J]. 海洋科学, 2015, 39(2):25-34. Li XM, Zhang J, Ma Y, Ren GB. Research on the classification method of the hyper-spectral image based on principal component analysis and decision-level fusion[J]. Marine Sciences, 2015, 39(2):25-34.
[12] 林培治.福建长汀水土流失区马尾松种群动态及其机制研究[D].福州:福建师范大学, 2013:23-25. Lin PZ. Population dynamics and mechanism of Pinus massoniana Lamb. in the restoring ecosystem, Changting County of Fujian Province[D]. Fuzhou:Fujian Normal University, 2013:23-25.
[13] Chen JM, Leblanc SG. A four-scale bidirectional reflectance model based on canopy architecture[J]. IEEE T Geosci Remote,1997, 35(5):1316-1337.
[14] Zhu GL, Ju W, Chen JM, Gong P, Xing B, Zhu J. Foliage clumping index over China's landmass retrieved from the MODIS BRDF parameters product[J]. IEEE T Geosci Remote, 2012, 50(6):2122-2137.
[15] Chen JM, Leblanc SG. Multiple-scattering scheme useful for geometric optical modeling[J]. IEEE T Geosci Remote, 2001, 39(5):1061-1071.
[16] Hasegawa K, Matsuyama H, Tsuzuki H, Sweda T. Improving the estimation of leaf area index by using remotely sensed NDVI with BRDF signatures[J]. Remote Sens Environ, 2010, 114(3):514-519.
[17] 张霞, 张兵, 胡方超, 童庆禧. 航天成像光谱仪CHRIS辐射与光谱性能评价[J]. 中国科学E辑:技术科学, 2006, 36(B07):85-93. Zhang X, Zhang B, Hu FC, Tong QX. Evaluation of CHRIS imaging radiation characteristics[J]. Science in China(Seris E):Technological Sciences, 2006, 36(B07):85-93.
计量
- 文章访问数: 1084
- HTML全文浏览量: 3
- PDF下载量: 1298
下载:
