复杂地形条件下联合光学数据和激光雷达数据森林生物量协同反演研究

项目名称： 复杂地形条件下联合光学数据和激光雷达数据森林生物量协同反演研究

项目编号： No.41201371

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 地理学

项目作者： 池泓

作者单位： 中国科学院测量与地球物理研究所

项目金额： 25万元

中文摘要： 森林生物量是整个森林生态系统运行的能量基础和营养物质来源，区域和全球森林生物量估算已成为森林碳储量、森林扰动以及森林生态系统动态变化研究的关键问题。本研究以森林生长模型、基于三维场景的激光雷达回波模型和可模拟复杂三维场景辐射传输过程的辐射传输模型为基础，借助激光雷达对垂直结构的敏感性和光学遥感丰富的光谱信息，探索基于光学遥感数据和激光雷达数据的三种森林生物量反演方法研究。（1）通过激光雷达回波模拟模型提取冠层高度和生物量信息，采用统计回归分析生物量与光学模型BRDF的关系建立生物量反演模型。（2）建立激光雷达波形数据库，通过实际波形数据反演树高和覆盖度，以此作为附加约束条件，从光学BRDF模拟数据库中反演生物量。（3）从实际波形数据出发，研究生物量由点及面外推的非参数化方法。通过本项目研究探索森林生物量反演方法的适用性，提高反演精度，为进一步探讨森林动态变化与全球气候变化提供帮助。

中文关键词： 森林地上生物量；大脚印激光雷达；ICESat/GLAS；光学遥感；森林资源调查数据

英文摘要： Forest biomass as one of key property of forest ecosystem is the source of energy and nutrition for whole ecosystem.Global or regional forest biomass estimation has become a critical issue of research on forest carbon stocks, forest disturbance and dynamics of forest ecosystems. Forest growth model(Zelig) simulation of forest stands were used to parameterize a three-dimensional vegetaion lidar waveform model and a radiative transfer model (RTM) in this research. Due to its measurement principle,light detection and ranging (lidar) is particularly suited to assess the horizontal and vertical canopy structure of forests, while the spectral measurements of imaging spectrometry are specifically rich on information for biophysical canopy properties Three methods for forest biomass inversion will explored based on the above analysis. (1) The model will develop by the canopy height and biomass retrieved from vegetaion lidar waveform model with RTM BRDF using statistical regression method. (2) The lidar data will used to retrieve tree height and fractional cover from the simulated waveform dataset. Then, biomass is inversed in the simulated BRDF dataset by restricting the LUT based on prior information of the canopy and fractional cover. (3) Non-parametric approaches of extrapolating the samples of forest bioamss data to

英文关键词： forest aboveground biomass；large footprint lidar；ICESat/GLAS；optical remote sensing；forest inventory data