文献类型：会议论文

英文题名：Model Based Terrain Effect Analyses on ICEsat GLAS Waveforms

作者：Pang Yong[1] Li Zengyuan[1] Lefsky, Michael[2] Guoqing Sun[3] Yu Xinfang[4]

第一作者：庞勇

通信作者：Pang, Y[1]

机构：[1]Chinese Acad Forestry, Inst Forest Resource Informat Tech, Beijing 100091, Peoples R China;[2]Colorado State Univ, Dept Forest Range & Watershed Steward, Ft Collins, CO 80523 USA;[3]Univ Maryland, Dept Geog, College Pk, MD 20742 USA;[4]Chinese Acad Sci, Inst Geog Sci & Natl Resources Res, Beijing 100101, Peoples R China

会议论文集：IEEE International Geoscience and Remote Sensing Symposium (IGARSS)

会议日期：JUL 31-AUG 04, 2006

会议地点：Denver, CO

语种：英文

外文关键词：terrain; large footprint lidar; waveform from forests; ICEsat GLAS

年份：2006

摘要：Over the past decade there have been dramatic improvements in lidar technologies and the technology has been used successfully to estimate many forest parameters. While airborne small footprint lidar systems have been operational used in forest inventory and topographic mapping, large footprint lidar systems have demonstrated very potential results in forest parameters estimation, which is feasible for spaceborne and covers over large area. As the lidar waveform is a record of return signal as a function of time from sensor to target, the terrain within lidar footprint will affect the waveform through changing the distance of target to sensor and make effective footprint area larger. With the larger footprint size, the terrain within it is perhaps not flat but with some slopes. The Geoscience Laser Altimeter System (GLAS) aboard the lee, Cloud and land Elevation Satellite (ICESat) has a 70 m diameter laser footprint. This paper simulated the terrain effects on large footprint lidar waveform using a three dimensional lidar waveform model. The slopes vary from 0 to 45 degree with 5 degree step in the model. A simulated waveform database has been built for forest covered slopes. Then the GLAS laser2a waveforms over Northeast China have been processed. The analyse of ICEsat GLAS waveforms shows the same phenomena with simulated results. The results show that both of the ground return signals and vegetation return signals are widened and the peak decreases with slope increasing. The return singles begin earlier with slope increasing. Meanwhile, the slope angle from the signal beginning to first peak of the waveform decreased. These two phenomena cause the error in detecting signal beginning and the following forest height estimation. The recorded lidar waveform will contain false vegetation peak at certain slope level. The relationship between waveform length and terrain slope is near linear. So the terrain effects on lidar wavefrom can be corrected or minimized by subtracting a factor of slope or terrain index.