中国南方森林冠顶高度Lidar反演——以江西省为例

董立新; 李贵才; 唐世浩

doi:10.11834/jrs.20110254

遥感应用 | 浏览量 : 0 下载量: 1125 CSCD: 6

PDF
导出
分享
收藏
专辑

中国南方森林冠顶高度Lidar反演——以江西省为例
Inversion of forest canopy height in south of China by integrating GLAS and MERSI:The case of Jiangxi province in China
2011年15卷第6期页码：1301-1314
纸质出版：2011
DOI： 10.11834/jrs.20110254
稿件说明：

移动端阅览

[1]董立新,李贵才,唐世浩.中国南方森林冠顶高度Lidar反演——以江西省为例[J].遥感学报,2011,15(06):1301-1314. DOI： 10.11834/jrs.20110254.

Inversion of forest canopy height in south of China by integrating GLAS and MERSI:The case of Jiangxi province in China[J]. Journal of Remote Sensing, 2011, 15(6): 1301-1314. DOI： 10.11834/jrs.20110254.

摘要

激光雷达（Lidar）与光学遥感的有效结合对中国南方区域森林冠顶高度反演意义重大

而国产卫星将为中国森林生态研究提供新的数据源。本文联合利用大脚印激光雷达GLA和国产MERSI数据

在实现GLAS波形数据处理和不同地形条件下森林冠顶高度反演算法基础上

建立了区域尺度不同森林类型林分冠顶高度GLAS+MERSI联合反演关系模型

进行了江西地区森林冠顶高度反演。总体上

GLAS激光雷达森林冠顶高度估算精度较高;且在与MERSI250m数据的联合反演模型中

针叶林模型精度较好（R2=0.7325）;阔叶林次之（R2=0.6095）;混交林较差（R2=0.4068）。分析发现

考虑了光学遥感生物物理参数的GLAS+MERSI联合关系模型在区域森林冠顶高度估算中有较高精度

且在空间分布上与土地覆盖数据分布特征非常一致。

Abstract

For inversion of forest canopy height in large scale

it is of great signi? cance to integrate space-borne Lidar and optical remote sensing data effectively.The homemade satellite will provide a plentiful datum for forest ecological researches.In this paper

the processing of GLAS waveform data and the algorithm of forest canopy height in different terrain were implemented.The GLAS + MERSI joint inversion model of canopy height of different forest types in regional scale was established and used to map the forest canopy height of Jiangxi province.Overall

high accuracy was observed for the canopy height estimated by GLAS + MERSI joint inversion model with R2=0.733 for the needle-leaf forest

following by the broadleaf forest（R2=0.610）.The results showed that the established model was workable.It was found that the GLAS + MERSI joint inversion model which considers the optical remote sensing of biophysical parameters can provide good estimates of forest canopy height at regional scale.The space distribution characteristic was found consistent with the data of land cover.