|
|||||
|
|
| 基于不同空间插值模型的川西南山地常绿阔叶林叶面积指数估测 | |
| 引用本文： | 赵安玖,陈昆,郭世刚. 基于不同空间插值模型的川西南山地常绿阔叶林叶面积指数估测[J]. 自然资源学报, 2014, 29(4): 598-609. DOI: 10.11849/zrzyxb.2014.04.005 |
| 作者姓名： | 赵安玖 陈昆 郭世刚 |
| 作者单位： | 1. 四川农业大学林学院, 四川雅安625014; 2. 荣县林业局, 四川荣县643100 |
| 基金项目： | 国家重点科技攻关项目“长江上游低山丘陵区生态综合整治技术与示范”（2011BAC09B05）。 |
| 摘 要： | 叶面积指数（LAI） 是森林经营管理的一个重要参数，亦是众多基于土壤、大气和植被相互作用的森林经营管理动态模型的关键输入变量之一，但目前获得大尺度、高空间分辨率的LAI精确估计仍十分困难。以川西南山地常绿阔叶林为研究对象，基于地面调查的83 个20 m×20 m样地和SPOT5 遥感数据，运用不同的方法，估测了区域森林有效叶面积指数（LAIe）。利用遥感数据作为辅助变量，比较了来自遥感数据的直接辐射关系法（DRR） 和地统计学方法协同克里格（CK）、回归克里格（RK） 的LAIe 空间分布差异。此外，运用反距离权重（IDW）、全局多项式（GPI）、普通克里格（OK）、泛克里格（UK） 等方法也对LAIe进行了估计。结果表明LAIe 与归一化植被指数（NDVI） 回归拟合效果相对不理想（R2=0.679，P<0.001）；由于研究区森林景观不是连续变量，LAIe呈低空间自相关性，各方法间差异不显著，对区域LAIe 估算精度提升不明显；但DRR 和CK、RK 方法能降低统计误差值，且DRR与CK、RK的相关性极高（相关系数分别可达0.91和0.92）。尽管相比于DRR，地统计学方法并没有提高LAIe估计精度，为提高预测的准确性，应探讨更密集的抽样方案和不同的辅助变量。 |
| 关 键 词： | SPOT 5 有效叶面积指数 地统计 山地常绿阔叶林 |
| 收稿时间： | 2012-11-30 |
| 修稿时间： | 2013-07-05 |
Estimation LAI of Montane Evergreen Broad-Leaved Forest in Southwest Sichuan Using Different Spatial Prediction Models |
|
| ZHAO An-jiu,CHEN Kun,GUO Shi-gang. Estimation LAI of Montane Evergreen Broad-Leaved Forest in Southwest Sichuan Using Different Spatial Prediction Models[J]. Journal of Natural Resources, 2014, 29(4): 598-609. DOI: 10.11849/zrzyxb.2014.04.005 | |
| Authors: | ZHAO An-jiu CHEN Kun GUO Shi-gang |
| Affiliation: | 1. College of Forestry, Sichuan Agricultural University, Ya'an 625014, China; 2. Forestry Bureau of Rongxian County, Rongxian 643100, China |
| Abstract: | Leaf Area Index (LAI) is a critical variable for forest management, and there are several dynamical models of forest management, based on the modeling of the interactions between the soil, the atmosphere and the vegetation. LAI is a critical input variable for these models. Currently, it is difficult to obtain accurate LAI estimations of high spatial resolution over large areas. Effective leaf area index (LAIe) of montane evergreen broad-leaved forest stands estimation was carried out in a region located in Southwest Sichuan, by means of different approaches including field inventory data, SPOT 5 imagery and spatial prediction models, LAIe was inventoried and assessed in a total of 83 sample field plots. And using remotely sensed data as auxiliary variables, LAIe spatial distribution, which is derived from Direct Radiometric Relationships (DRR), the geostatistical method Co-Kriging (CK) and Regression-Kriging (RK), respectively, were compared. Also, Inverse Distance Weighted (IDW), Global Polynomial Interpolation (GPI), Ordinary Kriging (OK), and Universal Kriging (UK) estimations were performed and tested. The results show that since forest landscape is not a continuous variable, the tested LAIe variables showed low spatial autocorrelation, which makes Kriging methods unsuitable to these purposes. But DRR, CK and RK methods produced lower statistical error values, and presented high spatial correlation existing between DRR and CK, RK methods. Despite the geostatistical method RK did not increase the accuracy of estimates developed by DRR, denser sampling schemes and different auxiliary variables should be explored, in order to test if the accuracy of predictions is improved. |
| Keywords: | effective leaf area index SPOT 5 geostatistics montane evergreen broadleaved forest |
| 本文献已被 CNKI 等数据库收录！ | |
| 点击此处可从《自然资源学报》浏览原始摘要信息 | |
| 点击此处可从《自然资源学报》下载全文 | |