论文部分内容阅读
林业研究中的主要兴趣点之一在于通过经验或半经验模型建立林分参数与遥感影像数据间的相互关系来估测林分参数。基于覆盖美国佛罗里达州东北Duval县的遥感数据和两块样地清查数据,论文探讨了所选林分参数与TM影像光谱DN值间的相关性。相关性分析结果表明,单波段或植被指数对林分参数的解释能力低于50%,为此构建了林分参数与影像多波段间多元回归模型来估测林分参数。预测结果通过另一组数据验证,除林分密度外,其它参数估测可信度达75%以上。论文最后探讨了预测模型不足和需改进的地方,并指出该研究有助于更好地理解影像光谱值和林分参数间的关系。图1表2参9。
One of the main points of interest in forestry research lies in estimating stand parameters by establishing correlations between stand parameters and remote sensing data using empirical or semi-empirical models. Based on the remote sensing data and two sample inventory data covering Duval County, northeastern Florida, USA, the paper discusses the correlation between selected stand parameters and TM spectral spectral DN values. The results of correlation analysis showed that the interpretation ability of single-band or vegetation index to the stand parameters was less than 50%. Therefore, the stand parameters and the multi-band multi-band regression model were constructed to estimate the stand parameters. The prediction results are verified by another set of data. Except for the stand density, the confidence of other parameter estimates is over 75%. Finally, the dissertation explores the shortages of prediction models and the areas to be improved, and points out that this study will help to better understand the relationship between image spectral values and stand parameters. Figure 1 Table 2 Reference 9.