论文部分内容阅读
以空间异质性较强的枯水期鄱阳湖为研究区,以搭载于同一卫星平台、具有同一观测时间和较高空间分辨率的ASTER数据为参照,分析研究了MODIS数据在土地覆盖分类中由空间尺度带来的不确定性。首先基于MODIS三角权重函数,建立了从ASTER到MODIS的尺度转换方法;然后对不同空间分辨率的数据进行土地覆盖分类,并基于误差矩阵和线性模型分析了MODIS土地覆盖分类结果的误差来源。结果表明,空间分辨率和光谱分辨率与成像方式这两类因素对MODIS与ASTER分类结果差异的贡献比例约为(6.6—11.2):2;MODIS像元尺度对研究区水体的分类不确定性影响较低,而对森林的不确定性影响可达63%。由此可见,在基于MODIS数据的土地覆盖分类研究中,空间尺度所产生的不确定性是比较显著的。这些研究结果对于土地覆盖分类及变化检测、尺度效应和景观生态学不确定性研究,有积极的参考意义。
Taking Poyang Lake in the dry season with strong spatial heterogeneity as the research area and ASTER data on the same satellite platform with the same observation time and high spatial resolution as the reference, the spatial distribution of MODIS data in the land cover classification Uncertainty of scale. First, based on the MODIS trigonometric weight function, a scale conversion method from ASTER to MODIS was established. Then the land cover classification of the data with different spatial resolution was carried out, and the error sources of MODIS land cover classification results were analyzed based on error matrix and linear model. The results show that the contribution ratio of spatial resolution, spectral resolution and imaging method to the difference between MODIS and ASTER classification results is about (6.6-11.2): 2; the classification uncertainty of MODIS pixel scale in the study area The impact is low, while the impact on forest uncertainty can reach 63%. Thus, in the study of land cover classification based on MODIS data, the uncertainty of spatial scale is more significant. These results have positive reference meaning for land cover classification, change detection, scale effect and landscape ecological uncertainty.