基于多源数据的地表覆盖制图与分析——以广东省广州市为例

Mapping and Analysis of Land Cover Based on Multi-source Data——Taking Guangzhou City in Guangdong Province as an Example

摘要:

地表覆盖数据是关于土地利用信息的重要来源，在地理国情监测、气候变化研究、生态环境评估等方面发挥着重要的作用。本文以广东省广州市作为研究区域，利用OSM数据结合FROMGLC数据和Landsat8遥感影像数据，对研究区域地表覆盖进行分类并完成地表覆盖制图。实验结果表明，利用OSM数据进行地表覆盖制图结果总体精度较高，具有可用性，但是OSM数据空白区域占比较大，有数据区域个别要素数量较少，在参考选取样本时，样本代表性不够高，部分类型的地表覆盖分类精度较低。

Abstract:

Land cover data is an important source of land use information, which plays an important role in monitoring geographical conditions, climate change research, ecological environment assessment and so on. Taking Guangzhou city in Guangdong province as the study area, by using OSM data, FROM-GLC data and LandSAT-8 remote sensing image data, land cover in the study area has been classified and land cover mapping has been completed. It is showed that the overall accuracy of land cover mapping by using OSM data is high and has usability. But the blank areas of OSM data occupy a large proportion, the number of different elements in the data areas is small, the sample representativeness is not high enough in reference sample selection, and the accuracy of some types of land cover classification is low.