2015年黄土高原地区土地利用/覆被栅格数据

Grid data on land use and land cover of the Loess Plateau region(2015)

土地利用/覆盖是反映地表开发利用和覆被状态的重要参数,是许多地理和生态研究的基本参数。本数据集以Landsat 8卫星OLI无云或少云(云量小于10%)的遥感影像为数据源,根据影像数据满足质量要求的影像覆盖情况,利用ArcGIS软件进行解译分区。土地利用/覆被的提取采取面向对象分类方法。经eCognition8.7软件支持下的多尺度分割后,选择空间、光谱、纹理、形状作为分类特征,采取基于规则集的计算机自动分类和人工目视修改相结合的方法进行分类。利用实测的1028个样本点进行精度评估,得到数据总体精度和Kappa系数分别为85.4%和0.807。将eCongnition里解译的每个分区结果导出为shapefile格式的矢量数据,通过Arc/info AML语言脚本程序对矢量数据进行矢量–栅格转换,使用ArcGIS软件对数据按照研究区范围拼接、裁剪、重采样,得到最终分辨率为250 m的土地利用/覆被栅格数据。

Land use/cover data reflects the cultivation and coverage state of land surface.It is the basic data for geographical and ecological researches.The Landsat 8 OLI satellite images with cloud cover less than 10%were employed for the production of this dataset.The Loess Plateau region was divided into sub-regions according to the spatial cover of qualified satellite imageries using ArcGIS software.The object-oriented classification method was used to extract land use/cover classes with the support of eCognition 8.7 software.After a multi-scale segmentation procedure,parameter collection including metrics on space,spectrum,texture and shape was established.Based on these parameters,a rule set was constructed to conduct an automatic classification and a manual check and modification.The accuracy of classifications was assessed by using the 1,028 field samples.It is showed that the data set has an overall accuracy of 85.4%,and the corresponding Kappa coefficient was 0.807.The classifications of each sub-region were exported as vector data in shapefile format.Then a vector-grid conversion was conducted by Arc/info AML procedure.Stitching operation on all of the obtained grid data of sub-regions was accomplished by using ArcGIS software.After clipping and resampling procedure,the land use/cover grid data of the Loess Plateau with a resolution of 250 m was generated.