闽东南沿海地区景观指数粒度效应的高分辨率遥感分析

发布时间：2018-12-30 10:25

【摘要】：以闽东南沿海地区3个流域SPOT 5影像解译所得的土地覆被类型分布图为原始矢量数据,通过优势规则聚合为一系列不同粒度大小的栅格类型图,统计各粒度下28种常用景观指数值;采用尺度检测图分析各种景观指数的粒度变化响应规律,对比中分辨率数据与高分辨率数据所建立尺度函数的差异,评价预测响应型指数尺度下推的相对误差。结果表明:闽东南沿海地区景观指数的粒度效应明显,可分为预测响应型、分段预测型、阶梯变化型和波动变化型4类,其中分段预测型指数由已有报道的预测响应型分化而来,其尺度转折点位于5 m、7.5 m或10 m等粒度下,即高分辨率影像对景观指数粒度效应的分类有重要影响。在预测响应型指数中,高分辨率影像对ED、SHAPE_MN、CONTIG_MN和AI 4种指数的尺度函数影响较大,由中分辨率数据所建立的尺度函数进行尺度下推,会带来较大误差;这些指数的区域差异性也较大,对同一指数来说,不同区域可选用不同的尺度函数进行尺度下推。
[Abstract]:Based on the original vector data of land cover type map interpreted by SPOT 5 image of three watersheds in southeast Fujian coastal area, a series of grid type maps with different granularity were aggregated by dominant rules. Statistic 28 common landscape index values under different granularity; The response law of granularity variation of various landscape indices is analyzed by using scale detection map, the difference of scale function established by medium resolution data and high resolution data is compared, and the relative error of predictive response index scale is evaluated. The results show that the granularity effect of landscape index in southeast Fujian coastal area is obvious and can be divided into four types: predictive response type, piecewise prediction type, step change type and fluctuating change type. The Piecewise Predictive Index is differentiated from the previously reported predictive response type, and its scale turning point is at the size of 5 m or 10 m, that is, the high-resolution image has an important influence on the classification of the granularity effect of the landscape index. In the prediction response index, high resolution image has a great influence on the scale function of ED,SHAPE_MN,CONTIG_MN and AI index, and the scale function derived from the medium resolution data will bring big error. For the same index, different scale functions can be used to push down the scale.
【作者单位】：福建师范大学地理科学学院;湿润亚热带山地生态国家重点实验室培育基地;福建省陆地灾害监测评估工程技术研究中心;
【基金】：福建省高校产学合作项目(2015Y4002)
【分类号】：P901;P237

【相似文献】