摘要
空间尺度问题是定量遥感重要而基础的问题之一,文章针对“分形方法是否适用于定量遥感地表参数的尺度转换研究”的问题进行实验验证。基于传统及改进的Chen NDVI尺度转换模型,获得NDVI不同空间升尺度影像,进而利用分形尺度转换模型分析NDVI尺度转换结果,以获取“分形方法在NDVI尺度转换研究中的适用性结论”。以厦门为研究区,利用上述分形模型进行实验,结果表明:(1)若NDVI输入类型为均值,NDVI尺度转换特性更符合双对数直角坐标系下的线性函数关系,即分形特性;若NDVI输入类型为方差、方差/均值,NDVI尺度转换特性更符合直角坐标系下的对数函数关系;(2)农田在两种Chen NDVI模型及两种空间直角坐标系下所得拟合模型皆表现出显著的线性规律,尤其是其NDVI尺度转换分形特性表现显著;(3)综合而言,NDVI尺度转换结果具有分形特性,但NDVI尺度转换具有更为显著的直角坐标系下对数函数关系特性。文章所提出的融入精细地类信息的“类NDVI”地表参数通用尺度转换模型及尺度转换分形特性研究方法具有一定的代表性,值得参考。
The spatial scale problem is one of the important and fundamental problems of quantitative remote sensing.The paper verifies the problem of whether the fractal method is suitable for the scale conversion of quantitative remote sensing surface parameters.Based on the traditional and improved Chen NDVI scale conversion model,the NDVI upscaled images on different spatial resolutions are obtained,and then the fractal scale conversion model is used to analyze the NDVI scale conversion results above to obtain the“Applicability Conclusion of Fractal Method in NDVI Scale Conversion Study”.Taking Xiamen as the research area,the above fractal model is used to test.The results show that:(1)If the NDVI input type is mean,the NDVI scale conversion characteristics are more in line with the linear function relationship in the double logarithmic Cartesian coordinate system,that is,the fractal characteristics;if the NDVI input type is variance or variance/mean,the NDVI scale conversion characteristics are more in line with the logarithmic function relationship in Cartesian coordinate system.(2)Farmland in both Chen NDVI models and two spatial Cartesian coordinate systems show a significant linearity of the fitted models.In particular,the fractal characteristics of its NDVI scale conversion are significant.(3)In general,NDVI scale conversion results have fractal characteristics,but NDVI scale conversion has a more significant logarithmic function relationship in Cartesian coordinate system.In the paper,the fractal characteristics research methodology for scale transformation,which is based on universal scale transformation model for NDVI-like land surface parameters fusing fine ground classifications information,is rather typical and can be referred in other relative research.
作者
栾海军
邢宸硕
张荣凯
何原荣
贺秋华
LUAN Haijun;XING Chenshuo;ZHANG Rongkai;HE Yuanrong;HE Qiuhua(College of Computer and Information Engineering,Xiamen University of Technology,Xiamen 361024,China;Digital Fujian Institute of Big Data for Natural Hazards Monitoring,Xiamen University of Technology,Xiamen 361024,China;Hunan Key Laboratory of Remote Sensing of Ecological Environment in Dongting Lake Area,Changsha 410008,China)
出处
《遥感信息》
CSCD
北大核心
2022年第3期12-20,共9页
Remote Sensing Information
基金
国家自然科学基金项目(41601350)
福建省自然科学基金项目(2020J01261、2019J01853)
洞庭湖区生态环境遥感监测湖南省重点实验室开放课题(DTH Key Lab.2021.016)
厦门理工学院“科研攀登计划”项目(XPDKT19010)。
