地球信息科学学报 ›› 2013, Vol. 15 ›› Issue (5): 752-760.doi: 10.3724/SP.J.1047.2013.00752

• 遥感科学与应用技术 • 上一篇    下一篇

农田表面粗糙度参数的测量与精度分析

郑兴明1,2, 赵凯1,2, 李晓洁1   

  1. 1. 中国科学院东北地理与农业生态研究所, 长春 130102;
    2. 中国科学院长春净月潭遥感实验站, 长春 130102
  • 收稿日期:2012-07-10 修回日期:2013-07-12 出版日期:2013-09-29 发布日期:2013-09-29
  • 通讯作者: 赵凯(1962-),男,研究员,主要从事微波辐射计研制及微波遥感基础理论研究。E-mail:zhaokai@neigae.ac.cn E-mail:zhaokai@neigae.ac.cn
  • 作者简介:郑兴明(1984-),男,博士,助理研究员,主要从事被动微波遥感辐射传输理论及地表参数反演研究。E-mail:zxm984913@163.com
  • 基金资助:

    国家“863”计划子专题(2012AA12A305-5-2);中科院知识创新项目(KZCX2-YW-340)。

Accuracy Analysis of Agriculture Soil Surface Roughness Parameter

ZHENG Xingming1,2, ZHAO Kai1,2, LI Xiaojie1   

  1. 1. Northeast Institute of Geographical and Agroecology, CAS, Changchun 130102, China;
    2. Changchun Jingyuetan Remote Sensing Test Site, CAS, Changchun 130102, China
  • Received:2012-07-10 Revised:2013-07-12 Online:2013-09-29 Published:2013-09-29

摘要:

表面粗糙度是有效解释雷达后向散射系数和微波辐射亮度温度的关键参数之一。表面粗糙度参数的测量精度受到测量方法、测量仪器、数据预处理等的影响,如何获取到表面粗糙度的“真”值是地表粗糙度测量急需解决的问题,并且有助于提高利用微波遥感技术反演地表参数的能力。本文利用激光扫描仪的二维高度数据和蒙特卡罗方法模拟的一维表面高度数据,分析了重复采样次数、采样间隔、采样剖面长度、空间自相关函数类型和大尺度结构(数据倾斜和农田垄行结构)对表面粗糙度精度的影响,研究表明:在大于20次重复采样、小于10mm的采样间隔、200倍相关长度的剖面长度的条件下,农田表面粗糙度参数的测量精度约为80%;分形相关函数与实测农田表面的空间自相关系数的吻合性要高于高斯函数和指数函数;数据倾斜和农田垄行结构严重影响表面粗糙度参数的结果,在进行表面粗糙度参数的计算之前,需从剖面高度分布数据中去除以上两个因素的影响。

关键词: 表面粗糙度, 微波遥感, 相关长度, 农田

Abstract:

Agricultural soil surface roughness parameters including surface root mean square height (rms h)and surface correlation length (cl)are key for effectively explaining radar backscattering coefficient and passive microwave brightness temperature, and its measurement accuracy is affected by measured method, measured instruments and data preprocessing. For surface roughness field measurement, the urgent problem is how to evaluate its measurement accuracy. This paper uses two-dimension surface height data from laser scanner and one-dimensional surface height data from Monte Carlo simulation, and analyzes the factors which influence the accuracy of surface roughness including sampling times, sampling interval, profile length, correlation function and large-scale structure (data skew and row structure). Four regions with different surface rough condition are used in this paper for analyzing surface roughness measurement method and accuracy. The result shows that, the measurement accuracy of agriculture soil surface roughness is about 80% under the condition of more than 20 samples, less than 10mm sampling interval and more than 200 cl profile length; the agreement between fractal correlation function and experimental spatial correlation coefficient is higher than Gaussian and exponential correlation function; and the skew and row structure of surface height data seriously affect the result of surface roughness parameter which needs to be removed before surface roughness calculation.

Key words: correlation length, agriculture soil, microwave remote sensing, surface roughness