Characterizing L-Band Scattering of Paddy Rice in Southeast China With Radiative Transfer Model and Multitemporal ALOS/PALSAR Imagery
Characterizing L-Band Scattering of Paddy Rice in Southeast China With Radiative Transfer Model and Multitemporal ALOS/PALSAR Imagery
复制标题
DOI：
10.1109/tgrs.2008.2008309
复制
发表时间：
2009-03
影响因子：
8.2
通讯作者：
Cuizhen Wang;Jiaping Wu;Yuan Zhang-;G. Pan;J. Qi;W. Salas
中科院分区：
文献类型：
--
作者：
Cuizhen Wang;Jiaping Wu;Yuan Zhang-;G. Pan;J. Qi;W. Salas
水稻是中国东南部的主要粮食供应。随着人口和城市化的增加，可靠的水稻测绘在该地区至关重要。由于水稻生长期云量和降水频繁，光学遥感技术难以进行大面积的水稻监测。L波段合成孔径雷达具有全天候昼夜成像和树冠穿透能力，提供了一种独特的替代办法。在这项研究中，一阶辐射传输模型被开发来模拟L波段水稻散射特性。利用2007年6月28日、8月13日和9月28日获得的3幅高级陆地观测卫星（ALOS）/相控阵L波段合成孔径雷达（PALSAR）双偏振模式（HH和HV）图像，对水稻后向散射的时间变化进行了分析。结果表明，株高和叶片质量是影响水稻后向散射的两个主要结构参数。模拟HH后向散射的变化与样本场中的PALSAR观测结果相匹配，尽管模拟后向散射系数比图像提取值低约3 dB。叶体积散射和叶-地双反射是L波段HH极化的两个主要散射分量，并随叶层高度和密度的增加而增加。这表明，L波段HH后向散射比VV后向散射对水稻的结构变异更敏感，因此在水稻作图和建模研究中可能更有用。
Rice is a major food supply in southeast China. With increased population and urbanization, reliable rice mapping is critical in this region. Because of frequent cloud cover and precipitation during the rice-growing season, it is difficult to conduct large-area rice monitoring with optical remote sensing techniques. L-band synthetic aperture radar (SAR), with its all-weather day and night imaging and canopy penetration capabilities, provides a unique alternative. In this study, a first-order radiative transfer model was developed to simulate L-band scattering properties of paddy rice. Three Advanced Land Observing Satellite (ALOS)/Phased-Array-Type L-band Synthetic Aperture Radar (PALSAR) images in dual-polarization mode (HH and HV) acquired in early tillering (June 28, 2007), tillering (August 13, 2007), and heading (September 28, 2007) stages were processed to test the temporal variation of rice backscatter. It was found that plant height and leaf mass amount were the two major structural parameters that contributed to rice backscatter in PALSAR images. The variation of the simulated HH backscatter matched with PALSAR observations in sample fields, although the simulated backscatter coefficients were around 3 dB lower than image-extracted values. Leaf volume scattering and leaf-ground double bounce were found as the two major scattering components in L-band HH polarization and increased with leaf layer height and density. This paper demonstrated that L-band HH backscatter was more sensitive to rice's structural variation than the VV backscatter and may therefore be more useful in rice mapping and modeling studies.