TY - JOUR
T1 - Copolarized and Cross-Polarized Backscattering From Random Rough Soil Surfaces From L-Band to Ku-Band Using Numerical Solutions of Maxwell's Equations With Near-Field Precondition
AU - Liao, Tien Hao
AU - Tsang, Leung
AU - Huang, Shaowu
AU - Niamsuwan, Noppasin
AU - Jaruwatanadilok, Sermsak
AU - Kim, Seung Bum
AU - Ren, Hsuan
AU - Chen, Kuan Liang
N1 - Publisher Copyright:
© 2015 IEEE.
PY - 2016/2
Y1 - 2016/2
N2 - We extend the 3-D numerical method of Maxwell's equation (NMM3D) for rough soil surface scattering from L-band to C-, X-, and Ku-bands. We illustrate the results for copolarization, cross-polarization, and polarization ratio (HH/VV). Copolarized and cross-polarized backscattering coefficients from NMM3D are analyzed for frequency dependence, incident angle dependence, and soil moisture dependence. We also cross compare results from analytical and empirical models. The 16 × 16 squared wavelength (λ2) of rough surface is applied for NMM3D using 256 processors on NSF Extreme Science and Engineering Discovery Environment clusters. Polarization ratio, HH/VV, is studied to address the feature of dependence on frequency for same fields (same physical parameters for the model). HH/VV is shown useful to provide additional information to study land surface. Results from NMM3D are also validated with POLARSCAT measurement data-1. NMM3D shows good agreement with data and better performance while considering copolarization, cross-polarization, and polarization ratio (HH/VV) together. The key advancement in computation efficiency in this paper is the implementation of a physically based near-field precondition algorithm in NMM3D to accelerate parallel computation. With precondition, the computation time is faster by ten times for larger root-mean-square height.
AB - We extend the 3-D numerical method of Maxwell's equation (NMM3D) for rough soil surface scattering from L-band to C-, X-, and Ku-bands. We illustrate the results for copolarization, cross-polarization, and polarization ratio (HH/VV). Copolarized and cross-polarized backscattering coefficients from NMM3D are analyzed for frequency dependence, incident angle dependence, and soil moisture dependence. We also cross compare results from analytical and empirical models. The 16 × 16 squared wavelength (λ2) of rough surface is applied for NMM3D using 256 processors on NSF Extreme Science and Engineering Discovery Environment clusters. Polarization ratio, HH/VV, is studied to address the feature of dependence on frequency for same fields (same physical parameters for the model). HH/VV is shown useful to provide additional information to study land surface. Results from NMM3D are also validated with POLARSCAT measurement data-1. NMM3D shows good agreement with data and better performance while considering copolarization, cross-polarization, and polarization ratio (HH/VV) together. The key advancement in computation efficiency in this paper is the implementation of a physically based near-field precondition algorithm in NMM3D to accelerate parallel computation. With precondition, the computation time is faster by ten times for larger root-mean-square height.
KW - Near-field precondition
KW - numerical Maxwell solution
KW - polarization ratio
KW - rough surface
KW - soil moisture
UR - http://www.scopus.com/inward/record.url?scp=84939500221&partnerID=8YFLogxK
U2 - 10.1109/TGRS.2015.2451671
DO - 10.1109/TGRS.2015.2451671
M3 - 期刊論文
AN - SCOPUS:84939500221
SN - 0196-2892
VL - 54
SP - 651
EP - 662
JO - IEEE Transactions on Geoscience and Remote Sensing
JF - IEEE Transactions on Geoscience and Remote Sensing
IS - 2
M1 - 7254168
ER -