In: Wagner W., Székely, B. (eds.): ISPRS TC VII Symposium - 100 Years ISPRS, Vienna, Austria, July 5-7, 2010, IAPRS, Vol. XXXVIII, Part 7B 
317 
THE SENSITIVITY OF MULTI-FREQUENCY (X, C AND L-BAND) RADAR 
BACKSCATTER SIGNATURES TO BIO-PHYSICAL VARIABLES (LAI) OVER CORN 
AND SOYBEAN FIELDS 
Xianfeng Jiao, Heather McNaim, Jiali Shang and Jiangui Liu 
Agriculture and Agri-Food Canada, 960 Carling Ave., Ottawa, Ontario K1A 0C6 
Key words: Agriculture, Crop, Analysis, SAR, Multifrequency 
ABSTRACT: 
The objective of this study is to investigate the sensitivity of synthetic aperture radar (SAR) backscatter signatures to crop bio 
physical variables. The experimental data were collected over com and soybean fields in eastern Ontario (Canada) during the 2008 
growing season. Remote sensing acquisitions consisted of TerraSAR-X dual-polarized stripmap data (X-band), RADARSAT-2 Fine 
beam quad-polarized data (C-band) and ALOS PALSAR dual-pol data (L-band), as well as the Compact Airborne Spectragrahic 
Imager (CASI) and SPOT-4 multi-spectral data. Plant variables, such as leaf area index (LAI) and surface volumetric soil moisture 
were measured to coincide with these acquisitions and key phenological growth stages. Analyses were conducted based on statistical 
correlation and a simple backscatter process model (the water cloud model). The results of this study show that the lower frequency 
bands, such as L and C, were closely related with LAI. For both com and soybean crops, most C-band linear (HH, VV, HV) 
backscatter coefficients were correlated with LAI; backscatter increased with increasing LAI. L-band backscatter at HH and HV 
polarizations produced the highest correlations with com LAI (r=0.90—0.96). Conversely, these L-band polarizations were only 
weakly correlated with soybean LAI. The higher frequency X-band was poorly correlated with both com and soybean LAI. Based on 
these findings, the water cloud model was used to express C-band and L-band backscatter for the whole canopy as a function of LAI 
and surface soil moisture. 
1. INTRODUCTION 
The monitoring of crop bio-physical variables is a very 
important task in agricultural management and in yield 
forecasting. Information from satellites can be exploited to 
assist in estimating key crop growth indicators including Leaf 
Area Index (LAI), biomass and crop height. LAI is an 
important indicator of agricultural productivity and a critical 
variable in crop growth models. Optical remote sensing data 
have been used to estimate LAI (Baret and Guyot, 1991; Brown 
et al., 2000; Chen and Cihlar, 1996). However, operational 
productivity and yield monitoring activities that rely solely on 
optical imagery are vulnerable to data gaps during critical crop 
growth stages as a result of unfavourable atmospheric 
conditions. Synthetic aperture radars (SARs) are unaffected by 
atmospheric haze and clouds. In addition to this oft-quoted 
advantage, SAR data also provide complementary and unique 
characterizations of vegetation when compared with the 
information provided by optical imagery. 
SAR response is dependent upon the sensor configuration 
including incidence angle, frequency and polarization. Target 
characteristics, most notably the soil and crop dielectric and 
geometric properties, influence scattering behaviour and the 
magnitude of the radar backscatter. Shorter SAR wavelengths 
such as X-band (~3 cm) and C-band (~6 cm) interact mainly 
with the top part of the canopy layers while long wavelengths 
such as L-band (~20 cm) have a greater penetration depth, 
interacting with the entire crop canopy and resulting in greater 
scattering contributions from the soil (Ulaby et al., 1984). 
Penetration depth depends on whether the bio-physical 
parameters of the scatters within a vegetation layer (e.g., 
canopy water content, size and geometry of the canopy 
components) scatter or attenuate the incident microwaves. 
Inoue et al. (2002) compared backscatter responses from multi 
frequency (Ka, Ku, X, C, and L) data in the context of several 
bio-physical variables of paddy rice. The results showed that 
LAI was best correlated with HH- and cross-polarization 
backscatter at C-band, while fresh biomass was best correlated 
with HH- and cross-polarizations at L-band. Conversely, the 
higher frequency bands (Ka, Ku, and X) were poorly correlated 
with LAI and biomass. 
Many experimental studies have linked the bio- and geo 
physical characteristics of crops with backscatter recorded by 
SAR sensors. (Clevers and van Leeuwen, 1996; Ferrazzoli et 
al., 1999; McNaim, 2002; Taconet et al., 1996). Most of these 
studies were carried out using C-band SAR due to the 
availability of this radar frequency on the first generation of 
satellite SAR sensors (ERS-1/2, RADARSAT-1). The SAR 
sensors currently operational include TerraSAR-X (X-band), 
COSMO-SkyMed (X-band), PALSAR/ALOS (L-band), 
ASAR/ENVISAT (C-band), RADARSAT-1/2 (C-band), and 
ERS-2 (C-band). With access to such a wealth of SAR 
satellites, it is now possible to study the sensitivity of multi 
frequency and multi-polarization data to LAI through the entire 
crop growth cycle. Detailed understanding of radar response to 
crops characteristics as a function of SAR parameters 
(wavelength, incidence, and polarization) is the first essential 
step in developing robust methods to retrieve crop bio-physical 
variables such as LAI. 
This study investigates the sensitivity of TerraSAR-X, 
PALSAR/ALOS, and RADARSAT-2 to crop bio-physical 
variables. The objective is to assess the radar response of com 
and soybean crops with respect to radar wavelength (X, C, and 
L-bands) and polarization. Leaf area index (LAI) and surface 
volumetric soil moisture were measured to coincide with 
remote sensing acquisitions. In this paper, correlation analyses 
were conducted between radar backscatter and LAI. In addition, 
a semi-empirical backscatter process model (the water cloud 
model) was used to develop the relationship between SAR 
backscatter and target conditions, including LAI and soil 
moisture.