Meetings: Documents

Active/Passive Remote Sensing of Vegetation: Theoretical and Experimental Results
[11-Apr-2016] Guerrinero, L., Ferrazzoli, P., Vittucci, C., Aurizzi, M., and Vergata, T.
Presented at the 14th Specialist Meeting on Microwave Radiometry and Remote Sensing of the Environment
In the last decades, researches based on passive and active systems were mostly conducted separately. More recently, a joint use of the two techniques were investigated with the objective of exploiting both the good properties of SAR in terms of high resolution and the good sensitivity of radiometric observations to soil moisture. Signatures collected by Aquarius instrument mounted on the SAC-D satellite, although dedicated to ocean applications, are also used to investigate a joint use of active and passive L band observations over land, particularly in large homogeneous areas (Piles et al., TGRS, pp. 4700-4711, 2015). Moreover, L band SAR systems that are in orbit, such as ALOS, or planned, such as SAOCOM, can be exploited jointly with spaceborne L band radiometers. In this framework, several works focus on the estimation of the slope of a linear relation between emissivity and backscattering coefficient, for different land covers and soil/vegetation parameters. In the past years, some works demonstrated that the emissivity e and the backscattering coefficient of land surfaces are affected in different ways by soil and vegetation variables (Ferrazzoli et al., TGRS, pp. 772-778, 1989; Ferrazzoli and Guerriero, Proc. Microrad 2011). In particular, for several vegetation kinds and for an angle higher than about 30°, an increase of soil roughness or vegetation biomass produces an increase of both the emissivity and the backscattering coefficient, while an increase of soil moisture produces opposite effects on the two variables. A physical interpretation of this result is based on the relationships between backscattering coefficient, emissivity, and bistatic scattering coefficient. The present paper exploits this theoretical formulation to simulate backscattering coefficient and emissivity of bare soils, growing corn crops and broadleaf forests under various values of soil moisture. The advantage of this approach is that a unique physical formulation is adopted to compute the backscattering coefficient and the emissivity of the observed medium, through its bistatic scattering coefficient. The theoretical results are compared with experimental data collected by various systems over different kind of surface cover. First, we have included data from the Soil Moisture Active and Passive Experiment (SMAPEx), collected during eight flights in September 2011 over the Gillenbah forests in New South Wales (Australia), in the framework of SMAPEx-3 campaign (www.smapex.monash.edu.au). Then, data from the 1999 Southern Great Plains (SGP99), Soil Moisture Experiment 2002 (SMEX02) and Cloud and Land Surface Interaction Campaign (CLASIC) campaigns over agricultural crops in the United States have been considered (Colliander et al., RSE, pp.309-322, 2012). Finally, we have included data globally collected by Aquarius in a 3-day interval of time, from July 1 to July 3, 2012. Near simultaneous values of soil moisture and vegetation optical depth retrieved by Soil Moisture and Ocean Salinity (SMOS) radiometric mission (Level 3) are used to interpret the Aquarius results. Experimental data from both airborne and satellite acquisitions confirm the theoretical simulations. Finally, the model is also used to predict the different sensitivities of active and passive instruments to soil moisture variations for different land covers, as well as the coefficient of a linear relationship between emissivity and backscattering coefficient.