Modeling atmospheric precipitation impact on synthetic aperture radar imagery at X and Ka bands

摘要

Spacebome synthetic aperture radars (SARs) operating at X-band and above allow observations of Earth surface at very high spatial resolution. Moreover, recent polarimetric SARs enable the complete characterization of target scattering and extinction properties. Nowadays several spaceborne X-band SAR systems are operative, and plans exist for systems operating at higher frequency bands (i.e. Ku, Ka and W). Although higher frequencies may have interesting and distinctive applications, atmospheric effects, especially in precipitating conditions, may affect the surface SAR response in both the signal amplitude and its phase, as assessed by numerous works in the last years. A valid tool to analyze and characterize the SAR response in these conditions is represented by forward modeling, where a known synthetic scenario, which is described by user-selected surface and atmospheric conditions, is considered. Thus, the SAR echoes corresponding to the synthetic scenarios are simulated using electromagnetic models. In this work a 3-D realistic polarimetric SAR response numerical simulator is presented. The proposed model framework accounts for the SAR slant observing geometry and it is able to characterize the polarimetric response both in amplitude and phase. In this work we have considered both X and Ka bands, thus exploring the atmospheric effects for the present and future polarimetric systems. The atmospheric conditions are simulated using the System for Atmospheric Modeling (SAM) which is an high-resolution mesoscale model. SAM is used to define the three-dimensional distribution of hydrometeors which are among the inputs used in the Hydrometeor Ensemble Scattering Simulator (HESS) T-Matrix which allow simulating the SAR signal due to the atmospheric component. The SAR surface component is, instead, simulated by a Semi Empirical Model (SEM) for bare-soils conditions and SEAWIND2 two-scale model for ocean surfaces. The proposed methodology has been applied in this work to assess the sensitivity of the considered frequency bands to different hydrometeor spatial distributions above some examples surface backgrounds.