Multi-layer model simulations of backscattering and emission from snow covered soils S. Paloscia 1* , P. Pampaloni 1 , E. Santi 1 , S.Pettinato 1 , X.Chuan 2 , M.Brogioni 1 1 IFAC-CNR, Florence (Italy), S.Paloscia@ifac.cnr.it 2 RADI-CAS, Beijing, (China) Abstract In this paper an overview of experimental and theoretical investigations concerning snow cover by using both active and passive microwave sensors will be presented. In particular, the effects of multi-layer structure of snowpack and its temporal evolution on backscattering and emission have been investigated by model simulations based on the Dense Medium Radiative Transfer (DMRT). The implemented model has been validated using both X- band Cosmo SkyMed acquisitions and multi- frequency microwave radiometric data collected in the winters between 2007 and 2011 on a test area located in the Eastern part of the Italian Alps, and corresponding direct measurements of the main snow parameters. The effect of snow parameters on microwave backscattering at X- band and emission at X-, Ku- and Ka-bands has been analyzed by comparing experimental data and model simulations performed with a multi-layer version of the DMRT- QCA model. The study has been focused on the effect of layering structure of snowpack typical of the Alpine regions. It has been observed that the multi-layer model is generally able to account for the effects of complex stratigraphy, reproducing the measured backscattering and emission with a higher accuracy than the single layer one. 1. Introduction Microwave sensors are sensitive to snow properties and are able to give information on snow depth (SD) and snow water equivalent (SWE) as demonstrated by the pioneering works of the teams of Univ. of Berne and Univ. of Kansas [1, 2]. The key frequency channels in detecting the presence of snow were found to be at Ku- and Ka- bands, and most algorithms to retrieve snow parameters by using passive sensors are based on the difference between the brightness temperature (Tb) at these bands (i.e. 19 and 37 GHz). These frequencies are not available from the present in orbit Synthetic Aperture Radars (SAR), which operate at C- or X- band where wave penetration is quite high in dry snow and very low in wet snow. Recent investigations showed a noticeable sensitivity of X-band backscattering and SWE when SD is higher than 50-60 cm [3]. The case of emission from layered snow has been theoretically studied by Liang et al. [4]. In this paper, the effects of multilayer structure of snowpack on emission and backscattering have been investigated by using a physical model and experimental data collected on the Italian Alps. The model simulates snowpack by means of a stratified medium composed by spherical ice scatterers embedded in air, overlying a rough homogeneous half-infinite medium. Volume scattering/emission from snow is computed by using a multi-layer version of the Dense Medium Radiative Transfer Theory in the Quasi Crystalline Approximation (DMRT-QCA) with sticky particles [5]. Soil contribution is simulated by using the AIEM for the co-polar components and the semi-empirical approach by Oh for the cross-polar terms. Radiometric data were collected from C to Ka bands with ground based instruments in winter seasons from 2007 to 2009 on a site selected in NE Italy. More recently, X-band SAR images of a wide area surrounding the radiometer station were obtained from the Cosmo Skymed mission. Also snow profiles were obtained with conventional methods at convenient temporal and spatial sampling. In addition to the measured profiles, and in order to make this study as general as possible, a few “synthetic” snow profiles, representative of typical cases of dry snow cover in the Alps were realized. The parameters taken into consideration were: grain size, snow density and layer thickness. These “synthetic” profiles, that summarize the main characteristics for each period of the season, were obtained by analyzing 26 real snow profiles obtained in winter 2010-2011, when the snow depth was frequently greater than 60 cm. The performed sensitivity analysis carried out considering all the major parameters that affect emission and backscattering confirmed the high sensitivity of the model to grain size and pointed out the contribution of the layering structure of snowpack in particular to the polarized emission. 2. Experimental data IFAC ground based radiometric system at C-, X-, Ku-, and Ka-band, in V and H polarizations, was installed in a shelter on a test area on the Cherz Plateau. This site is located in the Dolomites mountains, inside the Cordevole watershed (a few Km 2 extension). In the area, snow is usually present from mid-November to mid-April, and the 978-1-4673-5225-3/14/$31.00 ©2014 IEEE