1D-Var multilayer assimilation of X-band SAR data into a detailed snowpack model - Université de Rennes Accéder directement au contenu
Article Dans Une Revue The Cryosphere Année : 2014

1D-Var multilayer assimilation of X-band SAR data into a detailed snowpack model

Résumé

The structure and physical properties of a snow-pack and their temporal evolution may be simulated using meteorological data and a snow metamorphism model. Such an approach may meet limitations related to potential diver-gences and accumulated errors, to a limited spatial resolu-tion, to wind or topography-induced local modulations of the physical properties of a snow cover, etc. Exogenous data are then required in order to constrain the simulator and im-prove its performance over time. Synthetic-aperture radars (SARs) and, in particular, recent sensors provide reflectivity maps of snow-covered environments with high temporal and spatial resolutions. The radiometric properties of a snowpack measured at sufficiently high carrier frequencies are known to be tightly related to some of its main physical parame-ters, like its depth, snow grain size and density. SAR acqui-sitions may then be used, together with an electromagnetic backscattering model (EBM) able to simulate the reflectiv-ity of a snowpack from a set of physical descriptors, in or-der to constrain a physical snowpack model. In this study, we introduce a variational data assimilation scheme coupling TerraSAR-X radiometric data into the snowpack evolution model Crocus. The physical properties of a snowpack, such as snow density and optical diameter of each layer, are simu-lated by Crocus, fed by the local reanalysis of meteorological data (SAFRAN) at a French Alpine location. These snow-pack properties are used as inputs of an EBM based on dense media radiative transfer (DMRT) theory, which simulates the total backscattering coefficient of a dry snow medium at X and higher frequency bands. After evaluating the sensi-tivity of the EBM to snowpack parameters, a 1D-Var data assimilation scheme is implemented in order to minimize the discrepancies between EBM simulations and observa-tions obtained from TerraSAR-X acquisitions by modifying the physical parameters of the Crocus-simulated snowpack. The algorithm then re-initializes Crocus with the modified snowpack physical parameters, allowing it to continue the simulation of snowpack evolution, with adjustments based on remote sensing information. This method is evaluated us-ing multi-temporal TerraSAR-X images acquired over the specific site of the Argentière glacier (Mont-Blanc massif, French Alps) to constrain the evolution of Crocus. Results indicate that X-band SAR data can be taken into account to modify the evolution of snowpack simulated by Crocus.
Fichier principal
Vignette du fichier
Phan2014_Cryosphere.pdf (2.61 Mo) Télécharger le fichier
Origine : Fichiers éditeurs autorisés sur une archive ouverte
Loading...

Dates et versions

hal-01118432 , version 1 (19-02-2015)

Identifiants

Citer

Xuan Vu Phan, Laurent Ferro-Famil, Michel Gay, Y Durand, M Dumont, et al.. 1D-Var multilayer assimilation of X-band SAR data into a detailed snowpack model. The Cryosphere, 2014, pp.1975-1987. ⟨10.5194/tc-8-1975-2014⟩. ⟨hal-01118432⟩
924 Consultations
233 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More