Dissertation Abstract
Snow microwave emission modeling: applications in Antarctica and Quebec
Brucker, Ludovic 2009
Environmental Earth Science System, University of Grenoble, France (France), 308 pp.
The cryosphere has a key role in the climate system, mostly because it reflects a significant fraction of solar energy reaching the Earth's surface and contains a large amount of freshwater in solid form. Due to the sensitivity of the cryosphere to climatic variables such as temperature and precipitation, the various elements of the cryosphere, like the Antarctic ice cap and the seasonal snow over the subarctic regions, are indicators of global climate change. However, our knowledge of the polar regions are limited by a lack of in situ observations reflecting the remoteness of these regions and the harsh weather conditions. The analysis of these sparse observations can be enhanced with remote sensing to reduce uncertainties on climate trends observed at high latitudes.
Passive microwave remote sensing is suited to interpret and monitor the snow physical properties. Indeed, the microwave radiation emanates from the ground or snow itself, and then propagates to the surface. Thus, the emerging radiation contains information on vertical variations of snow properties, such as temperature and microstructure properties (grain size and density). These three properties control the snow microwave emission of dry snowpacks. For wet snowpacks, liquid water content controls the emission. The temporal evolution and vertical variations of these different properties are defined by snow metamorphism. Their link with the microwave emission is considered by the radiative transfer.
This thesis aims at explaining the snow microwave emission by forward modeling to understand the evolution of the main snow physical properties. The radiative transfer in snow has been calculated with multilayer microwave emission models (MEMLS, DMRT-ML), based either on semi-empirical or theoretical approaches. The stratigraphic profiles of snow used as input were measured, randomly generated, modeled with a simple relationship of metamorphism or with the thermodynamic snow evolution model Crocus.
These models and approaches have been applied on two types of snow, permanent (in Antarctica) and seasonal (in Quebec). In the first case, the temporal evolution of microwave brightness temperatures has been modeled at Dome C from in situ measurements of snow properties. Using this approach, the emissivity is modeled from measurements and is therefore applicable locally. To model the emissivity through out Antarctica, the vertical variation of grain size appeared crucial to predict the emissivity in vertical polarization. This sensitivity was used to estimate, at the continental scale, the grain size profile, an important glaciological variable. The density profile and surface properties determine the difference between vertical and horizontal polarizations.
The microwave emission of a seasonal snowpack in Quebec has also been discussed. The specificity of the study is to predict the timeseries of brightness temperatures with a coupled snow evolution-emission model, here Crocus-MEMLS. This approach allows us to finely interpret the temporal evolution of the brightness temperatures measured with a surface based radiometer. In addition, this approach allows us to identify some physical relationships of the Crocus model needing refinements. The results highlight the complexity of the microwave signal for snowpacks evolving rapidly at temperatures near the melting point.
Passive microwave remote sensing is suited to interpret and monitor the snow physical properties. Indeed, the microwave radiation emanates from the ground or snow itself, and then propagates to the surface. Thus, the emerging radiation contains information on vertical variations of snow properties, such as temperature and microstructure properties (grain size and density). These three properties control the snow microwave emission of dry snowpacks. For wet snowpacks, liquid water content controls the emission. The temporal evolution and vertical variations of these different properties are defined by snow metamorphism. Their link with the microwave emission is considered by the radiative transfer.
This thesis aims at explaining the snow microwave emission by forward modeling to understand the evolution of the main snow physical properties. The radiative transfer in snow has been calculated with multilayer microwave emission models (MEMLS, DMRT-ML), based either on semi-empirical or theoretical approaches. The stratigraphic profiles of snow used as input were measured, randomly generated, modeled with a simple relationship of metamorphism or with the thermodynamic snow evolution model Crocus.
These models and approaches have been applied on two types of snow, permanent (in Antarctica) and seasonal (in Quebec). In the first case, the temporal evolution of microwave brightness temperatures has been modeled at Dome C from in situ measurements of snow properties. Using this approach, the emissivity is modeled from measurements and is therefore applicable locally. To model the emissivity through out Antarctica, the vertical variation of grain size appeared crucial to predict the emissivity in vertical polarization. This sensitivity was used to estimate, at the continental scale, the grain size profile, an important glaciological variable. The density profile and surface properties determine the difference between vertical and horizontal polarizations.
The microwave emission of a seasonal snowpack in Quebec has also been discussed. The specificity of the study is to predict the timeseries of brightness temperatures with a coupled snow evolution-emission model, here Crocus-MEMLS. This approach allows us to finely interpret the temporal evolution of the brightness temperatures measured with a surface based radiometer. In addition, this approach allows us to identify some physical relationships of the Crocus model needing refinements. The results highlight the complexity of the microwave signal for snowpacks evolving rapidly at temperatures near the melting point.