Antarctica represents the largest ice reservoir on the Earth. In the context of climate change, this amount of may play a major role for the evolution of sea level. The knowledge of current variations of the Antarctic surface mass balance (SMB) is thus a global challenge. However, the integrated value of SMB still lacks from solid constraints for Antarctica and the fifth IPCC assessment report (AR5) highlighted this uncertainty as one of the main scientific challenges in climate science. The ASUMA project proposes to assess the integrated SMB value over Antarctica, by filling the gap that exists in the coast to central plateau transition zone, where large variations of SMB are observed within small distances. In the plateau, satellite data have been calibrated for interpretation in SMB. However, snow metamorphism occurring in the transition region induces distortions in surface characteristics preventing interpretation of satellite data in terms of SMB. In this project, we propose an alternative method for quantifying the recent variations of SMB and their relationship with climate, atmospheric circulation and moisture origin. We will perform original field measurements of SMB and snow physics and robustly link them to satellite data. We will combine this information with the use of back-trajectories and regional to global modeling. Three field trips are planned during successive austral summers. Two field trips are planned in the first 50 km from the coast to study melting areas, and a long distance traverse is proposed for the 2016-17 summer, to collect 8 firn cores (25-100m each) and samples in about 10 snow pits. To retrieve robust estimate of the mean level and variability of SMB during last decades, we will accurately date firn cores from radiochemistry analyses, especially in the region where melting and erosion are identified. Interpolation of SMB data will be done with ground penetrating radar and satellite data. For this task, understanding how to assess the SMB at a given point with the help of satellite will be crucial. Thus, we will 1) perform an accurate analysis of surface snow physics in the study area, 2) define the physical relationships between the surface physical characteristics and the remote sensing signal, 3) validate SMB outputs form regional atmospheric circulation models. This will allow to answer to four overarching questions: 1. How does the SMB vary in the coast-to-plateau transition zone? 2. What are the processes responsible for these differences and how does the origin and transport of moisture affect the signal stored in firn and ice? 3. How are snow physical characteristics related to the local SMB? 4. How can remote sensing data be best used to infer snow physical characteristics and SMB? 5. What are recent and future SMB distributions over Antarctica? The ASUMA project proposes to combine numerous field data and original techniques to better constrain remote sensing data and modeling results in the transition zone from the coast to the Antarctic plateau. Innovative methodologies will be deployed for grains size measurements, for continuous roughness estimates and for reflectance measurements. The combined information of snow physics, chemistry and isotopic content will be used to constrain output of general circulation models that will in turn be used for estimating future Antarctic SMB. The project will also train two PhD students which will be co-supervised within three project partners providing complementary expertise, and reinforcing strong collaborations between three laboratories. The ASUMA project represents the glaciological part of a long term target of the French climate and cryosphere community, which aims to reduce uncertainties of the mass balance estimates of major ice caps, in order to better understand recent and future changes in sea level rise and Antarctic climate and water cycle.