It has been generally acknowledged that practically all mass-motions within and on the Earth generate asignature in the gravitational field. Current satellite gravity missions, e.g. GRACE, have improved ourunderstanding of the interactions and dynamics of the various components of the Earth's fluids. Future missionssuch as GOCE promise to enhance our understanding of the dynamics of these fluids even more. Nonethelessthe task of unambiguously unravelling the different sources of mass distribution and transport from theaccumulated gravity signal as it is observed remains nontrivial. In order to optimally benefit from the high-resolution (both in time and space) and high-accuracy gravity data coming from possible future gravity missions,a study has been initiated by ESA to investigate the potential for improving our ability to separate the variouscontributions. In this study, we will investigate the need for improved geophysical modelling, mission design andscenario, and methodology required to 1) to de-alias noise from real signal as well as to 2) separate thecontributions from the various sources (e.g. to separate ocean mass variability from atmospheric massvariability). This study is being performed by a consortium of nine European groups, combining expertise in allrelevant geophysical fields and in designing satellite gravity missions. The study started early 2007 and will lastuntil the end of 2008. This paper presents the background of the study, and explains the (simulation) approachthat will be taken to address the goals of the study. Major issues that play a role for the latter are the missionconcepts, the state-of-the-art geophysical modelling, the use of dedicated spatio-temporal sampling to tackle theseparability issue and the use of complementary data and models.