IN CONTRAST TO RAPID ARCTIC SEA ICE LOSS ANTARCTIC SEA ICE EXTENT (SIE) HAS SHOWN A SMALL BUT STATISTICALLY SIGNIFICANT INCREASE SINCE THE LATE 1970S. THE EXPANSION OF ANTARCTIC SIE IN A WARMING CLIMATE HAS PUZZLED THE SCIENCE COMMUNITY. SIMULATION OF THE OBSERVED ANTARCTIC SEA ICE CHANGE REMAINS A MAJOR CLIMATE MODEL DEFICIENCY. WE PROPOSE TO ADVANCE THE UNDERSTANDING OF ANTARCTIC SEA ICE CHANGE AND TO IMPROVE THE REPRESENTATION OF ANTARCTIC SEA ICE IN THE NASA GODDARD EARTH OBSERVING SYSTEM MODEL VERSION 5 (GEOS-5) USING COUPLED REPLAY SIMULATIONS AND OBSERVATIONS THE MAJOR DIFFERENCE BETWEEN THE PROPOSED WORK AND PREVIOUS MODELING STUDIES IS THAT WE WILL USE THE GEOS-5 REPLAY CAPABILITY. BY CONSTRAIN THE ATMOSPHERE TO NASA MODERN-ERA RETROSPECTIVE ANALYSIS FOR RESEARCH AND APPLICATION VERSION 2 (MERRA-2) THE COUPLED REPLAY SIMULATIONS WILL ENABLE US TO SEPARATE THE INFLUENCES OF ATMOSPHERIC AND OCEANIC PROCESSES ON ANTARCTIC SEA ICE MULTIDECADAL VARIABILITY. WE WILL QUANTIFY THE IMPACT OF BIASES IN ATMOSPHERIC SURFACE FORCING ON ANTARCTIC SEA ICE TRENDS IN GEOS-5 BY COMPARING FREE-RUNNING AND REPLAY SIMULATIONS. WE WILL ISOLATE THE EFFECTS OF OCEAN EDDY REPRESENTATION AND SURFACE FRESHWATER FORCING ON ANTARCTIC SEA ICE CHANGE AND TO QUANTIFY THE ATMOSPHERIC MODULATION ON THESE EFFECTS. WE WILL QUANTIFY AND IDENTIFY THE MECHANISMS THAT CONTROL MODEL S INTERNAL MULTIDECADAL VARIABILITY OF ANTARCTIC SEA ICE. WE WILL CONDUCT SENSITIVITY SIMULATIONS TO IDENTIFY THE FEEDBACK OF ANTARCTIC SEA ICE CHANGE ON THE ATMOSPHERE. THE PROPOSED WORK WILL REDUCE UNCERTAINTIES IN ANTARCTIC SEA ICE SIMULATION AND GUIDE THE IMPROVEMENTS OF SOUTHERN HEMISPHERE ATMOSPHERE AND SOUTHERN OCEAN PROCESSES IN GEOS-5 AND THUS DIRECTLY RELATED TO THE MAP PROGRAMMATIC PRIORITY. OUR RESEARCH WILL ADDRESS THE MAP RESEARCH THEME OF COUPLING IN THE EARTH SYSTEM BY IMPROVING THE UNDERSTANDING OF THE COUPLINGS BETWEEN ANTARCTIC ATMOSPHERE SOUTHERN OCEAN AND SEA ICE.