THE VAPOR AND PARTICULATE PLUMES ARISING FROM ENCELADUS AND MEASURED BY MULTIPLE CASSINI INSTRUMENTS ARE A KEY SIGNATURE OF WHAT LIES BELOW THE SURFACE. THE MOST LIKELY SOURCE FOR THESE EXTENSIVE GEYSERS IS A SUBSURFACE LIQUID RESERVOIR OF SOMEWHAT SALINE WATER AND OTHER VOLATILES BOILING OFF THROUGH CREVASSE-LIKE CONDUITS INTO THE VACUUM OF SPACE. AS WITH MANY REMOTE PLANETARY OBSERVATIONS THE OBSERVATIONAL DATA ARE SPARSE AND ONE MUST INFER UNSEEN CONDITIONS AND MECHANISMS. THESE ARE DIFFICULT TO ESTABLISH BY PHYSICAL MODELING DUE TO THE COMPLEXITY INHERENT IN THE GEOMETRIES AND FLOWS AND THE DIVERSE TYPES OF CONSTRAINING OBSERVATIONS. THERE ARE ALSO A LARGE NUMBER OF PARAMETERS IN THE PHYSICAL MODELS AND THESE MUST BE COMPATIBLE WITH THE FINITE NUMBER OF OBSERVATIONAL CONSTRAINTS. A CRITICAL NEED THEREFORE EXISTS TO EVALUATE THE REAL-WORLD RELIABILITY OF INFORMATION (PARAMETER VALUES WITHIN PHYSICAL MODELS AND THEIR UNCERTAINTIES) OBTAINED FROM FITTING THE UNDER-CONSTRAINED SIMULATION PARAMETERS TO EXISTING SPACECRAFT OBSERVATIONS. OUR OBJECTIVES IN THE CURRENT PROPOSAL ARE TO OPTIMIZE THE DETERMINATION OF THE PROPERTIES OF THE NEAR-SURFACE PLUME SOURCE USING APROMISING NEW APPROACH WITH A UNIQUE SOFTWARE PACKAGE DEVELOPED AT UT THAT EMPLOYS BAYESIAN INFERENCE AND MARKOV CHAIN MONTE CARLO (MCMC) MODELING TO DETERMINE THE PROBABILITY DISTRIBUTION OF POSSIBLE VALUES OF THE MOST IMPORTANT VENT BOUNDARY CONDITIONS TO THE OBSERVED PLUME. WE SEEK TO DETERMINE NOT JUST BEST FIT VALUES OF PARAMETERS OF INTEREST BUT THE PROBABILITY DISTRIBUTION OF THEIR VALUES. MOREOVER EVERY PHYSICAL/MATHEMATICAL MODEL HOWEVER SOPHISTICATED HAS FLAWS OR SHORTCOMINGS. HENCE WE ALSO SEEK TO UNDERSTAND AND QUANTIFY MODEL PLAUSIBILITY I.E. HOW WELL THE MODEL EXPLAINS THE OBSERVATIONS.GOAL 1: TO IMPROVE THE PHYSICAL UNDERSTANDING OF THE DYNAMICS OF THE ENCELADUS PLUMES. IN PARTICULAR DETERMINE HOW THE OBSERVED GRAINS ATOMS MOLECULES AND PHOTONS ARE RELATED TO THE FLOWFIELD AND DRIVING SOURCE PROPERTIES. THIS GOAL WILL BE ATTAINED USING MODELS AND CODES WE ALREADY HAVE RUNNING WITH SMALL ADJUSTMENTS OF THE MODELS AND ASSUMED PARAMETERS.GOAL 2: TO QUANTIFY THE SENSITIVITY OF THE COMPUTED VALUES OF OBSERVED QUANTITIES TO THE VALUES OF PHYSICAL PARAMETERS INPUT TO THE MODELS IN ORDER TO DETERMINE TO WHICH PARAMETERS THE SIMULATED OBSERVATIONS ARE MOST SENSITIVE THUS INDICATING WHICH PARAMETERS MAY BE BEST CONSTRAINED BY POST-CALIBRATIONS BASED ON THE AVAILABLE OBSERVATIONS.GOAL 3: TO DETERMINE THE PROBABILITY DISTRIBUTION OF EACH OF THE MORE IMPORTANT PARAMETERS GIVEN THE UNCERTAINTY INHERENT IN THE MODELING AND THE OBSERVATIONS AND THEREBY OPTIMIZE OUR INFERENCES ABOUT THE INTERIOR. THIS IS A NOVEL APPROACH TO INTEGRATING A BROAD VARIETY OF PLANETARY OBSERVATIONS THROUGH SOPHISTICATED SIMULATIONS TO PROVIDE PROBABILITY DISTRIBUTIONS OF CRUCIAL PARAMETERS.INHERENT IN THIS APPROACH IS AN ASSESSMENT OF THE PLAUSIBILITY OF THE MODELS I.E. A QUANTITATIVE EVALUATION OF THE NECESSARY DEGREE OF COMPLEXITY IN THE MODELING. THIS MEASURE OF PLAUSIBILITY OR MODEL INADEQUACY WILL POINT THE WAY TO IMPROVEMENT OF THE MODELS. OUR PROPOSED WORK WOULD BE THE FIRST TIME A BAYESIAN STATISTICAL APPROACH IS USED TO ANALYZE DIVERSE DATA FROM REMOTE PLANETARY ATMOSPHERIC OBSERVATIONS AND WOULD ADVANCE THE STATE OF THE ART BY EXTRACTING ENOUGH INFORMATION FROM THE CASSINI ENCELADUSOBSERVATIONS TO DISTINGUISH WHICH PARAMETERS ARE TIGHTLY FITTED AND THUS BEST CONSTRAIN SUBSURFACE CONDITIONS. THIS WOULD MATERIALLY ENHANCE THE SCIENCE FROM THE CASSINI MISSION.