ANALYSIS OF SEISMIC WAVES REMAINS THE MOST DIRECT WAY TO EXPLORE AND CONSTRAIN PLANETARY INTERIORS. THE STRUCTURE THEY SAMPLE WHILE TRAVELING THROUGH THE PLANET CAN BE INFERRED FROM MEASUREMENTS OF GROUND MOTION RECORDED BY INSTRUMENTS SUCH AS INSIGHT'S SEISMOMETER SEIS. THE KEY OBJECTIVE OF THIS PROPOSAL IS TO CONSTRAIN THE MID AND DEEP MANTLE OF MARS USING SEISMIC DATA WITH SENSITIVITY TO STRUCTURE THAT COMPLEMENTS THE ANALYSES ALREADY PLANNED BY THE INSIGHT SCIENCE TEAM. THEY HAVE PLANNED TO MEASURE FUNDAMENTAL MODE SURFACE WAVE DISPERSION TO GET PRELIMINARY MODELS OF THE SHALLOW MANTLE STRUCTURE AND NORMAL MODES TO ESTIMATE MARS GLOBAL AVERAGE STRUCTURE DOWN TO THE CORE. THESE MODELS WILL BE IMPROVED BY INCLUDING IMPACTGENERATED BODY WAVES. THOSE HAVE HOWEVER POOR VERTICAL RESOLUTION. SURFACE WAVE DISPERSION HAS MUCH GREATER DEPTH SENSITIVITY TO SHEAR-WAVE VELOCITY STRUCTURE BUT FUNDAMENTAL MODE SURFACE WAVES ONLY SAMPLE THE SHALLOW PART OF THE MANTLE. ON EARTH THEY ARE TYPICALLY MEASURED AT PERIODS OF 50-200 S WHICH CAN RESOLVE STRUCTURE DOWN TO ABOUT 200 KM DEPTH. SIMILAR DEPTH RESOLUTION IS EXPECTED ON MARS FROM THE INVERSION OF FUNDAMENTAL MODE RAYLEIGH WAVES (PANNING ET AL. 2015 2016). HIGHER MODE SURFACE WAVES HAVE SENSITIVITY TO DEEPER STRUCTURE THAN FUNDAMENTAL MODES ARE THUS A POWERFUL TOOL TO CONSTRAIN THE MANTLE AT LARGE DEPTHS (BEGHEIN&TRAMPERT 2004; YUAN AND BEGHEIN 2013). HIGHER MODE DISPERSION MEASUREMENTS THEREFORE PROVIDE UNIQUE CONSTRAINTS TO MID MANTLE STRUCTURE COMPLEMENTING OBSERVABLES FROM FUNDAMENTAL MODE SURFACE WAVES AND NORMAL MODE CENTRAL FREQUENCIES AND DO NOT REQUIRE SEISMIC EVENTS AS LARGE AS THOSE NEEDED TO EXCITE FREE OSCILLATIONS (ON EARTH ONE CAN DETECT HIGHER MODES WITH EVENTS OF MAGNITUDE 6 WHILE MAGNITUDE 7 OR LARGER IS PREFERABLE TO MEASURE NORMAL MODES). HIGHER MODES SURFACE WAVES ARE DIFFICULT TO MEASURE AND REQUIRE SOPHISTICATED TECHNIQUES. WE RECENTLY DEVELOPED A BAYESIAN NONLINEAR WAVEFORM MODELING METHOD BASED ON A MARKOV CHAIN MONTE CARLO (MCMC) APPROACH TO MEASURE THEIR DISPERSION WITH QUANTITATIVE UNCERTAINTIES AND RELIABILITY ASSESSMENT (XU&BEGHEIN 2017). SUCH FORWARD MODELING METHODS ARE NOT ONLY IMPORTANT TO RETRIEVE STRUCTURE AT DEPTH WHEN INVERTING A LIMITED DATASET BUT THEY ARE ALSO POWERFUL TO MEASURE AND RETRIEVE SIGNIFICANT INFORMATION FROM NOISY SEISMIC DATA OR FROM OVERLAPPING MODES AS PROPOSED HERE. THIS APPLICATION OF THIS METHOD TO THE INSIGHT SEISMIC DATA WILL RESULT IN HIGHER MODE DISPERSION CURVES WITH QUANTITATIVE UNCERTAINTIES THAT WILL BE INVERTED TO CONSTRAIN STRUCTURE AT DEPTH. THE PROPOSED WORK WILL BE DIVIDED INTO TWO MAIN TASKS: TASK 1: BAYESIAN DISPERSION CURVE MEASUREMENTS. WE WILL PERFORM NON-LINEAR MODELING OF FILTERED SEISMOGRAMS WITH A MCMC BAYESIAN TECHNIQUE TO EXTRACT DISPERSION CURVES OF FUNDAMENTAL AND HIGHER MODE RAYLEIGH WAVES USING EVENTS LOCATED BY OTHER MEMBERS OF THE INSIGHT TEAM FOLLOWING PANNING ET AL. (2016). THIS TASK WILL BE REPEATED WHEN NEW SEISMIC DATA BECOME AVAILABLE. TASK 2: NON-LINEAR MODELING OF MANTLE STRUCTURE. WE WILL JOIN FORCES WITH THE REST OF THE TEAM TO MODEL MARS MANTLE STRUCTURE WITH A NON-LINEAR MCMC METHOD (WEIDNER&BEGHEIN 2017). THIS WILL BE DONE BY COMBINING OUR MEASURED DISPERSION CURVES WITH CRUSTAL CONSTRAINTS OBTAINED BY THE INSIGHT TEAM FROM MEASUREMENTS OF RECEIVER FUNCTION BROADBAND RAYLEIGH WAVE ELLIPTICITY AND NORMAL MODE CENTRAL FREQUENCIES IF AVAILABLE. THIS WORK WILL CONTRIBUTE TO RESEARCH IN INSIGHT'S DEEP INTERIOR THEME GROUP. IT WILL HELP DETERMINE THE THICKNESS AND STRUCTURE OF THE MANTLE WHICH IN TURN WILL AFFECT OUR KNOWLEDGE OF THE THERMAL STATE OF THE INTERIOR. IT WILL HELP UNCOVER THE FORMATION AND EVOLUTION OF MARS AND OTHER TERRESTRIAL PLANETS WHICH IS ONE OF INSIGHT'S PRIMARY SCIENCE GOALS. HIGHER MODE SURFACE WAVE ANALYSIS WILL BE CONDUCTED AS PART OF MARS STRUCTURE SERVICE.