MOTIVATION AND OBJECTIVES. RECURRING SLOPE LINEAE (RSL) ARE NARROW FINGER-LIKE FEATURES THAT ARE DARKER THAN THEIR SURROUNDINGS AND HAVE BEEN OBSERVED TO INCREMENTALLY LENGTHEN FADE AND RECUR IN MULTIPLE MARS YEARS. RSL CANDIDATES HAVE BEEN IDENTIFIED AT HUNDREDS OF SITES GLOBALLY. OF THESE SITES 44 EXHIBIT CONFIRMED ACTIVITY (MULTIPLE IMAGES HAVE SHOWN INCREMENTAL LENGTHENING FADING AND RECURRENCE) AND HAVE BEEN QUALITATIVELY RATED AS EXCELLENT (INDICATING LARGE EASILY MEASURED RSL; LESS THAN 15% OF RSL SITES OBTAIN THIS RANKING). TO DATE THE FORMATION AND RESETTING MECHANISMS OF RSL ARE POORLY UNDERSTOOD. WET-BASED HYPOTHESES BEST MATCH OBSERVATIONS THAT CORRELATE INCREMENTAL LENGTHENING WITH HIGHER SURFACE TEMPERATURES BUT RECENT TOPOGRAPHIC MEASUREMENTS SUGGEST A DRY GRANULAR FLOW PROCESSES. .THIS PROPOSAL FOCUSES ON USING HIGH-RESOLUTION MAPPING OF THE BEST-MONITORED RSL SITES COMBINED WITH HIGH-RESOLUTION (30 CM/PX) DIGITAL TERRAIN MAPS (DTM). THUS THE OBJECTIVE OF THE PROPOSED WORK IS TO MIX GLOBAL-SCALE OBSERVATION OF RSL GEOGRAPHY AND LENGTHENING SEASONALITY WITH HIGH-RESOLUTION MAPPING OF 10 OF THE BEST RSL SITES TO ULTIMATELY BETTER DETERMINE IF RSL ARE FORMED VIA A WET OR DRY PROCESS. .METHODOLOGY .TO ACCOMPLISH OUR OBJECTIVE WE WILL FIRST UPDATE THE GLOBAL INVENTORY OF RSL OBSERVATIONS BY EXAMINING RECENT HIGH RESOLUTION IMAGING SCIENCE EXPERIMENT (HIRISE) IMAGES SPANNING KNOWN SITES (116 IMAGES) AND POSSIBLE NEW SITES (563 IMAGES). NEXT WE WILL USE A CUSTOM APPROACH TO GENERATE ~233 ORTHORECTIFIED IMAGES AND ~10 HIGH-RESOLUTION DTMS (~30 CM/PIXEL) FOR EXCELLENT AND CONFIRMED RSL SITES. UTILIZING UP TO NINE STEREO PAIRS AT EACH SITE OUR NOVEL METHOD CAN QUANTIFY THE ERROR IN THE DTM. WITH THESE HIGH-RESOLUTION DTMS WE WILL USE OUR PREVIOUSLY DEVELOPED MAPPING AND AUTOMATED ANALYSIS OF RSL (MAARSL) COMBINED WITH A NEW MACHINE LEARNING ALGORITHM TRAINED ON OUR PREVIOUSLY ANNOTATED RSL DETECTIONS TO AUTOMATICALLY IDENTIFY AND CHARACTERIZE RSL. WE WILL EMPLOY A POST-PROCESSING MANUAL REFINEMENT STEP TO ENSURE HIGH QUALITY OF THE RSL DETECTIONS. GEOSTATISTICAL PROPERTIES SUCH AS AREA RELATIVE ALBEDO AND SLOPES OF RSL WILL THEN BE COMPUTED GEOSTATISTICS FOR THE RESULTING RSL DATABASE. THESE RSL ATTRIBUTES AND THEIR SEASONALITY WILL BE COMPARED TO WET AND DRY RSL FORMATION MECHANISMS. LASTLY THE DTMS SLOPES MAPS ORTHOIMAGES AND POLYGONS OF THE MAPPED RSL WILL BE UPLOADED TO THE PDS. .RELEVANCE .OUR PROPOSAL IS DIRECTLY RELEVANT TO THE MDAP PROGRAM AS IT PRIMARILY ANALYZES IMAGES FROM HIRISE ONBOARD MRO (MARS RECONNAISSANCE ORBITER) TO ENHANCE THE SCIENTIFIC RETURN OF THIS MISSION BY STUDYING HOW RSL ARE FORMED. .