NUMERICAL ALGORITHMS FOR COOPERATIVE IMAGING OF COMPLEX GEOLOGICAL REGIONS

IN ORDER TO LESSEN THE EFFECT OF ILL-CONDITIONING AND NON-UNIQUENESS THAT AFFECTS THE SOLUTION OF MANY INVERSE PROBLEMS IN GEOPHYSICS, WEIDLINGER ASSOCIATES PROPOSES TO EXPLORE AND DEVELOP NUMERICAL METHODS THAT WILL TAKE INTO ACCOUNT MULTIPLE INFERENTIAL SOURCES FOR IMAGING A COMPLEX GEOLOGICAL REGION. THEY HAVE ALREADY DEVELOPED SOME ADVANCED SEISMIC METHODS FOR INVERSION, AND PROPOSE TO DEVELOP COMPARABLE CAPABILITIES IN ONE OR MORE OF THE OTHER APPROACHES COMMONLY EMPLOYED IN PROSPECTING: GRAVIMETRIC, MAGNETIC ELECTRIC, AND WELL LOGGING. THE IDEA IS TO COMBINESEVERAL OF THESE TECHNIQUES TO APPLY THEM CONCURRENTLY TO THE SAME EARTH VOLUME, IN ORDER TO INCREASE THEIR INDIVIDUALPOWERS OF RESOLUTION. OF COURSE, THIS CAN BE DONE PROFITABLY EVEN WITH SEISMIC TECHNIQUES ALONE, BY COMBINING THE RESULTS OF REFLECTION AND REFRACTION SURVEYS WITH VARIOUS KINDS OF VERTICAL SEISMIC PROFILING AND WELL-LOG DATA, AND THEY WILL BE PURSUING THOSE ASPECTS AS WELL, IN THIS AND IN THEIR CURRENT INVERSION PROJECT. THE MAIN INGREDIENT THAT THEY WILL LOOK FOR IN THE DIFFERENT MODELING TECHNIQUES IS THAT THEY SHARE SOME COMMON PARAMETERS, I.E., THAT THEY ARE COUPLED IN PARAMETER SPACE, IN ORDER TO EFFECTIVELY INFLUENCE EACH OTHER. THE FACT THATOPTIMIZATION TECHNIQUES WILL BE USED TO FIT THE DIFFERENT DATA SETS WITH THEIR RESPECTIVE MATHEMATICAL MODELS, NATURALLY BRINGS OUT THE ISSUE OF MULTI-OBJECTIVE OPTIMIZATION; ANOTHER OF THEIR PROPOSED TASKS WILL BE TO LOOK INTO CURRENT THEORY AND PRACTICE IN THAT FIELD IN ORDERTO IDENTIFY USEFUL TECHNIQUES THAT APPLY TO THIS PROBLEM.