OPTIMIZATION TECHNIQUES FOR INTEGRATING SPATIAL DATA

Two optimization techniques to predict a spatial variable from any num ber of related spatial variables are presented The applicability of th e two different methods for petroleum-resource assessment is tested in a mature oil province of the Midcontinent (USA). The information on p etroleum productivity, usually not directly accessible, is related ind irectly to geological, geophysical, petrographical, and other observab le data. This paper presents two approaches based on construction of a multivariate spatial model from the available data to determine a rel ationship for prediction. In the first approach, tile variables are co mbined into a spatial model by an algebraic map-comparison/integration technique. Optimal weights for the map comparison function are determ ined by the Nelder-Mead downhill simpler algorithm CI multidimensions. Geologic knowledge is necessary to provide a first guess of weights t o start the automatization, because the solution is nor unique. In the second approach, active set optimization for linear prediction of the target under positivity constraints is applied Here, the procedure se ems to select one variable from each data type (structure, isopachous, and petrophysical) eliminating data redundancy. Automating the determ ination of optimum combinations of different variables by applying opt imization techniques is a valuable extension of the algebraic map-comp arison/integration approach to analyzing spatial data. Because of the capability of handling multivariate data sets and partial retention of geographical information, the approaches con be useful in mineral-res ource exploration.