Three dimensional gravity inversion with a priori and statistical constraints
The need to interpret increasingly complex geologic targets gravity is reemerging as a geology interpretation tool. Geology drives this revisionist approach to the geologic 3-D gravity inversion in two ways; (1) the overall approach formally test for consistency between multiple geologic hypotheses and the gravity; (2) densities and density distributions for the geologic units near the gravity stations are based on the mapped stratigraphic and structural framework at the scale of the best available digital elevation models. Bouguer and terrain corrected gravity data can contain unresolvable and ambiguous "noise". To address these problems, I am developing an a priori and statistically constrained rock density inversion to test for consistency between the gravity and a density distribution. I use 3-D gravity data and interpreted structural geologic models using superimposed regularly gridded surfaces of arbitrary fineness. Each surface used in a geologic model has an associated expected density and uncertainty. ^ To test the inversion I generated four bounding test cases using a partially imbedded 1 km diameter sphere in a 1 km thick layer. First to test the inversion implementation, I use perfect data and a consistent perfect geologic model. Second to test for stability, convergence and resolution I use perfect data and inconsistent perfect geologic model. This is similar to traditional forward gravity modeling. Third to test the inversion’s ability to diagnose inconsistency, I use perfect data and an inconsistent imperfect geologic model. Fourth to test the inversion using simulated real world data I use imperfect data and a consistent imperfect geologic model. After successfully completing the synthetic test I have applied the inversion algorithm to modeled and observed geology and gravity data in the surrounding region around Jackass Flats, Nevada. After successfully building a regional geologic and density interpretation I will build two geologic interpretations for the Jackass Flats basin. This will illustrate building detailed interpretations from the piecewise continuous regional interpretation and the testing of multiple hypothesis. First I built a single body interpretation of Jackass Flats that proved to be inconsistent with the gravity data. Second, I built a two body interpretation of Jackass Flats that proved to be consistent but imperfect. ^
Crain, Kevin Doyle, "Three dimensional gravity inversion with a priori and statistical constraints" (2007). ETD Collection for University of Texas, El Paso. AAI3242138.