Quantitative characterization and modeling of lithologic heterogeneity

The fundamental goal of this thesis is to gain a better understanding of the vertical and lateral stratigraphic heterogeneities in sedimentary deposits. Two approaches are taken: Statistical characterization of lithologic variation recorded by geophysical data such as reflection seismic and wireline logs, and stochastic forward modeling of sediment accumulation in basins.; Analysis of reflection seismic and wireline log data from Pleistocene fluvial and deltaic deposits in the Eugene Island 330 field, offshore Gulf of Mexico reveal scale-invariant statistics and strong anisotropy in rock properties. Systematic quantification of lateral lithologic heterogeneity within a stratigraphic framework, using reflection seismic data, indicates that fluvial and deltaic depositional systems exhibit statistical behavior related to stratigraphic fabric. Well log and seismic data profiles show a decay in power spectra with wavenumber, k, according to {dollar}ksp{lcub}-beta{rcub}{dollar} with {dollar}beta{dollar} between 1 and 2.3.; The question of how surface processes are recorded in bed thickness distributions as a function of basin accommodation space is addressed with stochastic sedimentation model. In zones of high accommodation, random, uncorrelated, driving events produce a range of spatially correlated lithology fields. In zones of low accommodation, bed thickness distributions deviate from the random forcing imposed (an exponential thickness distribution). Model results are similar to that of a shallowing upward parasequence recorded in 15 meters of offshore Gulf of Mexico Pleistocene core. These data record a deviation from exponentially distributed bed thicknesses from the deeper water part of the cycle to the shallow part of the cycle where bed amalgamation dominates.; Finally, a stochastic basin-fill model is used to explore the primary controls on stratigraphic architecture of turbidite channel-fill in the South Timbalier 295 field, offshore Louisiana Gulf Coast. Spatial and temporal changes in topography and subsidence rate are shown to be the main controls on turbidite channel stacking pattern within this basin. The model predicts the deposition of thick, amalgamated turbidite channel sands in the basin during a period of high initial subsidence followed by deposition of thinner, less connected sands when basin subsidence rate and accommodation space are low.