Contents

stratigraphy – concepts
Marine sequence stratigraphy
Nonmarine sequence stratigraphy
Basin and reservoir modeling
Reflection

of our ideas how things work
Conceptual models (qualitative models)
Physical

models (experimental models)
Flume-operated simulations of sedimentologic or stratigraphic phenomena at scales ranging from bedforms to basins
Mathematical models (computer models)
Deterministic models (physically-based or process-based) have one set of input parameters and therefore yield one unique outcome
Stochastic models have variable input parameters, commonly derived from probability-density functions (pdf’s), and therefore have multiple outcomes; as a consequence model runs must be repeated many times (realizations) and subsequently ‘averaged’

and responses in a system that has specified (assumed) initial

boundary conditions (e.g., the evolution of a sedimentary basin given an initial configuration)
Inverse models use observations as a starting point and aim to estimate initial boundary conditions and combinations of processes and responses that have operated to produce the observed conditions (i.e., flip side of forward models)

What is the goal of modeling in sedimentary geology?
Understanding processes and responses in sedimentary systems (experimental and process-based models)
Prediction of sedimentary architecture and stratigraphy (primarily stochastic models)

environments (e.g., alluvial architecture); different approaches are possible, involving different

kinds of equations:
Physical
Empirical
Probabilistic

simulate basin-scale stratal patterns (e.g., sequence stratigraphy):
In geometric models the

sediment surface is represented by one or more surfaces with predetermined geometry
Many models are based on a diffusion equation that relates rates of sediment transport to topographic slopes

Slideshow

has been to start from first principles (i.e., basic, small-scale

processes of sediment transport) and multiply this to the desired spatial and temporal scale (‘upscaling’)
The outcomes of this approach have been very disappointing (i.e., upscaling is a very complicated procedure)
There is no law of nature that says that “complexity + complexity = greater complexity”!

subsurface sediments or sedimentary rocks from the perspective of fluid

flow through porous media, including issues related to resource recovery (e.g., groundwater, hydrocarbons)
The net-to-gross ratio (proportion of permeable units) is one of the most basic parameters in reservoir studies
The connectedness between permeable units is another important parameter

scale of sedimentary architecture; they are mostly stochastic
Object-based models simulate

the distribution of objects, defined by specified geometries, in 3D space; simulations are usually constrained by well data
Geostatistical models predict sedimentary facies at unvisited sites, based on the quantified spatial facies variability derived from well data (e.g., sequential indicator simulation)
Conditioning model output to observations is more easily done in stochastic models, but process-based models have the advantage that they tend to provide sedimentologically more realistic output

scale of sedimentary architecture; they are mostly stochastic
Object-based models simulate

the distribution of objects, defined by specified geometries, in 3D space; simulations are ususally constrained by well data
Geostatistical models predict sedimentary facies at unvisited sites, based on the quantified spatial facies variability derived from well data (e.g., sequential indicator simulation)
Conditioning model output to observations is more easily done in stochastic models, but process-based models have the advantage that they tend to provide sedimentologically more realistic output

scale of sedimentary architecture; they are mostly stochastic
Object-based models simulate

the distribution of objects, defined by specified geometries, in 3D space; simulations are usually constrained by well data
Geostatistical models predict sedimentary facies at unvisited sites, based on the quantified spatial facies variability derived from well data (e.g., sequential indicator simulation)
Conditioning model output to observations is more easily done in stochastic models, but process-based models have the advantage that they tend to provide sedimentologically more realistic output

to mimic real-world conditions as well as possible (scaling); this

becomes increasingly difficult with increasing spatial and temporal scales (compare bedforms vs. sedimentary basins)
Grain size (e.g., how to simulate clays?)
Grain properties (e.g., how to simulate cohesion of sediment grains?)
Fluid mechanics (e.g., how to keep the Froude number reasonable?)
Experimental models are increasingly used to simulate sedimentary architecture and basin-scale stratigraphy
One important outcome of experimental modeling is the recognition of non-linear responses

to mimic real-world conditions as well as possible (scaling); this

becomes increasingly difficult with increasing spatial and temporal scales (compare bedforms vs. sedimentary basins)
Grain size (e.g., how to simulate clays?)
Grain properties (e.g., how to simulate cohesion of sediment grains?)
Fluid mechanics (e.g., how to keep the Froude number reasonable?)
Experimental models are increasingly used to simulate sedimentary architecture and basin-scale stratigraphy
One important outcome of experimental modeling is the recognition of non-linear responses

Animation