Practical 4 Development of a simple stratigraphic forward model
Note that this practical is assessed and work 50% of the module mark
The aims of this practical session are to:
• Introduce you to the theory and practical considerations involved in constructing a simple
stratigraphic forward model
• Begin to apply the coding skills developed in the module sessions completed so far
• Introduce you to some more basic elements of numerical forward models, including simple
finite difference methods, numerical stability and grid boundary issues
• Develop and assess your ability to use previously developed code fragments in functions to
help construct a coding solution to a particular problem
• Develop and assess your ability to develop code from an algorithm specified in a flowchart
and following other guidance, as specified below
Instructions
1. First, briefly recap the various content of the lecture and seminar session this week that
introduced some of the theory of stratigraphic forward models, and went through an
example of various aspects of one of these models.
2. Pay particular attention to the flowchart and other instructions from the seminar
presentation:
Your main task in this practical is to write code that
implements this flowchart.
And remember, there should also a sub loop within the man time loop that calculates the
diffusional erosion and depositional change to the elevation of a surface, using a shorter
time step than used in the main loop that stores the layers of strata, so
Key elements that the model code completes should be:
• Definition of an initial topography – see the initial topography shown in the seminar
presentation, this is the one you need to use
• Calculation of diffusional erosion and deposition across the topography through multiple
forward-in-time steps
• Representation of the strata produced by this erosion and deposition, including a layer
property such as grain size based on transport distance
• A calculation of total eroded and total deposited volume per time step, with a warning
printed if these two values are not the same – differences could be due to an open grid
boundary…
• Plotting of the results as a 3D final model surfaces, cross-section and chronostratigraphic
diagram
Details of all these elements were given and talked through in the seminar slides, and all of
the initial variable values you need are actually defined for you in the diffusionStratModel.m
code
3. Following this guidance and the various other information provided in this week’s seminar
material, and using all the material from previous practical sessions, you should code the
model, debug, test and run it to produce 100 layers, with a shorter sub-loop iteration to
ensure numerical stability of the diffusion finite difference solution.
4. When the code is complete and working, you need to do some work to explore how the
model behaves and how it produces different results with different input parameters, so:
• Change the grid size to be 100 x 100 cells. Run the model with this new grid size to
check that it works ok
• Now you are going to assess how the strata changes with different values of the
diffusion coefficient . Start with a value =0.05, but also run examples with
=0.025, =0.10, =0.20. Look carefully at the results in each case. Is mass
conserved on the grid for all these diffusion coefficient values? If not, what does this
mean for the validity of the model? Could you still use it to make meaningful
predictions of erosion and deposition?
• Now write a very short report, maximum word limit 200 words, to explain
i. how the modelled strata change with changing values of ,
ii. Is mass always conserved in the model and does it matter if it is or is not?
You should look at the marking rubric for this work before you complete this report, to
make sure you understand how your report is going to be assessed and marked.
5. Paste this report text into the end of your Matlab code as a comment and save the .m file.
You can then submit your .m file source code via the practical 4 assignment submission on
Canvas. Please make sure the file submitted is a plain text .m file – if you submit any other
file type it might be difficult to mark your work.