Incorrect Initial Values at the Bottom of the Model Caused by FastScape

Dear Experts,

I am currently using ASPECT version 3.0.0 to run a continental rift model. I have attempted to integrate FastScape for sediment layer simulation, but I’ve encountered an issue with the two fields I added: grain_size_evolution and water_content. Specifically, the values in the bottom 5 km of the model are initialized to zero. However, when I disable FastScape in the mesh deformation subsection, this issue does not occur, and the fields are initialized correctly.

Here are the parameters I have configured for FastScape. I would greatly appreciate your advice on how to modify these settings to resolve this issue.


subsection Mesh deformation
set Mesh deformation boundary indicators = top : fastscape
set Additional tangential mesh velocity boundary indicators = left, right

subsection Free surface
set Free surface stabilization theta = 0.5
end

subsection Fastscape
set Number of fastscape timesteps per aspect timestep = 5
set Maximum timestep length = 1000
set Vertical exaggeration = -1
set Additional fastscape refinement = 0
set Average out of plane surface topography in 2d = true
set Fastscape seed = 1000
set Maximum surface refinement level = 1
set Surface refinement difference = 0
set Use marine component = true
set Y extent in 2d = 40e3
set Use ghost nodes = true
set Uplift and advect with fastscape = true
set Node tolerance = 0.001
set Sediment rain rates = 0.03e-3,0.01e-3
set Sediment rain time intervals = 15e6
set Initial noise magnitude = 5

subsection Boundary conditions
  set Front = 1 # Front (bottom) boundary condition, where 1 is fixed and 0 is reflective.
  set Right = 1 # Right boundary condition, where 1 is fixed and 0 is reflective.
  set Back  = 1 # Back (top) boundary condition, where 1 is fixed and 0 is reflective.
  set Left  = 1 # Left boundary condition, where 1 is fixed and 0 is reflective.
end


subsection Erosional parameters
  set Drainage area exponent                          = 0.4  
  set Slope exponent                                  = 1    
  set Multi-direction slope exponent                  = -1  
  set Bedrock deposition coefficient                  = 1   
  set Sediment deposition coefficient                 = 1
  set Bedrock river incision rate                     = 1e-5
  set Sediment river incision rate                    = -1
  set Bedrock diffusivity                             = 5e-3
  set Sediment diffusivity                            = -1
  set Orographic elevation control                    = 2000
  set Orographic wind barrier height                  = 500
  set Elevation factor                                = 1
  set Wind barrier factor                             = 1
  set Stack orographic controls                       = true
  set Flag to use orographic controls                 = false
  set Wind direction                                  = west
  set Use a fixed erosional base level                = false
  set Erosional base level                            = 0
end

subsection Marine parameters
  set Sea level                        = -500
  set Sand porosity                    = 0.0
  set Silt porosity                    = 0.0
  set Sand e-folding depth             = 0
  set Silt e-folding depth             = 0
  set Sand-silt ratio                  = 1
  set Depth averaging thickness        = 1e2
  set Sand transport coefficient       = 200
  set Silt transport coefficient       = 200
end

end
end

Thank you for your assistance!

Chenghui Zeng

@zengchenghui -

Here are the parameters I have configured for FastScape. I would greatly appreciate your advice on how to modify these settings to resolve this issue.

Can you post the full PRM file, it won’t be possible to debug this issue without seeing quite a few more sections? Ideally, please first try and break down the parameter to the simplest case that can reproduce the issue.

Specifically, the values in the bottom 5 km of the model are initialized to zero. However, when I disable FastScape in the mesh deformation subsection, this issue does not occur, and the fields are initialized correctly.

When you disable Fastscape, were any other parameters changed? At first thought, this sounds like an issue with the composition boundary conditions.

zch_testing.prm (56.2 KB)
Thank you for your answer.
I have uploaded the prm input file. I used some plugins to initialize the water and grain field, and these initializations are correct if Fastscape is not turned on, but if Fastscape is turned on, the problem I had earlier occurs. Next I will simplify the model parameters, keep only the plugin that initializes the values, and see where the problem occurs.

Dear professor, based on your suggestion, I simplified the structure of the model. I added the FastScape functionality in the continental_extension.prm file located in the cookbooks folder. At the same time, I added a field called grain_size_evolution to the file, which is not involved in any calculations but is only meant to observe whether the initialization works correctly.

I found that for any fields related to the bottom boundary, including mantle_lithosphere and grain_size_evolution, the initialized bottom boundary values are set to 0, as shown in the attached image. As a result, after trying several options, I determined that the issue originates from the set Fixed composition boundary indicators option in the Boundary Composition model. When this option includes bottom, the initialization issue occurs. If the bottom is removed from this setting, the issue does not arise.

zch_fastscape_test.prm (19.2 KB)

Hi Zengchenghui,

I think it looks like the problem is that only the function describes the boundary composition, and it says to fix it to zero for all fields. Originally since no fields were here for other coupled models it was fine to have everything set to zero, but for fields you want to have a set bottom inflow value you could set them to something like if(y<80e3, 1, 0), which should hopefully work.

Hi Derek,
I modified the boundary conditions according to your suggestion, and it really worked. Thank you very much for your guidance. In addition, I would like to ask about the “set Fixed composition boundary indicators = top,bottom”, Can the “bottom” in this option be removed? Does not selecting this option affect model calculations?

Best wishes,
Chenghui

Hi Chenghui,
Yes, it is possible that if you have compositions that are quickly different above the fixed boundary it could cause issues, but I don’t think this is an issue in most cases. In general when having compositional fields on an inflow boundary you will want to fix the boundary, as otherwise you can have the input of incorrect compositional fields (or temperatures if the temperature isn’t fixed on an inflow boundary).