Re: [EXTERNAL] Trouble Loading Restart Concentration Field from ATS Checkpoint File
Coon, Ethan
First, if you’re using release version 1.5, can you try switching to master? This will not let you use your restart file, but it hopefully would mean that future restarts will not have this problem (and you’ll see lots of other bug fixes as well).
As of tomorrow, I’d encourage anyone doing transport who is willing to beta-test to start using master. Several major PRs have landed in the last week or two, and there is likely two more coming prior to release of 1.6 -- one changing a variable name (hopefully coming later today), and the other changing how sources are parsed. But it’s nearly ready and I’d love for people to break it and get us feedback before release.
The second thing: Tc-99 is an invalid variable name – see https://amanzi.github.io/ats/dev/input_spec/introduction.html#variable-naming
Hyphens should not be used in variable names, with the lone exception being if it is the last character(s) in a component name, e.g. NO3- where it may be used to indicate charge.
I know that Tc-99 used to be a common tracer name in demo input files, and probably still appears in our demos somewhere. But please switch to using Tc99 (if you’re actually tracing Technetium) or Tracer (if you’re just wanting a generic passive tracer).
Ethan
From:
ats-...@googlegroups.com <ats-...@googlegroups.com> on behalf of Xueyuan Kang <xueyuan...@gmail.com>
Date: Wednesday, May 21, 2025 at 1:04 AM
To: Amanzi-ATS Users <ats-...@googlegroups.com>
Subject: [EXTERNAL] Trouble Loading Restart Concentration Field from ATS Checkpoint File
Hello all, I’m trying to restart a solute transport simulation in ATS from a previously saved checkpoint file. In both the original XML file (used to generate the checkpoint) and the current restart XML, I’ve defined the component name as "Tc-99".
Hello all,
I’m trying to restart a solute transport simulation in ATS from a previously saved checkpoint file. In both the original XML file (used to generate the checkpoint) and the current restart XML, I’ve defined the component name as "Tc-99".
However, when I restart the simulation, the log file shows the following messages:
ATS | Beginning initialize stage...
State | initializing data through initial conditions...
Field total_component_concentration.cell.0 is not found in hdf5 file.
subsurface transpo | Number of components: 1
subsurface transpo | cfl=1 spatial/temporal discretization: 1 2
subsurface transpo | Initalization of PK is complete.
subsurface transpo |
Field surface-total_component_concentration.cell.0 is not found in hdf5 file.
surface transport | No BCs were specified.
surface transport | Number of components: 1
surface transport | cfl=1 spatial/temporal discretization: 1 2
surface transport | Initalization of PK is complete.
surface transport |
From the results at the first timestep, it seems the initial concentrations were not properly loaded from the checkpoint.
I inspected the checkpoint file (checkpoint1637990.h5), and it does contain the fields:
- total_component_concentration.cell.Tc-99
- surface-total_component_concentration.cell.Tc-99
In the XML file, I’ve defined the initial condition for both subsurface transport and surface transport PKs as follows:
<ParameterList name="initial condition" type="ParameterList">
<Parameter name="restart file" type="string" value="../3-transient_2012/checkpoint1637990.h5" />
</ParameterList>
It seems ATS is still attempting to read the field ...cell.0 instead of ...cell.Tc-99, even though the component name (Tc-99) is set in both "surface transport" and "subsurface transport" PKs. I’d appreciate any suggestions on why this might be happening or what additional configurations might be required to ensure that the solute concentration is correctly loaded from the checkpoint.
Thanks for your help!
Best regards,
Xueyuan
--
You received this message because you are subscribed to the Google Groups "Amanzi-ATS Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to
ats-users+...@googlegroups.com.
To view this discussion visit
https://groups.google.com/d/msgid/ats-users/0ce7128e-b185-4033-b5ad-f8766ffcae04n%40googlegroups.com.
