search

danigeos / tisc

TISC is a code to model isostasy, erosion/sediment transport/sedimentation, and tectonic kinematics, mostly for the study of sedimentary basins and source to sink systems.
https://sites.google.com/site/daniggcc/software/tisc
16 stars 2 forks source link

Precipitation not working correctly #12

Open jennawest opened 8 years ago

jennawest commented 8 years ago

Hi Daniel,

I am having a very similar problem to my last where the erosion and sediment transport were not working correctly except now it is the precipitation and resulting drainage pattern. I noticed that when I run your demo 7_climate I get a very different drainage pattern. Whenever I do any run where hydro_model is anything except zero, I get many lines of errors. The model completes the run and produces a .ps but as I said, something seems to not be working correctly. I've attached my file output when I run 7_climate as well as the command line (filename climate_commandlines) output for the last time section of the run.

Thanks

7_climate3.zip

danigeos commented 8 years ago

I have difficulties to follow your question. 'Different' relative to what? You output files seem to be ok, except of course that you have removed the orographic precipitation etc. (hydro_model 1) relative to the demo 7 Also, did you download my last uploaded TISC version (May 18th)? I arranged some things in the gmt scripts that should make them more understandable. Finally, when do your errors start? I run for some Myrs iand i got nothig comparable (using your setting as you sent it): output.txt

jennawest commented 8 years ago

My results are different relative to your results. I attached the wrong file of my input and output last time, sorry. Attached are the correct files. You can see that even though all my settings are the same as yours (including hydro_model 3) I get a different drainage pattern than you. I know some of our GMT settings are different so our .ps will look different but what bothers me is the drainage pattern differences. Yes, I downloaded the updated version of TISC. The error messages begin immediately and continue until 4.4 Ma. 7_climatecorrect.zip

danigeos commented 8 years ago

Ok, I have three questions:

  1. What is your platform? linux? MAc?
  2. Use the same script (job) as linked in the original version (script/tisc.topo+climate.gmt.job, linking this file to tisc.gmt.job in your working directory), so that we can see the precipitation.
  3. Also, can you run with the option -Pc as I do in the demo? (probably not the cause for the differences, but then we can see all time steps as jpegs)
  4. I get no error messages here. Please use this linux trick to copy your screen output to a txt file: tisc climate -Pc | tee climate.outputscreen.txt.
    This will not include the errors, so you have to manually copy them to another file (only the first 100 lines or so is ok).

I'm afraid this is going to be a compilation issue that may be hard to find :-S

jennawest commented 8 years ago

In response to your questions:

  1. I am running a virtual machine in Ubuntu
  2. I'm using the original version 3, 4. I ran the command you said and have attached the files. There were so many errors that my command window wouldn't go back to the first 100 lines of error so these are from time slice 4.2 but are representative of all the time slice error messages (in file errormessages.txt) 7_climate_3.zip
danigeos commented 8 years ago

you file is corrupted, Also, to get the first error, add -tf0 to stop the program after the first step.

In the meanwhile, i would suggest running in another native Ubuntu machine or mAc to see if you get rid of these issues.

as for 2.: somehow tisc is not calling the right job for the figure. in the original download, you should have a symbolic link in your demo/7_climate/ directory (link file named tisc.gmt.job), pointing to a tisc.climate.gmt.job script in the tisc/script directory. Check if that is correct. If you moved your project directory, the link is probably not working properly.

jennawest commented 8 years ago

Hi Daniel, I've gotten a Mac and compiled TISC on the mac. All of the demos produce the same .ps as yours except 8_Ebro_evol_Z which produces a different drainage and erosion pattern than you so the program is running much better but still not perfect (I still get hundreds of lines of error messages). I'd like to replicate your environment and was hoping you could tell me your specific environment conditions/how to set it up on OS X since the readme instructions are for a linux.

danigeos commented 8 years ago

I use a C-shell and there are no specific setups that i can think of. Can you send the stdout and stderr outputs of a specific demo to a file and pass it to me? then i can figure out.
If the results are ok (ignore the differences for the Ebro model) then the larger number of error messages should be linked to the misuse of the -V option, it's the only

jennawest commented 8 years ago

I have attached stdout and stderr in the attached file: climateLog.txt

If I may ask, why can I ignore the Ebro model? It is more similar to the type of model I am constructing so it seems like a good demo.

it's the only...?

danigeos commented 8 years ago

Because i saw recently that the results in the file are not perfectly in agreement with the PRM file (iow, i did not run the last PRM before building the version). Not really significant.

Remember to pack your directory and send it again, the file you sent on 7/26 was corrupted.

Also, to solve the problem, i need to know exactly what changes you do to the distributed version to succeed in compilation, and if you get any warnings.

About your linux comment, i should have said that the config instructions is taking mac as a version of linux.

jennawest commented 8 years ago

To get the program running on my computer I had to download Xcode, homebrew, gmt4, ksh, tcsh, and gcc (which has tofrodos). I downloaded the newest version of tisc from the github. I changed the path to include these new programs as well as tisc/script and tisc/bin as instructed. I did not modify the 'tisc/config.mk' or 'tisc/src/Makefile'. I get no errors when I type "tisc make".

When running demo/7_climate I did not change anything. I simply copied your file and ran the copy. The result is that there are hundreds of lines of error from 0 Ma to 3.5 Ma then there are no more errors. I get the same results as you (other than slight differences in GMT settings). I've attached the entire file of the copied 7_climate that I ran. I've once again included the climateLog.txt of the stderr stdout of the first time slice in addition to climateLogend.txt which shows how the errors stop after 3.5 Ma.

7_climate1.zip

jennawest commented 8 years ago

I have to correct myself:

I originally tried to change config.mk but I got errors and the program wouldn't make.

When I try to change the line in config.mk that shows DEFS = -DSURFACE_TRANSPORT #-DTHIN_SHEET
to DEFS = -DSURFACE_TRANSPORT -DTHIN_SHEET the program won't make. The output is shown in thinsheet_on.txt. When I leave #DTHIN_SHEET the program makes successfully and I get the output shown in thinsheet_off.txt.

thinsheet_off.txt thinsheet_on.txt

danigeos commented 8 years ago

You are not going to need the thin sheet crustal deformation, so you can leave it off.

As for the errors: i cannot find the cause right now. Nothing is different between your system and mine, apparently. Since the results are surprisingly not changing, i would suggest that you edit the src/surf_proc.c file, delete or comment lines 495,496 (starting at "if (!its_sea)"), and recompile. Not the most elegant solution, i know, but i am traveling this month and somewhat busy.