Ocean Modeling Discussion

Message from OSU: CEOAS and POA will be holding an event in recognition of Professor John Allen next week on Monday, October 23 starting at 3:00 pm (PT). Speakers will include Tuba Özkan-Haller, Doug Caldwell, Ken Brink, and Dudley Chelton. This event will be held in person in Burt 193 and also ...

I am so sorry to hear about this. Thanks for sharing, John, W. We have lost a giant of the community. I didn't get to interact too much with John Allen. But I have read a lot of his papers, which have helped me a lot, particularly, on the topic of coastal-trapped waves. I vividly remember all the ...

Woods Hole Oceanographic Institution is currently searching for a Research Associate with extensive ROMS experience to join the Applied Ocean Physics & Engineering Department. This is a regular, full-time, non-exempt position, and is eligible for full benefits. The initial appointment is for 12 ...

I am saddened by the news. Dale Haidvogel was the person who brought me to the U.S. and into the oceanography community, and let me pursue my own interest. For that, I am forever grateful! I was in Fluid Mechanics when I was working on a master's degree at Zhejiang Univ in China 20 years ago. By ...

As chairs and vice chairs of the Gordon Research Research Conference (GRC) on Coastal Ocean Dynamics we would like to invite you to participate in the 2023 Coastal Ocean Dynamics GRC to be held 18-23 June 2023 at Bryant University in Rhode Island. The entire conference program is now available at ...

Hi, all, The Gordon Research Conference (GRC) - Coastal Ocean Dynamics is coming back after 4 years, and we would like to invite you to attend. It will take place at Bryant University in Rhode Island on June 18-23, 2023: https://www.grc.org/coastal-ocean-dynamics-conference/2023/ There will also be ...

It is really strange to see Robinson and Dong's claim that the ROMS GLS code has issues for simulating internal tides. I have been using GLS for simulations of internal tides at a shelf edge for years. I have no problem with the code. The modeled viscosity and diffusivity look fine to me, and the ...

Hi, all,
I just came across this new paper about running a Python-based ocean model on GPUs achieving better performance than traditional CPU-based parallel simulations:
https://agupubs.onlinelibrary.wiley.com ... 21MS002717

Any thought on this?

Gordon

Hi, all, We are currently looking for a (and maybe more) Research Associate with a background in ocean modeling to join the Woods Hole Oceanographic Institution. This is a regular, full-time, non-exempt position, and is eligible for full benefits. The initial appointment is for 18 months with the ...

Thanks, Hernan. I am restarting a simulation from a perfect restart file and asking the model to output multiple history and average files. The problem occurs when the model tries to create a new average file during the middle of the simulation. But it has no problem with creating new history files ...

Hi, I am running into the same problem when I restart a simulation and ask the model to generate multiple average files. The problem occurs when it tries to create a new average file. The error message is: OUTPUT - multi-file counter AVG(ng)%load = 2, is greater than Nfiles = 1 dimension in ...

Woods Hole Oceanographic Institution is currently searching for a Postdoctoral Investigator to join the Coastal Ocean Fluid Dynamics Laboratory in the Applied Ocean Physics & Engineering Department. This is a temporary, full-time, exempt position, and is eligible for benefits. JOB SUMMARY ...

Line 1148 in mod_mixing.F, the CPP flag FORWARD_MIXING is misspelled as FOWARD_MIXING.

Gordon

Hey, If you want your calculation of the residence time to be mathematically precise and also resolve its temporal and spatial variability, you should definitely look into the Constituent-oriented Age and Residence time Theory (CART; http://www.elic.ucl.ac.be/repomodx/cart/ ). The forward model ...

We don't need FORWARD_BULKFLUX. Yes, the linearization of bulk flux is problematic. The way to get around that is to save the surface tracer and momentum fluxes in the forward nonlinear simulation (remember to save the shortwaver radiation separately). The ROMS adjoint and tangent linear models will ...

Is Bac_F1 part of the initial condition? If it is, you must have bad data in your initial condition since model encounter that problem while writing the initial record.

I've never used 2.2. But for the latest version of ROMS we can get using svn, there are more than one place we need to change to add point source for a passive tracer. Some changes in Nonlinear/get_idata.F is required to turn on SOURCES(ng)%Lsrc and SOURCES(ng)%Ltracer for the added passive tracer ...

You need to add passive tracer concentration variable in your initial condition file. This variable should have the same dimension as temperature (temp) and salinity (salt). The name of the variable should be consistent with that in varinfo.dat (the default one is 'dye_'). If you stick with the ...

which version of pgi are you using?

Hey, Zhou Feng,

Check your permission under the directory you are compiling. Maybe you don't have the permission to create the scratch directory (./Build/). But I am not sure.

You need to install svn on the machine you are compiling on. Type 'which svnversion' to see whether you have it installed. Check the 'makefile' file, there is a line like: SVNREV ?= $(shell svnversion -n .) This tells the system which version of the code you are working on, and the version number ...

FengZhou wrote:

Hi, Weifeng,
Thank you for the suggestions. I will bother you if necessary:)
Do I know you? I have a friend has the close name with you from Zhejiang University. He is in U.S.A now:)

Yes, it is me.

Hey, Zhoufeng, You should really consider to update the code. There has been a lot of changes both in model and in compilers since then. Personally, I don't have any experience with roms2.3, but I wouldn't be surprised by some problem between very old version of roms and new version of compilers ...

I tried the IEEE arithmetic standard ('-Kieee' option for pgi), the model results with different versions of pgi (pgi6.1, pgi7.0) became identical exactly as Rich said. But the results from different compilers (pgi, ifort and g95) are still different by same amount as before even all of them are ...

hey, Natalie, The pgi6.1 I have gives the right answer as pgi7.0. I also tried pgi6.0 which does give wrong boundary indices as you said. So, the bug has been corrected in my pgi6.1. For the problem I was talking at the beginning of this thread, I was using pgi6.1 and pgi7.0. So, that is not the ...

try "Zhou\ Wei"?

We recently updated our pgi compiler, and I am comparing the results from models compiled with different versions of pgi, 6.1 and 7.0. I ran the two nonlinear models for 120 time steps forced by exact same atmospheric forcings and initial conditions. For both of them, Flather and Chapman are used ...