Difference between revisions of "SELFE FAQ"

From ccrmwiki
Jump to: navigation, search
 
(3 intermediate revisions by the same user not shown)
Line 1: Line 1:
−
<OL>
 
−
  <LI> <A "#anchorA" NAME="faq1" href="#faq1_2">How do I set up a vertical grid?</A>
 
−
  <LI> <A "#anchorA" NAME="faq12" href="#faq12_2">My results show platform/compiler/CPU dependency.
 
−
  <LI> <A "#anchorA" NAME="faq13" href="#faq13_2">Run crashed with a fort.11 error message"QUICKSEARCH: no intersecting edge....".
 
−
</A>
 
−
  <LI> <A "#anchorA" NAME="faq10"  href="#faq10_2">How to do a tidal simulation with SELFE?</A>
 
−
  <LI> <A "#anchorA" NAME="faq5"  href="#faq5_2">How do I make GOTM libraries?</A>
 
−
  <LI> <A "#anchorA" NAME="faq2"  href="#faq2_2">How do I choose the flags in lqk.gr3?</A>
 
−
  <LI> <A "#anchorA" NAME="faq3"  href="#faq3_2">My river and plume are too fresh.</A>
 
−
  <LI> <A "#anchorA" NAME="faq4"  href="#faq4_2">I'd like to add some passive tracer transport to SELFE. How can I interface my own code to it?</A>
 
−
  <LI> <A "#anchorA" NAME="faq6"  href="#faq6_2">How to impose river discharge if the depth is negative there?</A>
 
−
  <LI> <A "#anchorA" NAME="faq8"  href="#faq8_2"> How do I set up sflux/ using NARR (North America only) files in the
 
−
        Utility Library for atmospheric model (nws=2)? </A>
 
−
  <LI>  <A "#anchorA" NAME="faq7"  href="#faq7_2"> How do I prepare my own netcdf files for atmospheric model (nws=2)? </A>
 
−
  <LI> <A "#anchorA" NAME="faq9"  href="#faq9_2">  My non-hydrostatic run blows up. </A>
 
−
  <LI> <A "#anchorA" NAME="faq14"  href="#faq14_2"> I have large velocity at open boundary </A>
 
−
  <LI> <A "#anchorA" NAME="faq15"  href="#faq15_2"> My run crashed; how can I find out
 
−
    why?</A>
 
−
</OL>
 
−
<HR>
 
−
<HR>
 
  
−
<HR>
 
−
    <p> </p>
 
  
−
<p> <b><font color="#FF0000"><a name="faq1_2">How do I set up a vertical grid? </a></font></b></p>
+
==My results show platform/compiler/CPU dependency==
−
    <p> </p> You should set it up according to your application. The first question to ask is whether a "pure S" model is
+
Due to some intricate differences between different compilers/platforms some minor differences in results are expected. Bit-by-bit match of results using different numbers of CPUs is impossible also. But the differences should be small and should stablize over time iteration.  
−
sufficient for your problem. We found it's adequate for most problems (99.9%) except when strong stratification encounters
 
−
steep bottom slope (e.g., strongly stratified Columbia River plume flows over a shelf break
 
−
  with slope in excess of 1/20). In the S part of vgrid.in, adjust the constants h_c, theta_b, theta_f to resolve appropriate
 
−
boundary layers. Note that h_c also determines the boundary between S-zone and sigma-zone (where traditional sigma-coordinates
 
−
are used becasue S-transformation is invalid when h &lt h_c).
 
  
−
If (in the unlikely case that) a SZ model is needed, h_s may be chosen as the max. depth of the shelf, and make sure there is
+
==Run crashed with a fort.11 error message "QUICKSEARCH: no intersecting edge...."==
−
a smooth transition between S and Z coordinates. You can get a preview of some sample z-coordinates at various depths
+
First you need to check the results before the crash (e.g. surface velocity) to see if anything is outrageously wrong.
−
by running the code with the pre-processor flag (ipre in param.in) turned on.
+
If the results look reasonable, this may be related to the version of SELFE you are using; contact the developers for more info.
  
−
<p> <a href="#faq1">go back</a></p>
+
 
 +
==How to do a tidal simulation with SELFE?==
 +
 
 +
The simplest way is to use SELFE 2D. If you have to use SELFE 3D, make sure the bottom boundary layer is reasonably resolved.
 +
 
 +
==My run crashed; how can I find out why? ==
 +
 
 +
See [[Debugging after a run crashes]].
 +
 
 +
<!--==I'd like to add some passive tracer transport to SELFE. How can I interface my own code to it?==
 +
==How to impose river discharge if the depth is negative there?==
 +
==How do I set up sflux/ using NARR (North America only) files in the Utility Library for atmospheric model (nws=2)?==
 +
==How do I prepare my own netcdf files for atmospheric model (nws=2)?==
 +
==I have large velocity at open boundary==
 +
-->

Latest revision as of 19:48, 15 September 2013


My results show platform/compiler/CPU dependency

Due to some intricate differences between different compilers/platforms some minor differences in results are expected. Bit-by-bit match of results using different numbers of CPUs is impossible also. But the differences should be small and should stablize over time iteration.

Run crashed with a fort.11 error message "QUICKSEARCH: no intersecting edge...."

First you need to check the results before the crash (e.g. surface velocity) to see if anything is outrageously wrong. If the results look reasonable, this may be related to the version of SELFE you are using; contact the developers for more info.


How to do a tidal simulation with SELFE?

The simplest way is to use SELFE 2D. If you have to use SELFE 3D, make sure the bottom boundary layer is reasonably resolved.

My run crashed; how can I find out why?

See Debugging after a run crashes.


Retrieved from "http://ccrm.vims.edu/w/index.php?title=SELFE_FAQ&oldid=731"