New XBeach and Runup Gauge Output

[Hans]

Hi, I have been running both the previous and the newest versions of
XBeach and found that on the newest one, having runup gauge output in
params.txt causes an error during simulations (see below).

If there is "nrugauge=1" included in the params.txt, the new XBeach
does not run.

Has anyone experienced the same problem? Would there be any way to
bypass this issue and still generate runup gauge output?

Thank you

-Hans

---

...
nrugauge = 1
npointvar = 0 (no record found, default value used)
forrtl: severe (59): list-directed I/O syntax error, unit -5, file
Internal List
-Directed Read
Image PC Routine Line Source
xbeach.exe 009B7939 Unknown Unknown Unknown
xbeach.exe 009B7797 Unknown Unknown Unknown
xbeach.exe 009B6974 Unknown Unknown Unknown
xbeach.exe 009B6DA9 Unknown Unknown Unknown
xbeach.exe 009AC383 Unknown Unknown Unknown
xbeach.exe 0073B448 Unknown Unknown Unknown
xbeach.exe 0073878B Unknown Unknown Unknown
xbeach.exe 0073429D Unknown Unknown Unknown
xbeach.exe 009A1687 Unknown Unknown Unknown
xbeach.exe 009F8B59 Unknown Unknown Unknown
xbeach.exe 009DB7C9 Unknown Unknown Unknown
kernel32.dll 7C817077 Unknown Unknown Unknown

Incrementally linked image--PC correlation disabled.
Press any key to continue

---

[Fedor Baart]

Hi Hans,

I tried to reproduce an error with nrugauge but wasn't able to. Can you perhaps mail me your input, then I'll do some more testing.

Cheers,

Fedor

> --
> You received this message because you are subscribed to the Google Groups "XBeach" group.
> To post to this group, send email to xbe...@googlegroups.com.
> To unsubscribe from this group, send email to xbeach+un...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/xbeach?hl=en.
>

[Robert McCall]

Hi Hans,

When you talk about the latest and the previous version of XBeach, do
you have revision numbers for these (the revision number should be at
the top of the log file). Also, could you post the section of the
params.txt file in which you specify the runup gauges and their
locations.

Cheers,

Robert

[Hans]

Robert,

The revision number of the version I am having trouble with is 2065.

Below is where I specified the output options within params.txt

----------------------------------------------------
Output options
----------------------------------------------------

nglobalvar = 4
H
zb
zs
u
nrugauge=1
2 4 x#H#zs#zb#
tstart=0.
tintg=30
tintm=30
tintp=2


Thank you.

-Hans

> > ---- Hide quoted text -
>
> - Show quoted text -

[Hans]

Fedor,

I just sent an email with the input files.

Thank you

-Hans

> > For more options, visit this group athttp://groups.google.com/group/xbeach?hl=en.- Hide quoted text -

[M.Onur Kürüm]

Hi Hans
I believe in the latest revisions you don't need to specify the outputs at the runup gauge you want separately like you did:
nrugauge = 1
2 4 x#H#zs#zb# (what is that x ? - maybe that's the problem)
instead just put:
nrugauge = 1
2 4

(2 and 4 should point to your "real" coordinates - not the grid index - XBeach will find the grid point closest to your defined "real" coordinate.)

after you complete the run you'll have rugau001.dat in the following format:
timestep (in terms of tint)
x location of the last wet point
y location of the last wet point (it will always be the same as what you defined in the params.txt - in this case 4)
runup
timestep
xlocation
y location
runup
...

you can extract all the other variables you need from the global variables you defined.
note:I am not using the windows version and I never used the tstart, tintg, tintm, tintp before so they might change the format that I described.

HTH
Onur

[Hans]

Onur,

Thank you for your input. Your advice worked and the model runs now,
but I still have some questions.

I am trying to get runup gauge output which varies in time with the
following four variables: x(the location of the last wetpoint), H(wave
height at the the last wetpoint), zs(water surface elevation at the
last wetpoint), and zb(seabed elevation at the last wetpoint). On the
older version of XBeach, I put "2 4" to specify the y grid (y grid had
three rows, and I wanted to choose the second one, thus 2) and the
number of the variables desired (x,H,zs, and zb, as mentioned above).
With the new version of XBeach, how do I specify in the params.txt to
obtain the same outputs? Or, is there any other way for the new XBeach
model to generate the same runup gauge output data?

Thank you

-Hans

[M.Onur Kürüm]

Hi Hans
I am glad it helped.

please look at this jpg
http://s1134.photobucket.com/albums/m619/onurkurum/?action=view&current=untitled.jpg

the last x point in the second y row is selected and it reads (in case
it is too small to read) X:819 Y:5 Z:3
these are real values - this is a 3x420 Grid - X:819 corresponds to
420th point in my x grid.
to put a runup gauge in this grid I would write in my params.txt:
nrugauge = 1
819 5 (NOTE: x first then y - also I always put the gauge at the very
last x point since it does not matter - XBeach will find the last wet
point on that row. neat! )
I guess that covers the location issue.

for the other questions:
let's say my grid is designed to have the following X,Z values in the
second row (Y=5)
X=750(say this is the 220th point in x grid), Z=1
X=751(say this is the 221th point in x grid), Z=2

and let's say I ran my simulation and got the output as
2000 (tint * timestep)
750 - X
5 -Y
1.3 -runup
2100 (tint * timestep)
751 -X
5 -Y
1.8 -runup
and so on

so here you see that XBeach will give you the exact runup between
(X=750,Z=1) and (X=751,Z=2) as 1.3 - but it won't give you the exact X
location - I am pretty sure (need to check the code to be sure) it
gives you the closest X value that corresponds to Z=1.3
same for the second output at t=2100 - you got Z=1.8 as runup and this
time X location is reported as X=751. closest X value to Z=1.8
again your runup gauge is all the way back at X=819 Y=5 Z=3 while you
get this output ( I did not test but I guess it does not matter where
it is, probably only thing that matters is the correct Y value)

so,
you need to interpolate your exact X location - you know the delta X
and delta Y and you know where your runup is. (actually this can be
done by some changes in the code, maybe in a future revision:) )

H zs zb: I am not sure what your question is with these,
zb = runup at the exact location - and the H and zs will be zero at
the exact runup location - otherwise runup wouldn't be there.

if you need zs, zb, H some other location, what I would do:
nglobalvar = 3
zs
zb
H

so you'll have zs.dat zb.dat H.dat --> you can extract all the
information you need at any timestep, at any point in your domain (use
matlab, python etc)

hope this helps, let me know if this is not clear
Onur

[Robert McCall]

Hi Hans,

The option to output other variables (such as wave height) in the
runup gauge data file was removed, because if the model is set up
correctly, such variables should be more or less zero (very little
short wave energy should exist at your most landward point in the
runup). Any significant wave energy left there should be regarded with
suspicion. As far as the bed level is concerned, it should be
approximately zs-par%rugdepth at the location of maximum runup. If you
absolutely do need these output variables, I could show you where to
look in the code to change things. Alternatively you can output
"regular" globalvar output at the same resolution in time as the runup
gauge output, and post-process the data to find the last wet (wetz==1)
data point.

If (many) other XBeach users are also interested in having more output
in the runup gauge data files, let us know here what you are using it
for, and we can look into fitting that into the next release.

Robert

[Dave Thompson]

Robert and/or Fedor,

MPI and runup gauge output seem to be incompatible for me. Is this due to a lack of MPI coding for the runup gauge or am I just running into something strange on my end?

Thanks!

Dave

[Robert McCall]

Hi Dave,

Actually the runup gauges are still working, but there was a problem
with a new variable called "runup" in MPI mode (hence the confusion).
This should be fixed now. I've tested it here on a simple simulation,
but let me know if you have any problems.

Robert

[Dave Thompson]

Works now!

Thanks Robert!

Dave