open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello, could someone possibly confirm the wind direction convention in ascii wind data for me, please?

I'm running Telemac-2D wind option=3, with spatially and temporally-varying winds, following the examples from wind_txy. It's all running fine (although outputs I view in BlueKenue not yet satisfactorily matching input ASCII data), but I'm confused after seeing conflicting posts on the forum re the convention of the input format.

My original data (ERA5) are meteorological convention (direction FROM) and I THINK that in the input ASCII file it is the oceanographic convention (direction TO), which is needed (WINDS and WINDD per site), with angle measured clockwise from 0deg being North (comment in meteo_telemac.f).

I cannot see a definitive answer in the manuals though, apologies if I've missed it, and would appreciate confirmation here so I can keep troubleshooting why output not quite matching input (some sort of matlab/telemac/bluekenue convention issue...??). Currently updating from v8p2 (with corrected meteo_telemac.f) to v8p4, not sure whether that's relevant...

With many thanks in advance

hi slr.

in the meteo_telemac.f fortran file that compute the wind (with other meteorological data) it is written :
" WINDS,WINDD : WIND
! > Measured in m/s and degree angle clockwise from north"
which is eventually the Meteorological convention.

But if you have a problem with those two conventions (meteorological and Nautical Convention) here is what I learnt after the explanation of other members and an other website:

=> Meteorological convention :



- Describing where the object (WIND) comes from
Eg: wind 0 degree = wind coming from North and going to South
Wind 135 deg= wind coming from SE going to NW (NO in french)
- international convention for wind, used by Copernicus (era-5) or Nasa power for example to describe The wind.
- used as input in the ascii file for TELEMAC

=> Nautical or Oceanographic convention




- Describing where the object (Wind or wave) goes to...

Eg: wind 0 degree = wind coming from south and going to north
Wind coming from SE and going to NW (NO) = wind 315 degree

- used by bleukenue to show wind results after the telemac 2d and 3d computation
- the convention written in the selafin .slf result file (I need confirmation by an Admin)
- used by an other module tomawac as the default input convention for WAVE


=> Remark

As telemac input and output are in two different conventions you can be lost in the beginning but you have to focus on 2 things :
- the arrows, showing you where the wind comes from and goes to
- and the conversion : Meteorological = Nautical + 180

* sorry for the modification I made to the original Image of the "Rose des vents" to describe the nautical convention but I didn't find any interesting original image for that.

*hope this help

Josias

Hello,

To complete/correct Josias's answer:
- during TELEMAC-2D or 3D computations, wind data are stored (and potentially converted) with x and y components (and the convention is wind to which direction, with angle = 0 along x to east),
- that is why for post-treatment when using X and Y for variables for graphic printouts, wind data are written along x and y,
- for the ASCII and binary atmospheric input files, the convention is wind blowing from which direction (as for meteorological convention).

I will try to clarify this in the user manuals.

Hope this helps,

Chi-Tuan

Hello,

Inputting in meteorological convention within the ASCII input file was the key confirmation I required - thank you both so much! I ran some more simple tests and have it working - knowing what you told me I was able to check/correct all graphical output ambiguities between different software conventions, and have scaled back up to full ERA datasets and it all looks good.

Josias - many thanks for your detailed reply, much appreciated.

Chi-Tuan - thanks also for the confirmation. Your comment re convention of 0 in east confused me somewhat, as this contradicts the comment in meteo_telemac.f ("Measured in m/s and degree angle clockwise from north") but I could be misunderstanding. Using ERA-5 convention of 0 being FROM the north, I now get outputted X/Y vectors pointing towards due south in B.K. (and so on), so I think it's sorted anyway.

For all this I followed the example ascii format in file fo1_wind_txy, which includes simple wind input data as WINDS and WINDD (but without headers). What I haven't tried is inputting headers WINDX and WINDY (and including ERA5 data as it comes in u and v) in an ascii wind input file for wind option=3, as indeed it may sort out this convention issue completely, negating the need for this conversion from x/y to speed/dir. One for another rainy day...

With many thanks!