#1 2016-10-27 19:22:35

Toozler
Member
Registered: 2016-10-27
Posts: 3

Automate conversion from .WRG to .GRD/.ASC

Hi!

I'm aware of WAsP Utility Program "rsf2grd.exe" (from versions prior to WAsP 10) to convert the WRG to the GRD format or even XYZ format.

I'm looking for a way to perform the same conversion in a Linux environment.

Is there any further documentation on how the rsf2grd tool works or even a Linux port of the code?

Thanks
-A

Offline

#2 2016-10-31 14:53:02

Toozler
Member
Registered: 2016-10-27
Posts: 3

Re: Automate conversion from .WRG to .GRD/.ASC

A quick update:

The .wrg format is well documented, so I was able to write a simple code to read the file and generate a XYZ grid with wind speeds based on the Weibull-A value from columns 44-48(total distribution for all sectors).

However, when comparing my XYZ grid with a XYZ grid generated using the rsf2grd tool, there is a consistent difference in the results (in my case, around ~1m/s).

Is there any transformation that I'm missing?

Offline

#3 2016-11-09 11:35:45

Duncan
WAsP team
From: Europe
Registered: 2009-05-14
Posts: 286

Re: Automate conversion from .WRG to .GRD/.ASC

Hello,

Linux is not supported by our software.

But in general you should be able to get wind speed results that make sense and are consistent. What are you generating, exactly? All-sector mean wind speed from the WRG 'Total'? How are you calculating it and with what are you comparing it?

Duncan..

Offline

#4 2016-11-09 15:42:15

Toozler
Member
Registered: 2016-10-27
Posts: 3

Re: Automate conversion from .WRG to .GRD/.ASC

Hi Duncan,
When converting a WRG to an XYZ grid with rsf2grd using the -u switch (mean wind speeds), how exactly does it calculate those values for each X and Y cell coordinate? Does it use the total distribution's Weibull A and K with a gamma function?

Offline

#5 2016-11-09 18:14:21

Duncan
WAsP team
From: Europe
Registered: 2009-05-14
Posts: 286

Re: Automate conversion from .WRG to .GRD/.ASC

This is really an archaeological question! We don't know about the code inside rsf2grd. But indeed, the old fashioned way of getting all-sector mean U was to use a Gamma function totalling of the sector wise A and K. It will be hard to reproduce the numbers exactly from a home made implementation. We found it was even sensitive to different Fortran and C++ compilers.

Can you give me some more idea about which numbers (inside WASP) you are interested to use, and what you want to do with them?

Offline

Board footer

Powered by FluxBB

[ Generated in 0.009 seconds, 10 queries executed - Memory usage: 522.43 KiB (Peak: 524.09 KiB) ]