Error in vertical cross sections when crossing 0º meridian

Errors and unexpected results
Post Reply
User avatar
Miguel A. Martinez
Posts: 5
Joined: Mon Apr 23, 2018 3:34 pm

Error in vertical cross sections when crossing 0º meridian

Post by Miguel A. Martinez »

I have created netCDF files with temperature, humidity, etc profiles on a region that covers Iberian Peninsula.

When I use McIDAS-V1.7u1 the vertical cross sections on left or right the 0º meridian are produced OK; as can be seen in the attached CV_ok1.png (position line) and CV_ok2.png (vertical cross section)
But when the start and end points are in different sides of the 0º meridian the vertical cross section appears as if the path is not the sort one and if would be using the great circle around the globe; as can be seen in the attached CV_KO_1.png and CV_KO_2.png.

Best regards
Miguel A. Martinez (AEMET)
Attachments
CV_ok2.png
CV_ok1.png
CV_KO_2.png
CV_KO_1.png
User avatar
bobc
Posts: 988
Joined: Mon Nov 15, 2010 5:57 pm

Re: Error in vertical cross sections when crossing 0º meridian

Post by bobc »

Hello Miguel,

Would you please post a sample netCDF file to our ftp location? Here are instructions on doing that. From a terminal/command prompt window:

Code: Select all

ftp ftp.ssec.wisc.edu
user: anonymous
password: your email address
bin
cd pub/incoming
put filename
bye

Please let me know the file's name when you post it and I'll take a look.

Thanks,
Bob Carp
McIDAS Help Desk
User avatar
Miguel A. Martinez
Posts: 5
Joined: Mon Apr 23, 2018 3:34 pm

Re: Error in vertical cross sections when crossing 0º meridian

Post by Miguel A. Martinez »

I have upload the example netCDF file "PGE00-VISIR_HYB_nwp_ALL_LargeIberica_C1748_0558_S0665_0511_2015-07-15T12-00-00Z_01x01_SAT_msg3.nc" in the pub/incoming directory of ftp.ssec.wisc.edu.

This size of this netCDF file is 395550108 bytes. It contain 2D and 3D arrays.

I have checked it with McIDAS-V 1.6. I have used similar files with previous McIDAS-V versions as you can see in http://www.nwcsaf.org/AemetWebContents/ ... 0pas00.gif

Thanks in advance
Miguel A. Martinez
User avatar
bobc
Posts: 988
Joined: Mon Nov 15, 2010 5:57 pm

Re: Error in vertical cross sections when crossing 0º meridian

Post by bobc »

Hello Miguel,

Thanks for your data file. I've replicated this bug in our McIDAS-V 1.7u1 build and verified that it worked in 1.6. I believe I may have narrowed down where this bug was introduced and the programmers have been made aware of it. I've written this up as Inquiry 2660 and I'll keep you up to date on any progress as far as a fix.

Thanks again for reporting this,
Bob
User avatar
bobc
Posts: 988
Joined: Mon Nov 15, 2010 5:57 pm

Re: Error in vertical cross sections when crossing 0º meridian

Post by bobc »

Hello Miguel,

Our programmers put in a fix for this problem in our 1.8beta1 nightly build, which you can download here (the login and password are both: mcv). I've checked this on Windows 7 and tested a variety of other gridded data files that cross the prime meridian and/or the dateline and things seem to be working well. Please note that these 1.8beta1 nightly builds are automatically created every day with all of the previous day's programming changes, so not everything in the build has been fully tested.

If you give the 1.8beta1 nightly build a try and run into any problems with cross sections crossing the prime meridian, or anything else, please let us know.

Thanks,
Bob
Post Reply