[RegCNET] A problem about RegCM-5.0.0
汪文超
202212010086 at nuist.edu.cn
Sun Mar 19 15:32:00 CET 2023
Dear RegCM community, I'm a postergraduate student named Wenchao Wang in Nanjing university of Information Science & Technology (NUIST) while my major is Meteorology.
When I use the latest version RegCM-5.0.0(Released),I find that the results are always biased everytime. However, I used the same driver data.
Fig1 shows the state at the beginning of the run and fig2 shows the state of integration after a period of time.(The left image is the state of the first run, and the right image is the state of the second run.)
The only difference between the two is the number of CPU cores on the high-performance computer I used.Fig 3 is the root-mean-square error of my previous 5 calculations.
Do different numbers of cores affect the results?
Wenchao Wang
NUIST
19 Mar
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.ictp.it/pipermail/regcnet/attachments/20230319/267f8230/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: fig1.png
Type: application/octet-stream
Size: 344650 bytes
Desc: not available
URL: <https://lists.ictp.it/pipermail/regcnet/attachments/20230319/267f8230/attachment-0003.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: fig2.png
Type: application/octet-stream
Size: 339526 bytes
Desc: not available
URL: <https://lists.ictp.it/pipermail/regcnet/attachments/20230319/267f8230/attachment-0004.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: fig3.png
Type: application/octet-stream
Size: 11936 bytes
Desc: not available
URL: <https://lists.ictp.it/pipermail/regcnet/attachments/20230319/267f8230/attachment-0005.obj>
More information about the RegCNET
mailing list