MRMS regridding script crahes

[HAOWEN YUE]

Dear all

I am trying to regrid MRMS forcing data using the regrdding script and example data provided on the website. However, the script crashes my laptop.

I also tried to run the script through the docker container, but it didn't work either. I have no idea where went wrong since there was no error message..

The last thing I tried is to run the script through Jupyter Notebook, which gave me the following error message. Please see the attached screenshots.

Did anybody else run the example MRMS script successfully?

Any help is welcome. Thank you!

Sincerely

Haowen Yue

[Katelyn]

Hi Haowen,

It looks like you’re using the patch option for interpolation. Have you tried using the bilinear or conservative options? These should be better tested.

That said, patch should work. I’ll take a closer look when I get a chance.

Katelyn

[HAOWEN YUE]

Hi Katelyn

Thank you for the reply.

I tried bilinear and patch, either didn't work. I can try conservative later.

Also, HRRR script also crashed once on my laptop. (I ran the HRRR script several times and it crashed once.) So I doubt it may be the problem of my own laptop.

Does the regridding process use a lot RAM? Or there is any specific requirement to run the script?

Thanks

Haowen Yue

[Katelyn]

Hi Haowen, 

I think you're correct that it's likely a memory issue.  The MRMS regridding script was also being killed in my Docker instance where the memory allocation is relatively small.  It looks like it requires around 14 GB of RAM.  Unfortunately, the MRMS grids are rather large and the underlying utilities used to generate the regridding weights have fairly significant memory requirements.  The NLDAS regridding example (and some of the others) should require significantly less memory to run.    

Thanks for raising this as an issue.  I'll see if we can add some notes about system requirements on the webpage.

Katelyn

[HAOWEN YUE]

Hi Katelyn

Thank you very much for the information. It makes sense to me.

Again, thank you for the help.

Best

Haowen Yue