Hello. I am trying to run AMET_v1.5 but there was no script for post only.
I cannot use pre-and-post script because it is so slow with the large CMAQ files.
Is it possible to submit the job with pre-post script or should I just run in the server?
or Will AMET_v1.4 script be compatible with AMET_v1.5 version?
The pre-and-post script provided with AMETv1.5 includes all the functions of the post-only script provided with AMETv1.4, but with the ability to also run combine (note that AMETv1.4 included a similar pre-and-post script as in AMETv1.5). So, if you’re just trying to accomplish the functions of the post-only script from AMETv1.4, you can do that using the pre_and_post script that is provided with AMETv1.5. You can simply use the flags to turn on-and-off the functions you want the script to perform.
The post-only script in AMETv1.4 starts after CMAQ has been run and the combine files have been created. To recreate the same function with the pre-and-post script provided with AMETv1.5, you can simply set the RUN_COMBINE and RUN_HR2DAY flags to F. Setting those to F assumes that you’ve already created the combine files required. You can then set the WRITE_SITEX, RUN_SITEX, CREATE_PROJECT, and LOAD_SITEX flags to T. That will run the site compare and database loading portions of AMET.
Let me know if that helps. And if I may ask, what function is so slow for you? Is it the creation of the combine files? Ultimately, you weren’t be able to skip anything using the post-only script that you need to do for the pre-and-post script. The post-only script was just an older script that I deprecated. So, let me know what functions are very slow for you. I removed the post-only script from AMETv1.5 because it really was a duplicate of the functions available in the pre-and-post script, and with the ability to set the flags, you can fully mimic the functions of the post-only script.
Wyat
Imagine that we are uisng CMAQ just a month hourly file.
For combine, it usually takes 15 mins with 1 layer.
For setting up database in SQL, it usually takes more than an hour with default networks.
For analysis with batch script, it also takes more than an hour with default batch settings.
Database and analysis with one script is convenient if you use your own personal or super computer.
However, when we run the script in server, the server automatically disconnected the connect with users after an hour without any commands.
We are using personal SQL server so I think we cannot submit the job to the server as well.
It might be possible by turning off all plotting options for only setting up database. (not sure)
Can you add some descriptions with updates in here for other users?
I think one script is a good idea for people who are familiar with AMET but it can be another huddle for new users. + if I do not combine with this script, the location is quite complicated and needed to modify with lots of effort.
Plus, AMET 15 does not process with different numbers of layer files.
With AMET14 script, it is working.
Hi Darby,
in your last post, are you specifically asking for feedback on the error you appear to be encountering when running combine as part of your CMAQ post-processing, either stand-alone or incorporated into the AMETv1.5 pre_and_post script? When you say that things were working with the AMET1.4 script, this probably doesn’t refer to running combine, because based on Wyat’s comments above, running combine wasn’t integrated into the AMETv1.4 script.
Prior to running combine as part of the AMET1.5 script, did you ever successfully run combine stand-alone for your case? The error message you show, as well as the grid spacing implied by your file names, suggest that you’re running into the same issue you encountered last year:
Back then, you updated the error checking in the combine code to work for your case. If running combine as part of the AMETv1.5 script, you need to make sure that the AMET script points to the same combine executable you created back then and that (I think) worked for you offline. And as was the case back then, the error isn’t caused by the different number of layers in ACONC and METCRO3D, but rather by non-robust grid consistency checking in combine for “odd” grid resolutions like 1.333 km.
Oh, you are right. I re-compile the combine so it is working.