Is it scientifically acceptable to start CMAQ-ISAM every two days with initial conditions from a global model?

[Hungnt]

Dear everyone,

I am running CMAQ-ISAM v5.5 for India. I am wondering whether it is acceptable to start a new CMAQ-ISAM run every two days using ICO from a global model. Specifically, I use ICO/BCO generated from WACCM, and I restart the model every two days rather than using CGRID files (i.e., running the first day with ICO from WACCM, the second day with the CGRID file from the first day, and the third day with ICO from WACCM).

Alternatively, should I always restart CMAQ-ISAM using only CGRID files after the initial start?

I have noticed that when I start a new run every two days, the contribution of ICO to total PM2.5 becomes very high. Therefore, I am unsure whether this approach is acceptable. I would appreciate your opinion on this.

Thank you.

[cjcoats]

No.

The purpose of running a finer-grid model is to be able to resolve features and chemical behavior that are not possible with the coarser grid of a global model. This is made worse by the fact that the default ICON and BCON processors used to create CMAQ initial and boundary conditions from coarser-grid files do not even interpolate: they just use the values for whatever coarse-grid cell the CMAQ-grid cell-center fits into – something that is completely at odds with the “high order algorithm” approach CMAQ claims to use – and does (mostly) use elsewhere.