Update E3SM-Project submodule

[xylar]

This merge updates the E3SM-Project submodule from 86f7e8b to 894b5b2.

This update includes the following MPAS-Ocean and MPAS-Frameworks PRs (check mark indicates bit-for-bit with previous PR in the list):

• (ocn) fix for submesoscale parametrization in standalone E3SM-Project/E3SM#5892 - 13 BFB checks fail in global ocean runs because this PR fixed the submesoscale parameterizaiton and 1 execution failure because of the issue fixed in Fix ISOMIP+ plotting #705
• (ocn) Accumulate data land-ice mass fluxes E3SM-Project/E3SM#5910 - 2 BFB checks fail because of the issue fixed in Update variables checked in DISMF performance test #694 and Fix parabolic bowl registry for MPAS-Ocean init mode E3SM-Project/E3SM#5930, 1 execution failure because of the issue fixed in Fix ISOMIP+ plotting #705
• (ocn) Fix parabolic bowl registry for MPAS-Ocean init mode E3SM-Project/E3SM#5930 - 2 BFB failures and 1 execution failure as above. Although this issue should fix the BFB failure, the baseline from the previous commit still has the issue.

[xylar]

I'm seeing large, non-bit-for-bit changes between E3SM-Project/E3SM#5857 and E3SM-Project/E3SM#5910. I need to find out why, since neither of these PRs explains the changes.

[xylar]

The non BFB changes mentioned above were from E3SM-Project/E3SM#5892 and are expected because that PR fixed the submesoscale parameterization for standalone runs.

All non-BFB changes are accounted for so it seems safe (and like a very good idea) to advance the submodule.

[xylar]

I'm holding off on this until I can finish making mesh files for #706, which is in progress. That PR uses the MPAS-Ocean build in my main branch, so I don't want to update it with non-BFB changes (especially accidentally) until that's done.