...
Similar to E3SMv2 with pg2 but respecting the conventions in Mapping file algorithms and naming convention and using TempestRemap (or mbtempest) for bilinear maps.
- fluxes: FV→FV aave for nearly all cases
- if source grid resolution >> target grid resolution: FV→FV aave
- if source grid resolution <= target grid resolution: V3 is considering adopting TR 2nd-order with non-linear correction ("nlmap").
- state and vector maps:
- ocn→atm: currently this map must be the same as used for fluxes (FV→FV aave)
- This restriction is because the range of the state and vector maps must match the range of the flux map.
- The "nlmap" approach should be used for this map as well (not sure if this was tested in time for V3?).
- lnd↔atm and atm→ocn
- if source grid resolution >> target grid resolution (fine-to-coarse): "FV→FV aave"
- if source grid resolution <= target grid resolution (nearly equivalent resolutions): Use TR bilin
- if source grid resolution << target grid resolution (coarse-to-fine): TR 2nd-order with non-linear correction ("nlmap"). Needed for atm→lnd in trigrid.
- Why do we still recommend aave or bilin above? intbilin or "nlmap" approach should be better in both cases.
- riv → ocn
- continue to use custom maps generated by E3SM/cime/tools/mapping/gen_mapping_files/runoff_to_ocn
- ocn→atm: currently this map must be the same as used for fluxes (FV→FV aave)
...