Minimum temperature limit creates issues for E3SM in aquaplanet mode

[whannah1]

I ran into this problem while trying to develop an RCE compset for E3SM. The valid temperature range for RRTMGP goes down to 180K, but for whatever reason when running in aqua planet the upper levels routinely see temperatures down to 150K.

For the time being I've enabled temperature clipping, but I'm not sure what the effect will be. It seems like a good idea to expand the look up tables to support these conditions.

[RobertPincus]

This has not turned out to be a high priority at AER but continues to stymie some folks. @szy21 and @RobertPincus have discussed the possibility of revising the data files by extending the band-integrated source term to lower temperatures (this is easily calculable) and leaving the within-band mapping to g-points fixed at values at the current lowest temperature. No timeline at this time.

[szy21]

I emailed @eli Mlawer but haven't heard from him yet. If I can get the source code and/or data for creating the band integral I would be happy to try extending the lookup table.