I'm setting up a few CESM runs where I'd like to nudge the CAM6 state towards a control simulation, following the instructions here:
I had originally planned on nudging all of U, V, T, and Q outside of a specific region, but the documentation has this warning against nudging Q above the surface: "Surface Nudging of Q: Since the nudging tendencies are applied separately from the convective parameterizations, nudging Q values in the interior of the model can lead to misleading results. Particularly in precipitation values. On the other hand, nudging Q at the surface layer is an effective proxy for surface fluxes of water vapor. The following settings for the vertical window illustrate how to nudge only at the surface for a 32 level model."
However, the instructions for the vertical window seem to apply that vertical window to ALL the nudged fields (U, V, and T as well as to Q) - and I definitely want at the very least my U and V fields to be nudged throughout the troposphere. So, is my best bet to only nudge U and V? I did a test run nudging T and not Q but that seems like I'm setting myself up for a situation where Q is always trying to "catch up" to the nudged T. Maybe that is fine, maybe it would be problematic.
Does anyone with experience in nudging CAM have any thoughts on only nudging U and V vs. also nudging T, or nudging T and Q (but keeping the U & V nudging across multiple vertical levels)?
Bonus question: can I have a more complex nudging mask where I don't nudge a 3-D box in lat/lon/lev space? (i.e. nudge all vertical levels in most regions, and only nudge vertical levels above some threshold in another region?)
I had originally planned on nudging all of U, V, T, and Q outside of a specific region, but the documentation has this warning against nudging Q above the surface: "Surface Nudging of Q: Since the nudging tendencies are applied separately from the convective parameterizations, nudging Q values in the interior of the model can lead to misleading results. Particularly in precipitation values. On the other hand, nudging Q at the surface layer is an effective proxy for surface fluxes of water vapor. The following settings for the vertical window illustrate how to nudge only at the surface for a 32 level model."
However, the instructions for the vertical window seem to apply that vertical window to ALL the nudged fields (U, V, and T as well as to Q) - and I definitely want at the very least my U and V fields to be nudged throughout the troposphere. So, is my best bet to only nudge U and V? I did a test run nudging T and not Q but that seems like I'm setting myself up for a situation where Q is always trying to "catch up" to the nudged T. Maybe that is fine, maybe it would be problematic.
Does anyone with experience in nudging CAM have any thoughts on only nudging U and V vs. also nudging T, or nudging T and Q (but keeping the U & V nudging across multiple vertical levels)?
Bonus question: can I have a more complex nudging mask where I don't nudge a 3-D box in lat/lon/lev space? (i.e. nudge all vertical levels in most regions, and only nudge vertical levels above some threshold in another region?)