You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Update User's Guide: Running CLM5.0 with irrigation #1640
The "Running with irrigation" section of the User's Guide contains incorrect instructions for CLM5.0. As described by @olyson in this DiscussCESM forum reply, the proper way to turn on irrigation is to set irrigate = .true. in the user_nl_clm namelist.
My suggestion is to have three separate subsections here, one for each of 4.0, 4.5, and 5.0.
Relevant to #509. Once that's implemented, the instructions for 5.0 will need to specify that irrigation is actually on by default for crop runs, but it can be turned off via the namelist.
I don't actually think it makes sense to maintain instructions for old versions of the code. In that case, this file would need to be changed to describe how to turn OFF irrigation in BgcCrop runs. At that point, do we even really need this file? Adding "next" to discuss.
Good idea on discussing this. It was needed when the behavior was different for clm4_0 and clm4_5/clm5_0. And it was needed when there was a special option to build-namelist to turn it on. Now with it being a simple namelist item it's no different than any other namelist item and hence may not be necessary.
That said irrigation is off except for Bgc transient historical cases for clm5_0 or clm6_0. But, it can be turned on for SP cases for example. It's also off for CMIP6 SSP future scenarios,, So the defaults for when it's off or on aren't necessarily obvious, so it might be good to have this simple example of showing how to control it.
Okay, well that's something we really need to talk about, because I think it should be ON by default for future scenarios! And probably for year-specific cases (e.g. 1850 and 2000) as well, although I can see why we'd want it off in 1850 for spinup purposes. But the "principle of least surprise" would suggest that if crops are on, irrigation should be on. (And since we have an irrigated generic crop, it should be on in non-Crop compsets too!)
Regardless of what we decide on default behaviors for different compsets, the User's Guide should provide instructions for (1) checking what a case is set up to do and (2) toggling it.
this seems like a broader discussion what we want to support for CESM3 and CMIP7. @samsrabin will you create an issue to iron this out in how we're defining defaults and compsets? Then our documentation should reflect these decisions.
Activity
samsrabin commentedon May 8, 2026
I don't actually think it makes sense to maintain instructions for old versions of the code. In that case, this file would need to be changed to describe how to turn OFF irrigation in
BgcCropruns. At that point, do we even really need this file? Adding "next" to discuss.ekluzek commentedon May 9, 2026
Good idea on discussing this. It was needed when the behavior was different for clm4_0 and clm4_5/clm5_0. And it was needed when there was a special option to build-namelist to turn it on. Now with it being a simple namelist item it's no different than any other namelist item and hence may not be necessary.
That said irrigation is off except for Bgc transient historical cases for clm5_0 or clm6_0. But, it can be turned on for SP cases for example. It's also off for CMIP6 SSP future scenarios,, So the defaults for when it's off or on aren't necessarily obvious, so it might be good to have this simple example of showing how to control it.
samsrabin commentedon May 10, 2026
Okay, well that's something we really need to talk about, because I think it should be ON by default for future scenarios! And probably for year-specific cases (e.g. 1850 and 2000) as well, although I can see why we'd want it off in 1850 for spinup purposes. But the "principle of least surprise" would suggest that if crops are on, irrigation should be on. (And since we have an irrigated generic crop, it should be on in non-
Cropcompsets too!)samsrabin commentedon May 10, 2026
Regardless of what we decide on default behaviors for different compsets, the User's Guide should provide instructions for (1) checking what a case is set up to do and (2) toggling it.
In addition to
irrigate, also explain:irrig_min_laiirrig_start_timeirrig_lengthirrig_target_smpirrig_depthirrig_threshold_fractionirrig_river_volume_thresholdlimit_irrigation_if_rof_enableduse_groundwater_irrigationirrig_method_defaultsamsrabin commentedon May 11, 2026
Also, we should document the various irrigation history outputs here.
wwieder commentedon May 28, 2026
this seems like a broader discussion what we want to support for CESM3 and CMIP7. @samsrabin will you create an issue to iron this out in how we're defining defaults and compsets? Then our documentation should reflect these decisions.
samsrabin commentedon May 28, 2026
Blocked until we resolve that (#4056).