CLM-CN / SIMGM spin up for deep, unconfined aquifer

[cory_pettijohn@oregonstate_edu]

I am running CLM-CN in offline point mode (driven by tower data using Stoeckli et al.'s Modelfarm) on a linux machine here at Oregon State. Currently I am focusing on the semi-arid Metolius Intermediate Ponderosa Pine Ameriflux site (US-ME2) in east-central Oregon.

I am not sure of the best way to spin up the SIMGM groundwater component of the model given the following regional hydrogeological conditions: (1) regional unconfined groundwater depth ~ 75 m below surface in complex volcanic & sedimentary facies; (2) local water well data indicating ~ 3-4 m soil depth, underlain by previously mentioned volcanic materials, including tuff & fractured basalt.

Do I prescribe the initial depth to groundwater before my ~800 yr. spin up for the CN component, or is the point of SIMGM to solve for the depth to this new 'aquifer'? Should I perform a separate hydro. spin-up prior to spinning up CN, considering Niu et al.'s (2007) suggestion of "following the definition of equilibrium described by Yang et al. [1995], the model needs to spin-up for at least 250 years to reach an equilibrium water table depth in arid regions (e.g., Sahara)..." (p. 5, Sec 3.1)? I am just concerned because after my 800 yr+ spin up, the equilibrium depth to groundwater solution = 5.7 m. Maybe I am taking SIMGM's definition of an aquifer too literally???

Thanks so much in advance for any input!
Cheers,
Cory

 

[slevis]

I do think that you are taking clm's below-soil water storage too strictly as an aquifer when, really, it's just another water reservoir. It's true that the closest such reservoir in nature may be the aquifer and that comparing to aquifer data should make sense, but don't expect agreement most of the time given the complexities that determine aquifer depth that the clm neglects.

Sam