Removing compounding factors on u2/v2 boundary conditions in ADC

[katsmith133]

Happy to do this PR, but just want to make sure which one we want to remove...
config_adc_up2_vp2_factor is only used on u2 and v2:

MPAS-Model/src/core_ocean/shared/mpas_ocn_adcReconstruct.F

Lines 273 to 274 in 5dc8d6e

	u2(k,1,iCell) = config_adc_up2_vp2_factor*config_adc_bc_const*frictionVelocity**2.0
	v2(k,1,iCell) = config_adc_up2_vp2_factor*config_adc_bc_const*frictionVelocity**2.0

config_adc_bc_const is used on u2, v2, tp2, and sp2:

MPAS-Model/src/core_ocean/shared/mpas_ocn_adcReconstruct.F

Lines 330 to 331 in 5dc8d6e

	tp2 = 0.4_RKIND * config_adc_bc_const * (wt(i1,1,iCell) / frictionVelocity)**2
	sp2 = 0.4_RKIND * config_adc_bc_const * (ws(i1,1,iCell) / frictionVelocity)**2

Do we want separate factors for u2/v2 and tp2/sp2?

[BrodiePearson]

If we are following the Andre et al. (1978) paper that Luke linked to in the other thread we should use config_adc_bc_const for both, as the second-order tracer moments should have the same coefficient as u2 and v2 vs. $u_{star}^2$ (but not $w_{star}^2$). But...

• That paper doesn't have a coefficient of 0.4 out the front of tp2 and sp2. (Should we remove it?)
• That paper relates tp2 and sp2 to $u_{star}^2$ (sfcFrictionVelocitySquared ) rather than frictionVelocity, which is a combination of $u_{star}^2$ and $w_{star}^2$. (Should we change that to tp2 = config_adc_bc_const * (wt(i1,1,iCell) / sqrt(sfcFrictionVelocitySquared))**2 etc?)
• At the moment tp2 and sp2 at the surface only affect the splat parameterization but don't affect the model profiles of these properties (t2 and s2). Should we rename them t2 and s2 so they feed into the simulated profiles?

[katsmith133]

@BrodiePearson

• Andre et al. (1978) does not have this extra coefficient, but the CLUBB implementation that Luke pulled it from does. Not sure where that 0.4 comes from for CLUBB. Perhaps its atmospheric related (?).
• Again, Andre et al. (1978) does not include the wstar^2 in the tp2 and sp2 term, but the CLUBB implementation does. And again, not sure what the motivation was.
• I tested this change and there appears to be almost no difference in the results for the cooling case with splatting turned on (both with the w2 BC on and off). Only a very small (~O(-4)) change in the surface value of t2 occurs and that appears to not affect much else.