Atmosphere energy budget is not zero

[fgibbon]

What version of the code are you using?
CESM2.1.5

Describe your problem or question:
I'm running an out of the box aquaplanet (QPC5) with the default prescribed SSTs for ten years. I want to calculate the atmosphere energy budget to check spin-up, so I calculated the difference between net surface flux and net top-of-model flux:
Fnet_sfc = FSNS - FLNS - LHFLX - SHFLX
Fnet_tom = FSNT - FLNT
atm_budget = Fnet_tom - Fnet_sfc

I expected the weighted global mean of atm_budget to be zero or close to zero, but it seems to fluctuate around 1.5 W/m^2 and I'm not sure why this is. (I have checked the weighting.) I think the model is spun up as individual flux variables (e.g. FSNT shown for 2 years) seem to jump quickly and then plateau.
Any insight would be greatly appreciated.

 

[brianpm]

I suspect that the residual is from not accounting for the surface energy flux of frozen precipitation. I think you need to include (PRECSC+PRECSL)*rho_water*L_fusion in your surface budget.

 

[fgibbon]

I suspect that the residual is from not accounting for the surface energy flux of frozen precipitation. I think you need to include (PRECSC+PRECSL)*rho_water*L_fusion in your surface budget.

Thank you so much, that has solved it!