Brief summary of bug

frac_sno appears to always be 0 for lake columns. This appears to cause incorrect results in various other lake variables.

General bug information

CTSM version you are using: ctsm1.0.dev055

Does this bug cause significantly incorrect results in the model's science? Yes

Configurations affected: All

Details of bug

I initially noticed this issue by examining LakeHydrology and seeing that frac_sno is never set there. I then confirmed that these three out-of-the-box initial conditions files have frac_sno identically 0 for all lake columns: clmi.B1850.0161-01-01.0.9x1.25_gx1v7_simyr1850_c190111.nc, clmi.I1850Clm50BgcCrop-ciso.1366-01-01.0.9x1.25_gx1v7_simyr1850_c190116.nc and clmi.I2000Clm50BgcCrop.2011-01-01.1.9x2.5_gx1v7_gl4_simyr2000_c180715.nc.

I then added this code (which isn't exactly what we want in the end) in order to see if these incorrect frac_sno values for lake have any effect on the simulation:

       if (snow_depth(c) > 0.0_r8) then
          frac_sno(c) = 1._r8
       else
          frac_sno(c) = 0._r8
       end if

This led to answer changes in many variables in a short test run. I have not yet investigated whether these differences are significant scientifically.

@swensosc thinks we should use the same snow cover fraction method for lakes as is used for non-lakes. I will plan to make that change. However, this will warrant a closer investigation to see the impact of this change.