Using TAVG in METGRID for small lakes but not Great Lakes - RESOLVED

[GlennDoodle]

Hello,

I am trying to run a simulation from back in 1983 on finer resolution grids (3km and 1km). The issue that I am having is that the finest resolution SST data that I have found is 0.25deg x 0.25deg. Using the default METGRID.TBL, I get a very blocky (not surprisingly) pattern in the center of the Great Lakes from the supplied SST and then the METGRID does an extrapolation of the rest of the lake temperatures toward the shoreline of the Great Lakes. I modified the METGRID.TBL so that the SST was spread out to the lake coastlines by changing
name=SST
interp_option=sixteen_pt+four_pt
fill_missing=0.
missing_value=-1.E30
flag_in_output=FLAG_SST

to
name=SST
interp_option=sixteen_pt+four_pt+wt_average_4pt+wt_average_16pt+search
fill_missing=0.
missing_value=-1.E30
flag_in_output=FLAG_SST

This worked for the Great Lakes, but the unresolved lakes (e.g. Finger Lakes in New York) became much warmer as did many lakes in Canada due to the extrapolation. So, I tried the TAVG method, grabbing a weeks worth of data to make a TAVG file to reduce my "hot lake" issue for the unresolved lakes. While I prefer the values supplied to the small unresolved lakes, this method overwrote the SST values over the Great Lakes and thus I lost some information over the Great Lakes (the TAVG reduced my Great Lake temperatures by up to 4K). So, my question is, can I use the TAVG method but mask out the Great Lakes so that I they do use the SST values which were resolved in the ERA5 data set? Is this an option in METGRID? Would I need to redefine the landuse values over the Great Lakes to be the same as ocean points instead of the special lake category that is used by METGRID.

For completeness sake the METGRID.TBL entry is
name=SKINTEMP
mpas_name=skintemp
interp_option=sixteen_pt+four_pt+wt_average_4pt+wt_average_16pt+search
masked=both
interp_land_mask = LANDSEA(1)
interp_water_mask = LANDSEA(0)
fill_missing=0.


Thanks for your time!

 

[GlennDoodle]

My next attempt was to redefine the landuse values over the Great Lakes to be the same as ocean points instead of the special lake category that is used by METGRID. I wrote a script to change the LU_INDEX values from 21 to 17 over the Great Lakes (and the update geo_em.d0X.nc files showed that the LU_INDEX was changed. Unfortunately, it appears that the Great Lakes are still being treated as lakes because my model output is showing skin temperatures that reflect the TAVG file. I am trying to figure out if I am missing a step (or variable to change).

 

[GlennDoodle]

Apparently I am using this forum to think out loud! My method worked as soon as I also adjusted the LANDUSEF field as well!

 

[kwerner]

Hi,
I first would like to apologize for the long delay in response. I've been traveling for the past few weeks and am just now catching up on the forum inquiries. I am very happy to hear, however, that you were able to solve the problem. Thank you for updating the post!