Its u-cj051 and the ticket is #6592 with the branch name branches/dev/hannahbryant/vn11.1_vn11.1_h2_noaa_field_emissions
There is this line in the ukca_chemistry_ctl: !*RESHAPE(y(:,nn_h2),(/row_length,rows/))*volume(:,:,k)*1.0e6/avogadro
where I was including the conversions factors seen in the asad_flux but the output was not correct. It is commented out but could be helpful!
Using instead the outputted grid cell volume and mass mixing ratio (converted to molc cm-3) as volume and y respectively gives a result more similar to the output deposition but not identical. In most regions it is off by a factor of 2-3 but in some areas the difference is larger.
I think I may have a clue regarding the x3 issue: I ran the model for 1 hour and outputted both depositions on TALLTS. The 3d output is once every hour, however, the 2d surface is output every 20 minutes but the first two of these is zero and the third is identical to the 3d output. The hourly mean of the 2d surface is therefore 1/3 of the 3d output!
There is still the decay with altitude to understand and the difference to the output deposition but hopefully this explains the 2d vs 3d plots.
Ah! You’ve found it. The time code for the 2D fields will be for every timestep - I guess you copied it from an emissions one. For deposition this needs to be 17 as this means output it on UKCA timesteps (every hour!).
See Appendix C of UMDP C04 (page 23)
Great thank you! I’m still not convinced on the difference to the real profile. For example, ozone has a real dry deposition of:
And the rate from the 2d output is found using this calculation:
o3_ppm=o3_mass_mixing_ratio*(28.9644 / 47.9982)1e6
Which gives this output (on the same scale):
This has the same general shape but has discrepancies over some regions (eg. himilayas are ~3x too high). I’m wondering if this still has something to do with timesteps (eg. when the deposition is calculated and whether this is because I am multiplying together the monthly mean values?).