I’m looking for some advice about greenhouse gas concentrations in the UM (vn13.2). Specifically, I’m trying to work out where the model reads CO2, CH4 and N2O concentrations from. From what I’ve found so far, they can either be read as mass mixing ratios (kg/kg) from the Gas MMRs page in the Radiation section of the UM Science Settings or from an RCPDAT file which defines concentrations as ppm/ppb. Reading of this file is controlled by i_ukca_scenario in the TraceGases page of Section 34 in UM Science Settings.
Test runs suggest that CO2 is being read from the Gas MMRs and that CH4 and N2O are being read from the RCPDAT file because the CO2 concentration in the model output is not the same as that in the RCPDAT file. However, this hybrid approach doesn’t seem possible given the current settings.
I have two principal questions:
Can you explain where CO2, CH4 and N2O concentrations are read from within the model?
Can you explain how the model converts from concentration in ppm/ppb to mass mixing ratio in kg/kg?There is some discrepancy between the two values for CO2 using a traditional conversion as follows:
CO2 [kg/kg] = CO2 [ppm] x ( 44.01 / 29.966 ) / 1e6
The suite I am currently working with is u-dd727, a copy of a UKESM1.1 piClim-Control suite.
Apologies if these are not questions that you can answer here but any advice on the subject would be much appreciated.
I’m afriad that my knowledge of CO2 isn’t as good as for CH4 & N2O - my understanding is that CO2 concentrations are defined in Rose, either in clim_fcg_levls_co2 (for time-varying concentrations) or in co2_mmr (for fixed concentrations). Values are already in kg/kg and are valid for 1st January for the year specified in clim_fcg_years_co2 (for time-varying) and then interpolated linearly between values. I’m not sure of the conversion factor used within the radiation scheme to convert from mass-mixing ratio to mole fraction (if this is ever done). These values are valid for the whole atmosphere.
For CH4 and N2O - when UKCA is on, the usual practice, as is used in UKESM (and as you have discovered) is to specify the lower boundary condition in mole fraction using the file defined in ukca_rcpdir and ukca_rcpfile when i_ukca_scenario=2. The units for each gas, e.g. ppm, ppt etc., are given at the top of each file, which will have many columns. CH4 and N2O are chemical tracers in UKCA, and their 3D fields are then read by the radiation scheme. The values specified are valid for the middle of the year, i.e. 1st July, and the values are linearly interpolated to each timestep when they are used. UKCA converts these to the necessary units used within the code, and the tracers themselves are in kg/kg. The conversion factors used can be found in the ukca_constants.F90 module, and are:
c_ch4 = 0.5523
c_n2o = 1.5188
where the calculation done is
vmr*c_species = mmr
These values are used to define the target surface concentration of these species, and then psuedo-emission fields are created online that are used to bring these tracers back to the target values using the UM’s TR_MIX routine.
UKCA also uses CO2 sometimes to create a constant field used in the chemical solver. When it does the conversion factor used is
Thanks for your reply. I was able to find out much the same by scouring the code myself but I wasn’t aware of the pseudo-emission fields or the TR_MIX routine. I’ve now set the GHG concentrations to the desired levels for my experiment.