Crashing on European tile 16 with infinity values for o3_bg coming from subroutine uEMEP_nonlocal_NO2_O3

[brucerd]

Crash output for European tile 16:

Calculating chemistry for NO2 (uEMEP_chemistry)

Photochemistry with time scale used
nox_bg no2_bg o3_bg nox_loc f_no2_loc J_photo temperature time_scale
1.29E-02 0.00E+00 Infinity 4.42E-04 1.00E-01 1.23E-04 2.86E+01 6.69E+02
f_no2 BB b b**2-4.*c
NaN NaN Infinity NaN
1

o3_bg from subroutine uEMEP_nonlocal_NO2_O3 is infinity. Should check inputs and outputs to this routine first

This is likely due to edge effects, as the model is searching outside of the EMEP grid for EMEP grids. Could be caused in a number of places prior to this call, including the crossreference routines and the EMEP grid interpolation routines. These you find in uEMEP_subgrid_EMEP and uEMEP_subgrid_meteo_EMEP. Current flag to interpolation is EMEP_grid_interpolation_flag.eq.6, so line 360 of uEMEP_subgrid_EMEP.

Also suspicious is the annual mean temperature of 28.6 C. Bit too high for this tile. temperatures are also interpolated to the integration subgrids for meteo_subgrid(). These are interpolated in uEMEP_subgrid_meteo_EMEP, line 120. There are no checks in this routine for out of bounds EMEP index values.

[eamousing]

As far as I understand, this error is ultimately caused by the uEMEP grid only being partially within the EMEP grid right? Then the solution is to make a check for overlap and exit the program earlier if this is the case? @brucerd