Following up on the results shared by Eemeli on the performance of the simplified chemistry scheme, I would like to share some old observations on the aqueous-phase sulfur chemistry reactions rates in the TM5 based chemistry version of the model (subroutine TM5_WETCHEM). I have posted these findings over three years ago on the TM5 portal (https://ci.tno.nl/gitlab/tm5/tm5-mp/-/work_items/13, for those who have access), and it seems they are still relevant for the chemistry in cycle 48r1. I copy the text below:
"Within CAMS the expressions for the aqueous-phase production of sulphate have been revised. While I think the new expression for the reaction with H2O2 is incorrect, also the current expression contains a typo.
The current expression is from Martin and Damschen (1981), with a temperature dependence from Nair and Peters (1989). However, the expression given in Table 2 of Nair and Peters uses a constant 3650. K in the exponential describing the temperature dependence, while in the code it is 3560. K.
An alternative, more recent expression for the rate of this reaction can be found in, e.g., Liu et al. (2020; https://doi.org/10.5194/acp-20-4153-2020)."
"At the lowest temperatures at which liquid droplets can exist (-38 degC), the relative error in the reaction rate due to this typo is +8.4%. At higher temperatures the error drops steadily, reaching 2.8% at 0 degC, crossing 0 at 298 K (~25 deg C) and reaching -2.3% at 50 degC."
"For future versions of the model I propose to use for the aqueous-phase reactions with both H2O2 and O3 the same rate expressions as in the IFS-AER version in which the oxidant fields are prescribed (see Table 2 in the attached report).
These are the expressions recommended by Seinfeld and Pandis (2006). They originate from the attached EPA report by Hoffmann and Calvert (1985). The activation energies in their expressions for reaction with O3 and H2O2, which govern the temperature dependence, are from Erickson et al. (1997) and McArdle and Hoffmann (1983), respectively.
The same expressions are used in several other papers, such as the study by Liu et al. mentioned above. Note that for O3 three reactions are included, whereas in the current code based on Maahs (1983) there are only two.
In general, the expressions from Hoffman and Calvert produce similar rates as the expressions in the current code. As mentioned by Martin and Damschen, their expression describes the data from pH 0 to 5, while cloud water pH generally varies between 3 and 7."
"For the record: the calculation of the reaction rate for the oxidation by O3 (following Maahs, 1983; attached) also looks suspicious (subroutine wetS in ebischeme.F90):
XSO2O3A=4.39e11*exp(-4131./temp)+2.56e3*exp(-966./temp) !S(IV)
XSO2O3B=2.56e3*exp(-966./temp)/hplus(i,j)
!divide by [H+]!S(IV)
It seems to me that the second term in XSO2O3A should be removed. Luckily, the resulting relative error in the rate is negligible, varying from 0.4% at -38 degC to 0.06% at 0 degC and decreasing further with temperature."
CAMS43_20217SC2_D43.1.1.1_201710_Sulphate.pdf
maahs1983.pdf
Hoffmann_Calvert_PB85198653.pdf
nair1989.pdf
robbinmartin1981.pdf
Following up on the results shared by Eemeli on the performance of the simplified chemistry scheme, I would like to share some old observations on the aqueous-phase sulfur chemistry reactions rates in the TM5 based chemistry version of the model (subroutine TM5_WETCHEM). I have posted these findings over three years ago on the TM5 portal (https://ci.tno.nl/gitlab/tm5/tm5-mp/-/work_items/13, for those who have access), and it seems they are still relevant for the chemistry in cycle 48r1. I copy the text below:
"Within CAMS the expressions for the aqueous-phase production of sulphate have been revised. While I think the new expression for the reaction with H2O2 is incorrect, also the current expression contains a typo.
The current expression is from Martin and Damschen (1981), with a temperature dependence from Nair and Peters (1989). However, the expression given in Table 2 of Nair and Peters uses a constant 3650. K in the exponential describing the temperature dependence, while in the code it is 3560. K.
An alternative, more recent expression for the rate of this reaction can be found in, e.g., Liu et al. (2020; https://doi.org/10.5194/acp-20-4153-2020)."
"At the lowest temperatures at which liquid droplets can exist (-38 degC), the relative error in the reaction rate due to this typo is +8.4%. At higher temperatures the error drops steadily, reaching 2.8% at 0 degC, crossing 0 at 298 K (~25 deg C) and reaching -2.3% at 50 degC."
"For future versions of the model I propose to use for the aqueous-phase reactions with both H2O2 and O3 the same rate expressions as in the IFS-AER version in which the oxidant fields are prescribed (see Table 2 in the attached report).
These are the expressions recommended by Seinfeld and Pandis (2006). They originate from the attached EPA report by Hoffmann and Calvert (1985). The activation energies in their expressions for reaction with O3 and H2O2, which govern the temperature dependence, are from Erickson et al. (1997) and McArdle and Hoffmann (1983), respectively.
The same expressions are used in several other papers, such as the study by Liu et al. mentioned above. Note that for O3 three reactions are included, whereas in the current code based on Maahs (1983) there are only two.
In general, the expressions from Hoffman and Calvert produce similar rates as the expressions in the current code. As mentioned by Martin and Damschen, their expression describes the data from pH 0 to 5, while cloud water pH generally varies between 3 and 7."
"For the record: the calculation of the reaction rate for the oxidation by O3 (following Maahs, 1983; attached) also looks suspicious (subroutine wetS in ebischeme.F90):
It seems to me that the second term in XSO2O3A should be removed. Luckily, the resulting relative error in the rate is negligible, varying from 0.4% at -38 degC to 0.06% at 0 degC and decreasing further with temperature."
CAMS43_20217SC2_D43.1.1.1_201710_Sulphate.pdf
maahs1983.pdf
Hoffmann_Calvert_PB85198653.pdf
nair1989.pdf
robbinmartin1981.pdf