On the upgrading of the modified Carbon Bond Mechanism IV for use in global Chemistry Transport Models

The modified Carbon Bond Mechanism 4 (Houweling et al., 1998) has been updated to include the most recent recommendations concerning both the input parameters used for calculation of both photolysis rates (J values) and the chemical reaction rates (Sander et al., 2006; Atkinson et al., 2006). First we present the results of a box model study that investigates the differences between the chemistry of TM4 and TM5. An inconsistency is introduced as a result of differences in the definition of the photolysis rate of methylglyoxal (CH3C(O)CHO). The most significant effects were found to occur for urban (polluted) scenarios throughout all seasons. In summary, the most important updates were: (i) a retuning of the photolysis rate for CH3C(O)CHO, (ii) the 30% increase in the photo-dissociation rate of formaldehyde (HCHO), (iii) changes to the reaction rates involving organic peroxy-radicals and (iv) an enhanced formation of ORGNTR. Applying these updates resulted in a decrease in the availability of reactive nitrogen which influences the production efficiency of ozone. When incorporating an operational set of these updates into the 3D global CTM TM4 it was found that both increasing the JMGLY value and introducing tracer transport for CH3C(O)CHO significantly reduced the surface concentrations over industrial regions by up to 80%. For the other trace gases there were generally decreases in tropospheric [OH], [HO2], [O3], [PAN], [HNO3] and [HCHO], with associated increases in [ORGNTR] and [CO]. In general, the oxidising capacity in the model atmosphere decreases as a result of the lower global production rate of OH. Comparisons made against selected ozonesonde profiles shows that TM4 generally over predicts surface ozone in the tropics whilst underestimating surface ozone in remote locations. For the upper troposphere TM4 generally under predicts tropospheric ozone indicating that the transport and/or release of reactive nitrogen is insufficient. For surface CO there are generally improvements in the correlation between values simulated in TM4 and a host of CMDL measurement sites. The atmospheric lifetimes of both CH4 and O3 become ~9.3 years and 23.6 days, respectively, where both these values are in the 1-sigma variability of the multi-model ensemble mean given in Stevenson et al (2006) for the IPCC 2000 simulations.