The TOMCAT global chemical transport model: Description of chemical mechanism and model evaluation

The TOMCAT 3-D chemical transport model has been updated with the emissions and chemical degradation of ethene, propene, toluene, butane and monoterpenes. The full tropospheric chemical mechanism is described and the model is evaluated against a range of surface, balloon, aircraft and satellite measurements. The model is generally able to capture the main spatial and seasonal features of high and low concentrations of carbon monoxide (CO), ozone (O3), volatile organic compounds (VOCs) and reactive nitrogen. However, model biases are found, some of which are common to chemistry models and some that are specific to TOMCAT and warrant further investigation. Simulated O3 is found to generally lie within the range of ozonesonde observations and shows good agreement with surface sites. The most notable exceptions to this are during winter at high latitudes, when O3 is underestimated, and during summer over North America, when O3 is overestimated. Global Ozone Monitoring Experiment-2 (GOME-2) comparisons suggest that TOMCAT sub-column tropospheric O3 in DJF may also be underestimated outside of the Arctic, particularly near tropical regions. TOMCAT CO is negatively biased during winter and spring in the Northern Hemisphere (NH) when compared to ground-based observations and MOPITT (Measurements Of Pollution In The Troposphere) satellite data. In contrast, CO is positively biased throughout the year in the Southern Hemisphere (SH). The negative bias in the NH is a common feature in chemistry models and TOMCAT lies well within the range of biases found in other models, while the TOMCAT SH positive bias is at the upper range of positive biases reported in other models. Using two simulations with different boundary conditions highlighted the sensitivity of model performance to the chosen emission dataset when simulating VOCs, nitrogen oxides (NOx) and peroxyacetyl nitrate (PAN). VOC measurements show winter/spring negative biases in C2-C3 alkanes and alkenes, which is likely driven by underestimated anthropogenic emissions. TOMCAT is able to capture the seasonal minima and maxima of PAN and HNO3. However, comparisons to an aircraft climatology show that PAN may be overestimated in winter and HNO3 may be overestimated in winter and spring in regions over North America. The model showed different biases in NOx, depending on location, with evidence of underestimated Asian emissions contributing to negative model biases over China and underestimated fire emissions contributing to negative biases in the SH. TOMCAT global mean tropospheric hydroxyl radical (OH) concentrations are higher than estimates inferred from observations of methyl chloroform, but similar to, or lower than, multi-model mean concentrations reported in recent model intercomparison studies. TOMCAT shows peak OH concentrations in the tropical lower troposphere, unlike other models, which show peak concentrations in the tropical upper troposphere. This is likely to affect the lifetime and transport of important trace gases and warrants further investigation.