Full Results Accompanying A Machine Learning Examination of Hydroxyl Radical Differences Among Model Simulations for CCMI-1
Files
Related Publication Link
Date
Advisor
Related Publication Citation
DRUM DOI
Abstract
The hydroxyl radical (OH) plays critical roles within the troposphere, such as determining the lifetime of methane (CH4), yet is challenging to model due to its fast cycling and dependence on a multitude of sources and sinks. As a result, the reasons for variations in OH and the resulting methane lifetime, both between models and in time, are difficult to diagnose. We apply a neural network (NN) approach to address this issue within a group of models that participated in the Chemistry-Climate Model Initiative (CCMI). Analysis of the historical specified dynamics simulations performed for CCMI indicates that the primary drivers of methane lifetime differences among ten models are the flux of UV light to the troposphere (indicated by the photolysis frequency JO1D), the mixing ratio of tropospheric ozone (O3), the abundance of nitrogen oxides (NOx=NO+NO2), and details of the various chemical mechanisms that drive OH. Water vapor, carbon monoxide (CO), the ratio of NO:NOx, and formaldehyde (HCHO) explain moderate differences in methane lifetime, while isoprene, methane, the photolysis frequency of NO2 by visible light (JNO2), overhead ozone column, and temperature account for little-to-no model variation in methane lifetime. We also apply the NNs to analysis of temporal trends in OH from 1980 to 2015. All models that participated in the specified dynamics historical simulation for CCMI demonstrate a decline in methane lifetime during the analysed timeframe. The significant contributors to this trend, in order of importance, are tropospheric O3, JO1D, NOx, and H2O, with CO also causing substantial interannual variability in OH burden. Finally, the identified trends in methane lifetime are compared to calculated trends in the tropospheric mean OH concentration from previous work, based on analysis of observations. The comparison reveals a robust result for the effect of rising water vapor on OH and methane lifetime, imparting an increasing and decreasing trend of about 0.5 % per decade, respectively. The responses due to NOx, ozone column, and temperature are also in reasonably good agreement between the two studies.
Notes
Two tarred and gzipped files contain the full results from the CCMI OH analysis conducted by Nicely et al., 2020. "CCMI_NN_OH_analysis_plots_spec_dynamics_refc1sd_results.tar.gz" contains results from analyzing the specified dynamics simulations, REF-C1SD in CCMI. "CCMI_NN_OH_analysis_plots_free_running_refc1_results.tar" contains results from the free-running historical simulations, REF-C1. Both files house subdirectories: "Plots_of_inputs" (containing maps of each input variable at 850 hPa, where applicable, or total column quantity, in the case of "O3 COL"), "Swap_plots" (containing figures showing the change in tropospheric column OH resulting from the swap of the indicated variable from one model into the neural network (NN) of another model), "Swap_tables" (containing a budgeting of the changes in methane lifetime that result from all swaps between two indicated models' NNs), and "Aggregate_plots" (containing plots depicting methane lifetime versus month, mean change in methane lifetime due to all neural network swaps, and tables providing the values of the latter plot, for all models). The Specified Dynamics file additionally contains a subdirectory "Time_series_analysis," which consists of various plots analyzing the temporal changes in methane lifetime from all models, including time series, trends and interannual variability bar charts, multi-model means of the above, and comparisons of the multi-model mean trends to past literature. For further detail concerning the analysis conducted, refer to the journal article that accompanies this data set, Nicely et al., "A Machine Learning Examination of Hydroxyl Radical Differences Among Model Simulations for CCMI-1," ACP, 2020.