Data, scripts and simulations for ProxyOH-[OH] analysis using ATom data and F0AM and AM3 simulations

This dataset provides the simulations and analysis code used in Baublitz et al., An observation-based, reduced-form model for oxidation in the remote marine troposphere, Proceedings of the National Academy of Sciences, 120. Code, package versions The code is generally written in Python and saved to a Jupyter Notebook (.ipynb) format, except for the component developing the Bayesian regressions, which is written in R. For improved accessibility, the code has also been printed to PDF format so that it may be readable without requiring access to Jupyter. The code used to create the main text figures is specified in the file names. When the primary focus of a script is to create supplemental figures, the figure names have also been specified in the script file name. The code for creating other supplemental figures is also available in the script corresponding to the section where that figure is referenced. The following packages and package versions were used to develop this analysis: Python (v3.10.0) anaconda 4.13.0 collections (native to anaconda installation) datetime os random jupyter 1.0.0, jupyter-core 4.9.1 matplotlib (visualization) 3.5.1 notebook 6.4.6 numpy 1.21.4 pandas 1.3.4 scipy 1.7.3 seaborn (figure formatting) 0.11.2 The full environment is specified in the YAML file "atom_env.yml." Anaconda users (not tested, potentially restricted to Windows) may load this environment with this file and the following command: $ conda env create -f atom_env.yml R (v4.1.2, includes package parallel) tidyverse 1.3.1 rjags 4-12 runjags 2.2.0-3 lattice 0.20-45 lme4 1.1-31 loo 2.4.1 ggpubr 0.4.0 matrixStats 0.61.0 Zipped directory contents The full set of global, hourly AM3 model simulations developed for this project are included in this repository (AM3_hourly_simulations_global_ATom1-4.zip) for reference and potential future application, though they are not used in the code. They are described here (vs listed) and span the dates for each campaign leg and are broken into four variable categories, concentrations and met fields ('stp_conc_v2'), individual reaction rates ('ind_rate'), integrated reaction rates ('all_rate') and deposition velocities or photolysis rates ('dep_jval'). Some of these files include all days in the range, while others include only the days that the campaign took measurements. In addition, a subset of the AM3 simulations that specifically include variables used in the manuscript analysis that have been sampled along the ATom flight is included, along with the 10 s ATom merge data (AM3_model_simulations_sampled.zip). This is the file that should be downloaded for reproducing the manuscript in the analysis. AM3_model_simulations_sampled.zip atom1_10s_ss_030122.csv atom2_10s_ss_030122.csv atom3_10s_ss_030122.csv atom4_10s_ss_030122.csv bayes_data.zip bayes_atom_10s_model_122022.csv bayes_ats_10s_remNOlsth2sigma_highlogNO_emulate_122022.csv bayes_ats_10s_remNOlsth2sigma_highlogNO_emulate_allPOH_030723.csv base/ .Rhistory atom_jags_010723.R atom_lmer_model_122122.R atom_sens_030723.R dat1_bins.csv dat1_OH.csv gelman_list_base.csv levels.csv log_pd.csv model_b0.csv model_b1.csv p.fit.csv p.mu.csv p.sd.csv r_prx_ytrue.pkl rjmt_B0.csv rjmt_B1.csv rjmt_proxy.csv rjmt_y_true.csv CH4_CO_HCHO_MHP/ .Rhistory atom_altCH4_CO_HCHO_MHP_031423.R atom_jags_altCH4_CO_HCHO_MHP_031423.R dat1_bins.csv dat1_OH.csv dat1_proxy_ch4_co_hcho_mhp.csv levels.csv log_pd.csv p.fit.csv p.mu.csv p.sd.csv r_prx_ytrue.pkl rjmt_B0.csv rjmt_B1.csv rjmt_proxy_ch4_co_hcho_mhp.csv rjmt_y_true.csv CO_HCHO/ .Rhistory atom_altCO_HCHO_031423.R atom_jags_altCO_HCHO_031423.R dat1_bins.csv dat1_OH.csv dat1_proxy_CO_HCHO.csv levels.csv r_prx_ytrue.pkl rjmt_B0.csv rjmt_B1.csv rjmt_proxy_co_hcho.csv rjmt_y_true.csv CO_HCHO_MHP/ atom_altCO_HCHO_MHP_031423.R atom_jags_altCO_HCHO_MHP_031423.R dat1_bins.csv dat1_OH.csv dat1_proxy_CO_HCHO_MHP.csv levels.csv r_prx_ytrue.pkl rjmt_B0.csv rjmt_B1.csv rjmt_proxy_CO_HCHO_MHP.csv rjmt_y_true.csv H2O2_O3_CH4_CO_HCHO_MHP/ .Rhistory atom_altH2O2_O2_CH4_CO_HCHO_MHP_122122.R atom_jags_altH2O2_O3_CH4_CO_HCHO_MHP_030723.R dat1_bins.csv dat1_OH.csv dat1_proxy_h2o2_o3_ch4_co_hcho_mhp.csv levels.csv r_prx_ytrue.pkl rjmt_B0.csv rjmt_B1.csv rjmt_proxy_h2o2_o3_ch4_co_hcho_mhp.csv rjmt_y_true.csv HCHO/ atom_altHCHO_031423.R atom_jags_altHCHO_031423.R dat1_bins.csv dat1_OH.csv dat1_proxy_HCHO.csv levels.csv r_prx_ytrue.pkl rjmt_B0.csv rjmt_B1.csv rjmt_proxy_HCHO.csv rjmt_y_true.csv MHP/ atom_altMHP_122122.R atom_jags_altMHP_122122.R dat1_bins.csv dat1_OH.csv dat1_proxy_MHP.csv levels.csv r_prx_ytrue.pkl rjmt_B0.csv rjmt_B1.csv rjmt_proxy_MHP.csv rjmt_y_true.csv F0AMv3.2.zip mean_ratio_OH_loss_bins_oce.npy mean_ratio_OH_prod_bins_oce.npy mean_ratio_OH_prod_loss_bins_oce.npy Data/ atom1/ atom1_output_alt.cs atom1_output_CO.csv atom1_output_H2O.csv atom1_output_lat.csv atom1_output_lon.csv atom1_output_lossOH_ppt_lump15.csv atom1_output_M.csv atom1_output_NO.csv atom1_output_OH.csv atom1_output_prodOH_ppt_lump15.csv atom1_output_startTime.csv atom1_output_sza.csv atom2/ atom2_output_alt.csv atom2_output_CO.csv atom2_output_H2O.csv atom2_output_lat.csv atom2_output_lon.csv atom2_output_lossOH_ppt_lump15.csv atom2_output_M.csv atom2_output_NO.csv atom2_output_OH.csv atom2_output_prodOH_ppt_lump15.csv atom2_output_startTime.csv atom2_output_sza.csv atom3/ atom3_output_alt.csv atom3_output_CO.csv atom3_output_H2O.csv atom3_output_lat.csv atom3_output_lon.csv atom3_output_lossOH_ppt_lump15.csv atom3_output_M.csv atom3_output_NO.csv atom3_output_OH.csv atom3_output_prodOH_ppt_lump15.csv atom3_output_startTime.csv atom3_output_sza.csv atom4/ atom4_output_alt.cs atom4_output_CO.csv atom4_output_H2O.csv atom4_output_lat.cs atom4_output_lon.cs atom4_output_lossOH_ppt_lump15.csv atom4_output_M.csv atom4_output_NO.csv atom4_output_OH.csv atom4_output_prodOH_ppt_lump15.csv atom4_output_startTime.csv atom4_output_sza.csv For any further questions on the model simulations or code included here, please contact the corresponding author (Colleen Baublitz, cbb2158@columbia.edu).The development of this dataset was supported by the ATom investigation under NASA's Earth Venture (NNX15AG58A) and ACMAP (NNX17AG40G) programs. Contributions by GMW and JMN were also supported by the NOAA Atmospheric Chemistry, Carbon Cycle and Climate program through grants NA19OAR4310164 and NA19OAR4310165, respectively. SML acknowledges support from the NASA FINESST program (80NSSC19K1301)