The Open-Source Photochem Code: A General Chemical and Climate Model for Interpreting (Exo)Planet Observations

Abstract

With the launch of the James Webb Space Telescope, we are firmly in the era of exoplanet atmosphere characterization. Understanding exoplanet spectra requires atmospheric chemical and climate models that span the diversity of planetary atmospheres. Here, we present a more general chemical and climate model of planetary atmospheres. Specifically, we introduce the open-source, one-dimensional photochemical and climate code Photochem, and benchmark the model against the observed compositions and climates of Venus, Earth, Mars, Jupiter and Titan with a single set of kinetics, thermodynamics and opacities. We also model the chemistry of the hot Jupiter exoplanet WASP-39b. All simulations are open-source and reproducible. To first order, Photochem broadly reproduces the gas-phase chemistry and pressure-temperature profiles of all six planets. The largest model-data discrepancies are found in Venus's sulfur chemistry, motivating future experimental work on sulfur kinetics and spacecraft missions to Venus. We also find that clouds and hazes are important for the energy balance of Venus, Earth, Mars and Titan, and that accurately predicting aerosols with Photochem is challenging. Finally, we benchmark Photochem against the popular VULCAN and HELIOS photochemistry and climate models, finding excellent agreement for the same inputs; we also find that Photochem simulates atmospheres 2 to 100 time more efficiently. These results show that Photochem provides a comparatively general description of atmospheric chemistry and physics that can be leveraged to study Solar System worlds or interpret telescope observations of exoplanets.

Figures (13)

• Figure 1: A schematic pressure-temperature profile to aid explanation of our climate model. On the right-hand-side, $j$ indexes each atmospheric layer, and the column labeled "Convecting with below" indicates if a given $j$ layer is convecting with the layer below it. Note that layer $j = 5$ is convecting with the layer above it ($j = 6$), but not with the layer below it ($j = 4$), so it is labeled "No" in the "Convecting with below" column. A new atmospheric grid $k$ can be defined that groups the convecting layers, simplifying the iteration to radiative-convective equilibrium (see text).
• Figure 2: Photochemical simulations of Venus's atmosphere as a closed system (lines) compared to observations (black dots). Table \ref{['tab:obs']} lists the sources for all observations. The solid lines are the nominal model, while the dashed lines are a sensitivity test that adds additional chlorine chemistry as described in Section \ref{['sec:venus_chem']}. The dotted lines are the thermochemical equilibrium of each atmospheric layer in the nominal model. Panel (k) is the H$_2$SO$_4$ cloud mixing ratio computed with $f_{\mathrm{H_2OSO_4}(l)} = n_{\mathrm{H_2OSO_4}(l)}/n$, where $n_{\mathrm{H_2OSO_4}(l)}$ is the molecules cm$^{-3}$ in the condensed phase and $n$ is the total gas phase number density. Overall, our nominal model reproduces the observed concentrations of CO, HCl, H$_2$SO$_4$, H$_2$S, and SO at all altitudes, but fails to explain the measured abundances of SO$_2$, H$_2$O, OCS, S$_3$ and S$_4$. Most of these model-data discrepancies are long standing unsolved problems in Venus photochemistry Rimmer2021Bierson2020.
• Figure 3: Climate simulations of Venus's atmosphere. (a) A climate model that uses the H$_2$SO$_4$ clouds and gas concentrations predicted by the nominal photochemical simulation in Figure \ref{['fig:venus']}. (b) The same as (a), except the model uses observed cloud optical properties from Crisp1986. (c) The same as (b), except the simulation uses Venus's observed SO$_2$ profile. In (a), (b) and (c), the top panels shows predicted temperature profiles compared to the Venus International Reference Atmosphere Seiff1985. Thick portions of the predicted P-T profile indicate convecting layers. The bottom panels are the net thermal and solar fluxes compare to the solar fluxes measured by Pioneer Venus Tomasko1980. Our climate model can reproduce Venus's energy balance (panel (c)) when we correct for inaccurate clouds and SO$_2$ predicted by the photochemical model.
• Figure 4: Photochemical simulation of modern Earth's atmosphere (lines) compared to observations (dots). Table \ref{['tab:obs']} lists the sources for all observations. For observations that do not give an error bar, we assume a one order-of-magnitude error to account for diurnal and spatial variations following Tsai2021. Photochem broadly reproduces the observed composition of Earth's atmosphere.
• Figure 5: Climate simulations of modern Earth's atmosphere with 193 and 500 DU of ozone (black and grey lines, respectively) compared to temperature profiles at the equator and $45^{\circ}$ N (red and blue lines, respectively) in January from the COSPAR International Reference Atmosphere 1986 (CIRA-86). Thick portions of the predicted P-T profiles indicate convection. Our model broadly reproduces Earth's pressure-temperature profile.