Elsevier

Planetary and Space Science

Volume 56, Issue 12, November 2008, Pages 1630-1643
Planetary and Space Science

Epistemic bimodality and kinetic hypersensitivity in photochemical models of Titan's atmosphere

https://doi.org/10.1016/j.pss.2008.05.016Get rights and content

Abstract

We show that photochemical models of Titan's atmosphere can give rise to bimodal distributions in the abundances of some major compounds, like C2H2 and C2H4. Sensitivity analysis enabled us to identify the causes and conditions of this bimodality. We propose several methods to control this behavior in photochemical models. In particular, we point out the importance of two key reactions and the needs for a critical evaluation of the kinetic data. We also show that the abundances of some compounds are hypersensitive to the ratio [CH4]/[H], suggesting that a time-dependent variation of this ratio might lead to a real bistability in the high atmosphere of Titan.

Introduction

The photochemistry of Titan is known to be very complex in the stratosphere and the ionosphere. The photolysis of methane (CH4) in the higher part of the atmosphere initiates an active hydrocarbon photochemistry. In the ionospheric region (above 1000 km), neutral hydrocarbons are dominated by CH4 and C2 compounds like acetylene (C2H2), ethylene (C2H4) and ethane (C2H6). Several observations are now available for these compounds from the Cassini mission (Shemansky et al., 2005, Waite et al., 2005, Yelle et al., 2006) and reanalysis of Voyager UVS observations (Vervack et al., 2004). In addition, the chemical scheme to produce C2 compounds is expected to be quite simple (see Hébrard et al., 2006 for a recent review). So, it should be possible, in principle, from the comparison of photochemical model results and observations, to extract some insight on the processes occurring in the higher atmosphere of Titan.

Hébrard et al. (2007) recently reported the first detailed analysis of the uncertainties carried by the reaction rate coefficients included in an up-to-date 1D photochemical model of Titan's atmosphere. They showed that the uncertainties on most of the computed abundances are much larger than the estimated uncertainties on abundances gathered from observations. This is the case even for basic hydrocarbons like CH4, C2H2, C2H4 and C2H6. Such large uncertainties preclude useful comparisons between observations and model results.

In the same work (Hébrard et al., 2007), the Monte Carlo sampling procedure used for uncertainty propagation produced outstanding abundance profiles of some compounds like CH4, C2H2 and C2H4. Considering the robustness of the algorithms used in this study, it is very likely that these features are not artifacts but result from specific zones in the space of input parameters (reaction rates) explored by Monte Carlo sampling. As indicated by sensitivity analysis, these neutrals play a major role in the large uncertainties observed in simulated ion mass spectra used to interpret Cassini INMS measurements (Carrasco et al., 2008). Again, this strongly weakens inferences about reaction scheme incompleteness or modeling biases. It is thus of paramount importance to identify the kinetic parameters at the origin of the outstanding profiles of neutral species.

In the present paper, we report the identification of the key parameters and the means we devised to control them in order to better constrain the photochemical models. The study focuses on C2H2 and C2H4 abundances in the high atmosphere of Titan (around 1200 km), predicted by 1D and 0D photochemical models. In Section 2, we briefly present results of the 1D model and we show the wide-ranging distributions of the abundances of these compounds. To better understand the behavior of the 1D model, we developed a representative 0D kinetic model. Results of this model are presented in Section 3. The sensitivity analysis used to identify the key parameters and derived modified Monte Carlo sampling strategies are presented in Section 4. Implications of this study for the Titan photochemistry are discussed in Section 5.

Section snippets

The model

Rate coefficients ki, photodissociation coefficients Ji and uncertainty factors Fi for each reaction i used in the present study are given in Hébrard et al. (2006). It is to be noted that for many processes, in the absence of experimental data at temperatures representative of Titan's atmosphere, reaction rates and uncertainties have to be extrapolated. This can lead to very large uncertainty factors (see Section 5).

Uncertainty propagation studies with large uncertainties on input parameters

The model

We have extracted from the 1D photochemical model of Hébrard et al. (2006) two different chemical schemes, using the same set of reactions, corresponding to two representative altitudes: 1200 km and 900 km. The rates change because the density and temperature are different. The chemical schemes are restricted to hydrocarbons and contain 74 compounds, 50 photodissociation processes and 351 reactions. We use a Monte Carlo procedure similar to Dobrijevic and Parisot (1998) to obtain distributions of

Tracking and tackling bimodality

Bimodality in the PDF of the concentration of a chemical species reveals that the system explores at least two different chemical regimes. This is a problem when there is experimental or observational evidence that the studied system should indeed explore only one of the regimes. In this case, it is necessary to identify the reactions implied in the “bifurcation” and see if their parameters can be constrained to provide the suitable chemical regime.

As stated earlier, it is expected from

Comments on the key reactions Ra and Rb

The relative importance of each reaction obtained from a sensitivity analysis is, in part, due to the value of the uncertainty factor assigned to these rates: a drastic overestimation might exaggerate the relative importance of some reactions. In any case, sensitivity analysis pinpoints reactions to be scrutinized in priority for a reevaluation of their uncertainty.

The best way to estimate each uncertainty factor, taking into account the technical constraints of different experiments and the

Conclusions

The study of the propagation of uncertainties carried by the reaction rate coefficients in 1D photochemical model of Titan's atmosphere shows outstanding distributions in the abundances of some major compounds like CH4, C2H2 and C2H4. In order to understand the origin of these profiles, we used a 0D kinetic model to overcome the problem of computational time. This model reveals bimodal densities at 1200 km and 900 km for C2H2 and C2H4. A sensitivity analysis was used to ascertain the chemical

Acknowledgments

We acknowledge the support received from the Centre National d’Etudes Spatiales (CNES) through postdoctoral positions for N.C. and E.H. This work was also supported by the Programme National de Planétologie (PNP).

References (43)

  • E. Hébrard et al.

    Photochemical kinetics uncertainties in modeling Titan's atmosphere: first consequences

    Planet. Space Sci.

    (2007)
  • D.F. Strobel et al.

    Titan's upper atmosphere—structure and ultraviolet emissions

    Icarus

    (1992)
  • R.J. Vervack et al.

    New perspectives on Titan's upper atmosphere from a reanalysis of the Voyager 1 UVS solar occultations

    Icarus

    (2004)
  • R.V. Yelle et al.

    The vertical structure of Titan's upper atmosphere from Cassini ion neutral mass spectrometer measurements

    Icarus

    (2006)
  • S.M. Anderson et al.

    Fast-flow studies of methylidyne radical kinetics at 290 K

    Phys. Chem.

    (1987)
  • D. Baulch et al.

    Evaluated kinetic data for combustion modeling: supplement II

    J. Phys. Chem. Ref. Data

    (2005)
  • D. Baulch et al.

    Evaluated kinetic data for combustion modeling

    J. Phys. Chem. Ref. Data

    (1992)
  • D. Baulch et al.

    Evaluated kinetic data for combustion modeling: supplement I

    J. Phys. Chem. Ref. Data

    (1994)
  • M.A. Blitz et al.

    Reaction of CH radicals with methane isotopomers

    J. Chem. Soc., Faraday Trans.

    (1997)
  • M.W. Bosnali et al.

    Studies with pulsed-radiolysis of the reaction of CH(2Π) with methane and other substances

    Z. Naturforsh. A

    (1971)
  • W. Braun et al.

    Flash photolysis of methane in vacuum ultraviolet. 2. Absolute rate constants for reactions of CH with methane hydrogen and nitrogen

    J. Chem. Phys.

    (1967)
  • Cited by (0)

    View full text