Chemical modeling and data inversion in presence of uncertainty on model and data

Details Chemical modeling and data inversion in presence of uncertainty on model and data Chemical modeling and data inversion in presence of uncertainty on model and data

Sep 2008

adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=2008epsc.conf..344p&link_type=abstract

European Planetary Science Congress 2008, Proceedings of the conference held 21-25 September, 2008 in Münster, Germany. Online a

Statistics

Computation

Scientific paper

Recent articles and reviews have insisted on the need to consider various sources of uncertainty which put strong limits on the prediction ability of photochemical models of Titan's atmosphere [14]. The central fact is that very few reaction rates have been measured in conditions relevant to Titan, and that extrapolation from the available measurements to those conditions is far from being straightforward [5]. We are presently in a state where the prediction uncertainty of mole fractions of minor components of Titan's atmosphere is notably larger than their measurement uncertainty, for instance by Cassini's INMS. In this context, any data inversion based on a chemical model has to deal with this prediction uncertainty in order to obtain reliable confidence intervals on the optimized parameters. Handling of these uncertainties in modeling introduces an additional layer of complexity, but it is unavoidable, and efficient tools have to be used to ensure consistency and minimize the computational overcost. Bayesian data analysis provides a rigorous framework to handle these problems, as well for uncertainty propagation as for data inversion [6]. We will introduce the present methods and issues in uncertainty description by probability density functions, uncertainty propagation, sensitivity analysis and stochastic optimization, relevant to Titan modeling. This approach will be illustrated on the simulation of Cassini's INMS ion mass spectra and on the recovery of densities of neutral species in Titan's ionosphere by inversion of these spectra with an equilibrium ionospheric model.

Affiliated with

Also associated with

No associations

Rating

Chemical modeling and data inversion in presence of uncertainty on model and data does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.

If you have personal experience with Chemical modeling and data inversion in presence of uncertainty on model and data, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Chemical modeling and data inversion in presence of uncertainty on model and data will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFWR-SCP-O-1793520