Taming the Uncertainty Monster: Lessons from Astrochemistry

Marie Gueguen Marie G.

Date: Thursday, April 1, 2021
Time: 11:30 am EST
Duration: 30 minute presentation, followed by 30 minute question/answer period
Host: Martin Zelko
Presentation Style: powerpoint

Abstract:

Astrochemistry is a young discipline that started with the surprising detection of molecules in the interstellar medium in the 1940’s. This was a surprise for such a low temperature, low density environment, with constant exposure to ionizing radiation, seems too hostile to host molecules. Astrochemical models that rely on extrapolated reaction rate constants for networks of chemical reactions often fail to reproduce observations. The spectroscopic observations, performed by ground and space-based telescopes, that permit the detection of molecules in space and the physical conditions in astrophysical media are known to be both incomplete and uncertain, despite recent significant progress in telescopic resolution. Hence, astrochemistry is faced with the challenging task of evaluating models known to be incomplete against uncertain data. Yet, models are essential for identifying where better experimental data are needed, for improving and guiding future observational campaigns observations, as well as for motivating further theoretical development. Thus, astrochemists have no choice but to develop methods to assess their models, to learn from departures between models and observations, and to decide when disagreements between the former and the latter should lead us to question the fundamental assumptions of astrochemical models. In this talk, I will present some specifics of model evaluation in astrochemistry as well as methods recently developed to make the most out of a models’ expected departures from observations. This crucial task of model evaluation requires interdisciplinary expertise, and philosophers have a key role to play, both in advancing these methods and in fleshing out an adequacy-for-purpose view of models that stands up to such taxing empirical circumstances.

Zoom link: https://westernuniversity.zoom.us/j/92980218208