Validación de observaciones y de simulaciones astrofísicas: un enfoque epistemológico.

Un enfoque epistemológico

Abstract

The classical concept of observation in contemporary astrophysics has undergone considerable changes, mainly due to certain observations carried out with computational simulations. Taking into account how this notion has evolved in the last decades, different philosophical reflections about observability will be analyzed. Here, topics such as the theoretical-observational dichotomy, the realism-antirealism debate and those current treatments closely related to scientific practices, alluding to the distinction between observable and unobservable entities, will be considered. In this article, and through a case study, the role of external validation will be analyzed in this field. Different mechanisms of data production and their eventual reduction in astroinformatics will be studied, considering data acquired both through traditional or observational ways or through simulation models. In relation to this point, the proposal offered by Reiss (2019) will be developed, examining the arguments for and against the external validation.Thus, current validation strategies in different data models will be identified, establishing the appropriate links for a quantitative and qualitative assessment of observations and simulations. From the interplay between observations and simulations, the status and configuration of the available evidence and its dependence on the design of current observations will be shown. It will be concluded that the interaction between observation and simulation models not only contributes to the development of heuristic guides, but also allows profitable feedback in the external validation processes of both models.