From Confirmation to Rejection: Philosophical Insights into Hydrological Modeling

Código

I-EBHE0116

Autores

Iporã Brito Possantti

Tema

WG 1.13: Ensuring evidence-based findings

Resumo

This work reviews the philosophical foundations for the application of hydrological models, highlighting two main approaches: the confirmation and the rejection of theories. The first approach is related to scientific realism, which argues that scientific theories aim to describe reality. The second approach is related to instrumentalism, a perspective that does not deny the existence of reality but asserts that scientific theories can only describe the available empirical information and never the complete reality. The realist approach to confirmation is based on Bayesian epistemology, which uses inductive inference to probabilistically adjust hypotheses in the face of random noise in empirical observations. This process is carried out through the application of Bayes' Theorem, which incrementally adjusts the probability distribution of hypotheses, seeking an approximate truth based on empirical evidence. The confirmation of hypotheses in this context aims to establish the closest possible correspondence between models and observed reality, despite inherent limitations and uncertainties. On the other hand, the instrumentalist approach, advocated by Keith Beven within the realm of hydrology, emphasizes the rejection of theories and models. This perspective is grounded in Karl Popper's thesis, which argues that reliable knowledge is acquired through deductive refutation, where empirical evidence is used to test hypotheses based on counterexamples that prove their falsity. In this context, scientific theories must be falsifiable, allowing for their own rejection. Following this instrumentalist line, Beven proposes the testing of predictions? encapsulation by observational uncertainty as a criterion for rejecting hydrological models. Models that pass the test are considered empirically acceptable and equivalent but not necessarily true (the problem of equifinality). These two approaches reflect the complexity and nuance in the application of hydrological models. While confirmation seeks a gradual approximation of truth through empirical adequacy, instrumentalist rejection focuses on the practical utility of models and their ability to handle observational uncertainty. Reviewing the philosophical foundations provides a solid theoretical basis for the critical and informed application of hydrological models, highlighting the importance of understanding the limitations and potentialities of both approaches in evidence-based decision-making.