Subspace-based dimension reduction for chemical kinetics applications with epistemic uncertainty

TL;DR

This paper introduces a subspace-based dimension reduction method for uncertainty quantification in complex chemical kinetics, demonstrating its effectiveness in identifying dominant variability directions with minimal parameters.

Contribution

The work applies active subspace analysis to chemical reaction networks, revealing significant dimension reduction and linking it with sensitivity analysis for better understanding of key parameters.

Findings

A 1-dimensional active subspace was identified in both 19- and 33-parameter cases.

Active subspace analysis effectively captures dominant sources of variability.

Links between active subspaces and global sensitivity analysis were established.

Abstract

We focus on an efficient approach for quantification of uncertainty in complex chemical reaction networks with a large number of uncertain parameters. Parameter dimension reduction is accomplished by computing an active subspace that predominantly captures the variability in the quantity of interest (QoI). In the present work, we compute the active subspace for a H2/O2 mechanism that involves 19 chemical reactions, using an efficient iterative strategy. The active subspace is first computed for a 19-parameter problem wherein only the uncertainty in the pre-exponents of the individual reaction rates is considered. This is followed by the analysis of a 33-dimensional case wherein the activation energies are also considered uncertain. In both cases, a 1-dimensional active subspace is identified, which indicates enormous potential for efficient statistical analysis of complex chemical systems. In addition, we explore links between active subspaces and global sensitivity analysis, and exploit these links for identification of key contributors to the variability in the model response.