PROPER ORTHOGONAL DECOMPOSITION ANALYSIS OF SPATIOTEMPORAL TEMPERATURE PATTERNS

The proper orthogonal decomposition (POD) technique decomposes a spatiotemporal signal into time-independent orthogonal spatial components and time-dependent amplitudes, a splitting that proves very useful in the analysis of intricate patterns. The technique is shown here to be useful in elucidating complex patterns observed in chemical systems. Its power and advantages are illustrated by analysis of two time-dependent temperature patterns measured by an infrared camera during the oxidation of hydrogen on a nickel disk in a mixed flow reactor. The POD analysis reveals that in both cases the surface temperature variations are induced by two ''pacemakers'', active regions at the edge of the disk, from which temperature fronts propagate. The synchronization and interaction between the two pacemakers determine the nature of the oscillations and of the propagating temperature fronts.