Behind the rhythms of free surface thermal signatures: A tale of two lengths scales in a submerged buoyant jet

Thermally scanned free surface flows appear striking — rhythmic, convoluted motions folding and fading as they evolve. Yet more striking is what they conceal: two length scales within the same flow, each behaving differently. This talk explores the utility of thermal imagery to uncover patterns within the thermal free-surface expressions of a submerged buoyant jet. Such a flow can be found in the meltwater plumes of marine-terminating glaciers, or industrial outfalls in the ocean. In our idealized experiments, the jet was sourced through nozzles with varying diameter, depth beneath the free surface, and flow rates. Thermal signatures were captured using a long-wave infrared camera. To filter measurement noise, the noise floor was estimated using the power spectra of the raw thermal images. The images were then reconstructed using orthogonally decomposed modes required to capture the cumulative energy corresponding to the pure part of the measured signal. These de-noised thermal fields were then systematically processed for thermal pattern tracking via a particle image velocimetry (PIV) algorithm to extract free-surface velocity fields. A correlation length scale analysis of the thermal fields reveal a consistent decrease in scalar (temperature)-based integral length scales with increasing source-based Reynolds number. In contrast, velocity fields counterintuitively show that momentum-based integral length scales remain “locked” to the flow geometry, independent of the source flow rates. A physical interpretation of these findings is developed by considering: (i) the arrest of vertical motions by inertial effects within the source layer of the free surface interface, diverting flow radially, and (ii) the parallel cascades of thermal and kinetic energy from large scales down to Batchelor and Kolmogorov scales, respectively. Overall, this presentation demonstrates how thermal imaging, combined with spectral and modal filtering as well as free surface physics, offers a potential remote sensing method that can replace perilous in-situ measurements.