Lava flows are well-studied natural phenomena, yet their behaviour remains not fully understood. Given lava’s significant impact on people, infrastructure, and agriculture, it is important to explore further lava flow dynamics. Lava flows encounter a variety of topographic features as they move down the flanks of volcanoes. This thesis employs both numerical and experimental methods to model lava flow using analogues. We developed numerical simulations of lava flow behaviour based on the shallow water equations. In addition, we explored the flow of analogue fluids with varying rheological properties, over and around different topographies. We developed and implemented new methods and novel techniques to measure flow properties suitable to inform inversion models to infer underlying rheology.

We compared a simple force balance model of the channelised flow of different rheologies around a bend to a shallow water equation model to analyse the effects of inertial displacement and the development of the wetted contact line between the channel wall and the fluid. This study extends our understanding of the force balance model by quantifying the effect of the sector angle of the bend’s influence on the flow behaviour and how the rheology and density of the flow, subtly affect the development of the wetted contact lines.

Our analogue experiments show that in the case of wax with temperature-dependent Newtonian rheology flowing over channel-spanning weirs, we discovered that basal cooling influences the apparent thickness of the fluid as it approaches a weir. The expected thickness of the flow differed from the thickness predicted by the Nusselt equations, due to the wax freezing on the channel base. Our novel methodology showed that the internal temperature profile for a cooling wax flow was quadratic across the thickness of the flow.

We completed analogue experiments that examined the flow of Herschel-Bulkley fluids as they navigated past surface-piercing occlusions, observing the impact of geometry on the flow. This study produced a dataset that can inform an inversion model to infer the flow’s rheology, a method tested and validated in Muchiri et al. (2024a).

We finally show the implications of our results providing examples from recent eruptions that produce similar responses to flow in various topographic scenarios. In the future, these responses could be quantified and compared to inversion models similar to those developed in association with this work to provide real-time rheological measurement of flowing lava.