James Stewart

Where to put a barrier?

And that's where engineers like Horn come in, and thank you so much to her, she helped us massively in researching this video.

With a background in hydrology and hydrodynamics, Horn's expertise was in modelling floods and river ice, both slow and fast-moving destructive flows shaped by gravity and terrain.

Yeah, those fundamental principles, it turned out, could be applied to lava.

By using flood simulation software originally designed by the US Navy, she was able to quickly predict, with remarkable accuracy, where molten rock was most likely to go next.

Teams rushed to build the first experimental barriers using whatever they could find in situ, literally anything, mud, gravel, soil from past eruptions.

Crews worked around the clock and they were close enough to feel the burn from the encroaching lava.

This is serious stuff.

Now the aim here was not to stop the lava outright, but to delay it, restrict it, and essentially buy time.

These early tests show that this approach could actually work, but they also revealed some critical lessons.

Firstly, lava type matters, and it matters a lot.

A'a lava is chunky and rough.

When this hit a barrier, it thickened and piled up like a geological porridge.

So if you want to stop it, your walls need to be built like a medieval fortress.

Pahoehoe lava, on the other hand, is smooth and much runnier, behaving more like honey, if we're doing the breakfast metaphor thing.

Interestingly, drone surveys showed that it flowed over dams, but could be redirected using angled burns.

And it was this piece of information that immediately suggested that a barrier doesn't have to stop lava flow, it just needs to guide it.

These initial field trials made up as they were going along led to the most ambitious lava defence project ever attempted.

And as volcanic activity shifted further west, closer to towns, the lessons became more critical.

Multi-kilometre barriers up to 24 metres high have now been built across the Requienes Peninsula.