Giuseppe Bilotta<p>So, I mentioned already that we cannot really model <a href="https://fediscience.org/tags/lava" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>lava</span></a> flows. The main reasons for that is that we don't actually know how lava behaves, at least not in sufficient detail.</p><p>Of course, lava is a fluid, and a (very) viscous one at that, so we know that it follows the Navier–Stokes equations. We also know that its behavior is heavily dependent on temperature, so we know that we also need the heat equation, with both kinds of boundary conditions (conduction to ground, and radiation on the free surface).</p><p>And that's all we know. Seriously.</p><p>OK, not really, but everything else is extremely uncertain. When modeling a viscous fluid (like lava, or any other geophysical flow for the matter), the first thing you need to know is what the viscosity is. And for lava, we don't know. There's a lot of things we do know, but not enough.<br>For example, we know that the viscosity depends on temperature, chemical composition, degree of crystalization, amount and types of volatiles in the melt, and so on and so forth. But we don't exactly know the laws relating the viscosity to all of these chemical and physical properties.</p><p>2/</p><p><a href="https://fediscience.org/tags/NavierStokes" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>NavierStokes</span></a> <a href="https://fediscience.org/tags/NavierStokesEquations" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>NavierStokesEquations</span></a></p>