I think there’s different ways: - take random samples from the area and statistically come up with that value - put a radar/sonar/ xray etc in a hole and get an outline of how big the deposit is.

I did not work in this field explicitly, but worked on Gaussian processes which were first used as a way to aggregate data like this from multiple sources to minimize the number of drillings required to find oil.

not a dumb question, as seen by the fact that the media/government minister etc. have blown this story out of proportion.

In reality its most likely 15M tonnes inferrred resource of 0.4% lithium content So much smaller amount of actual lithium.

They work this out by doing core drilling, and then inferring/extrapolating the size/shape of the orebody boundaries. (Im a mining engineer)