,
6 tweets,
2 min read
Read on Twitter
We have to flip the top layer of soil by shovel periodically to keep it loose like the upper layer of real lunar soil. On the Moon, impact cratering flips the soil, keeping it loose. We used to think the top millimeter gets flipped every 100,000 yrs, the top centimeter slower... 
2/ but that was an estimate based on the influx of micrometeorites and larger impactors from space. It neglected the secondary impacts. Secondary impacts are the ejecta that get thrown out from another impact. Those also hit the soil and flip it over. When you account for that,
3/ it turns out the upper centimeter gets flipped every 80,000 years or so. Down to 30 cm gets flipped over millions of years because the bigger meteorites that can flip that deeply are more rare, but they do arrive at a known rate and this predicts the flipping-depth rate.
4/ The Apollo missions measured how compacted the soil is at different depths. It is very random as you go down, because each layer of soil in the column is a random impact ejecta blanket flipped out from different sized impacts at different distances.
5/ However, if you plot all the data onto the same graph (all locations on the Moon aggregated together) you can fit a smooth curve onto it, and it shows that on average the soil compacts with depth (like an exponential decay) from very fluffy on top to very dense down deep.
6/ This matches the expectations from it being flipped over more often at shallow depths. So the top 30 cm is pretty loose, and the top several cm is very very fluffy and loose. This is what we try to simulate in our robotics competition.
