Bottom end of the range: 10^7Kg per year would be about 27.5 tonnes per day. The article references another piece[1] where they quote 54t per day, a similar num

Bottom end of the range: 10^7Kg per year would be about 27.5 tonnes per day. The article references another piece[1] where they quote 54t per day, a similar number. There's then an estimate made of how much of that is Aluminium - 1%. I guess the author ran the 54t / 100 calc and came up with 0.45 instead of 0.54 ... a simple typo. The key point is most of the natural stuff doesn't hang around up there. Top end (10^9 Kg) of the range we get about 2750t per day natural origin. And at 1% for Aluminium that's 27.5t per day. That number is not far off the 29t estimated daily burn for the full Starlink v2 constellation they quote. Also: I notice that the Lodders article focussed specifically on Aluminium as it had been proposed as a method for geoengineering albedo. There are other metals in the natural origin material, and I don't see why we should exclude those from affecting the Van Allen belts - is it only Aluminium that's bad? Comparing the Starlink burn to the Van Allen material mass seems irrelevant when there's perhaps a few hundred tonnes per day of naturally arriving metals around too. It's not obvious to me this topic can be waved away as not an issue though. [1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8137964/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8137964/