I'm not sure this actually provides the right perspective, in the sense of a useful intuition. Taking anything from the part of the Earth we live in (essentially, the troposphere plus the upper part of the crust) and visualizing it as a sphere will make it look small, because the troposphere + upper crust is a fairly thin shell, so doesn't take up a lot of volume when visualized as a solid sphere next to the earth.
I disagree. It is a useful perspective because it flies in the face of one's intuited perspective. How many of us would have estimated that the volume of water would be so small? We visualize the oceans as really deep gouges on the Earth's surface because to our puny asses a depth of 1 to 8 miles sounds amazingly deep. We'd never guess that even with miles deep depressions and miles high mountains, the Earth is smoother than a billiard ball:
Combining both of these counter-intuitive perspectives tells us that our oceans are akin to a very thin film of water on a wet billiard ball. A thin dirty film is all that billions of bacteria needs to thrive on a billiard ball, and a thin dirty film is all that billions of animals and plants need to thrive on Earth.
> How many of us would have estimated that the volume of water would be so small?
Well, I did, but that's because it's already been emphasized to me before (in some kind of geology class) how little volume the crust+troposphere shell takes up. It's not a water-specific thing, but a crust/troposphere thing, of which water is a subset. And actually I think the real underlying counterintuition is just a geometric one, that people don't realize little volume spherical shells occupy relative to solid spheres.
"How many of us would have estimated that the volume of water would be so small?"
I did. Before reading the article, I did the following very rough calculation in my head: earth is 40,000 km (equator length) by 20,0000 km (pole to pole). An average water depth of 1 km gives 800M cubic kilometers of water. From there, I guesstimated the diameter at about 1000km (a 1000km cube would be 1000M cubic kilometers). It might have ended up smaller if I hadn't approximated the surface of the earth to be a cylinder, but who cares about such a 'puny' rounding error, given that I totally guessed at that 1 km?
This gets at my reaction as well. It might be useful to have spheres of some of the precious metals (silver, gold, etc.) next to the water sphere to help set the frame of reference. I'm guessing they'll be much, much smaller than the water sphere.
At the least, this depiction jars against the common idea that water covers ~66% of the earth's _surface_. The mapping to this presentation is, presumably, the fact that surface of the earth is exceedingly thin with respect to its radius. Thus _delirium's point, that the absolute volume isn't as important as the manner of distribution in relation to other materials on the earth. The common adage that a human is 70% water doesn't, by itself, give a useful description of what it means to _be_ human.
> It might be useful to have spheres of some of the precious metals (silver, gold, etc.) next to the water sphere to help set the frame of reference.
I'm not sure how it compares to the total gold potentially available, but from what I can find, only about 8500 m^3 of gold has been mined in all of human history, which would make a completely invisible sphere on the scale of this image--- radius a bit over 50 m.
Starting with the earth's crust [1]: gold is one of the absolute rarest elements there -- at 72nd out of 78 [a] at between 1-4 parts per 10^9 (mass). The volume of the crust is on the order of 10^10 km^3 [2][b], with a density around 3 g/cm^3 [2], for a mass of gold on the order of 10^11 tonnes. Being extremely dense at 19 g/cm^3 [3], this would fill a compact sphere on the order of 1 km radius.
The mass of the earth [4] is about 200 times more than the crust -- almost 10^22 tonnes -- but the elemental composition of the core is unknown and very different from the crust. It's likely to be highly enriched in gold relative to the crust, because of gravitational separation (heavy metals sink down). Extrapolation from crustal abundance would give 10^13 tonnes, whereas one geologic estimate [5] gives 10^15 tonnes in the core alone -- a sphere 30 km in radius.
[a] excluding short-lived unstable elements and noble gases
[b] roughly, ~40% of the earth's surface (~5*10^8 km^2) is continental crust ~50 km thick
That is not a completely reliable measure of abundance when it comes to human purposes, after all it needs to be economical to extract these resources.
After all metals like lithium and most of the rest of the "alkali metals" and "rare earth metals" are relatively common, this issue is natural processes do not collect and concentrate them, as they are much more reactive and easily dispersed, so significant commercial deposits are rare compared to its crust concentration percentages.
Gold is a metal that naturally concentrates into deposits after aeons of time, as it is heavy and relatively non reactive, so there are many more commercially viable deposits than other metals with significantly higher abundance.
Indeed. The chemical properties of the elements will make some easier to extract than others. Metals that have been known since prehistoric times (copper, tin, lead, silver, zink and others) easily combine with sulphur in the crust, and can be found as easily extractable ores, while the rare earths are as abundant as these, but do not concentrate much in specific ores. The same goes for Uranium, that is surprisingly common in most minerals.
What matters is how easily the elements can be extracted and processed. Aluminium is 13% of the crust by weight, but was not possible to process before after around 1880, because electrolysis and large scale electricity production had to be in place first.
There are some smart[3] people on HN. I think it should be possible for someone to have an Earth globe, with a sidebar of statistically correct, usefully laid out information that the user asks for - how much alcohol is produced each year vs how much soya, or gold vs copper vs uranium; people living on less than $2USD per day vs people with average income of $20; amount of energy used by renewables and non renewables; amount of CO2 produced vs amount able to be 'sunk'; etc. Some data would need to be retrieved from other places (with sources given.) Caveats for biases and inaccuracies would need to be clear.
You'd monetise it by selling it to Wolfram Alpha.
I'm finding the spheres a bit tricky to get my head round.
How to Lie with Statistics covered this with the "money bags" example.[2]
I was told that the entire population of the Earth could fit on the Isle of Wight[1].
It might be useful to have spheres of some of the precious metals (silver, gold, etc.) next to the water sphere to help set the frame of reference. I'm guessing they'll be much, much smaller than the water sphere.
The actual amount of gold in the Earth is pretty poorly constrained, because most of it is expected to be at the core.
