The best places to launch into space from are Kismayo (Somalia), Macapa (Brazil), Galapagos Islands (Ecuador), Christmas Island (Kiritimati), Nauru, Sorong (Indonesia), Ternate (Indonesia), Gorontalo (Indonesia), Bontang Lng (Indonesia), and Singapore. Further, they really are the best places to launch from, not just a few degrees off of the Equator, but just about dead on it (Note 1).

But just because these 10 places are the best in terms of the physics, does not mean that they are the best in terms of sociogeography. What I mean is that each of them has pro and cons in terms of the rest of the Earth. Specifically, the geography of the Earth’s landmasses, it’s weather, the populations of the people of the Earth, and the Earth’s resources.

First, let’s look at the map of the places where physics says that it’s best to launch from:

Now, let’s draw some lines on the map. But not just any old line, these lines are going to make a Voronoi Diagram. This is when you put points on a plane (here we are using a sphere, but Mercator projected onto a plane) and then draw lines that are the same distance to one point or another. In the below picture, we see a bunch of black dots. The different colors are the ‘cells’ of that dot. Inside each of those colored polygons, any part of that colored ‘cell’ is closer to the black dot (in the cell) than to any other black dot. Eventually, as you move away from one black dot, you’ll get far enough away that you get closer to another black dot. That’s when you are in another colored cell. This way, if you pick some random point on a plane, you can easily see which black dot is closest.

Now, let’s take the idea of Voronoi diagrams and apply it to the points on the Earth that are great launch sites. Remember, since we are using a Mercator projection of the Earth on a flat plane, the lines should be warped. In fact, the North and South Poles will be where all the lines intersect, as the poles are each the furthest you can get from the equator.

                I’ve cut off the very tops and bottoms of the map, as it’s not very interesting. Still, you can see a slight curve to the lines (remember, these are boundary lines where you are equally distant to any one launch site).

What the map now shows you are the ‘starsheds’ of the Earth. I am making up a word here, so bear with me for a bit. I’m using starshed just like how a watershed is used to see where some random drop of rain will eventually flow out to the ocean. What these starsheds show us is where the closest launch site/starport is, i.e. where something on Earth will flow out to space. We see that the Atlantic and most of South America are in the Macapa starshed. Most of India, Indochina, and Southern China are in the Singapore starshed. The three Indonesian starsheds comprise the eastern Chinese coast, Japan, and Australia. The Nauru and Kiribati starsheds take really only some of Kamchatka, Alaska, British Colombia, and New Zealand. While the Galapagos starshed takes North America and some of the west coast of South America. Most of Europe, Africa, the Mediterranean, the Middle East and some of Russian and India are in the Kismayo starshed.

One thing that jumps out is the number of starsheds in Indonesia. Currently, those islands are all Indonesian. Most countries don’t have a space program, they tend to be really expensive, afterall. Even the US really only has one main spaceport (sorry Vandenberg!) where the majority of space-bound things depart.

So, now I am going to make a bold and probably debatable assumption: Singapore is going to be the only real starshed in that area of the world. Yes, there may be other launch sites in Indonesia, but Singapore is going to be the ‘main’ spaceport in that area of the world. I’ll get to why I think this as we go along; I promise it’ll make sense.

So, now that I have knocked out all the other starsheds in Indonesia, here is the list of spaceports: Kismayo (Somalia), Macapa (Brazil), Galapagos Islands (Ecuador), Christmas Island (Kiritimati), Nauru, Singapore. And here is the starshed map with the new, smaller, list of spaceports:

Now the Singapore starshed includes most of India, half of Australia, all of China, Indonesia, most of Japan and eastern Russia. The Nauru starshed now takes half of Paupa New Guinea and Australia, New Zealand, and some of Japan. By consolidating the Indonesian starsheds into Singapore, not much was lost, largely it was just portions of Japan and half of Australia. I believe that, for the cost of running a spaceport, this trade off from five spaceports to just one (that includes most of the previous ones anyways) is a good idea and makes practical sense.

However, when it comes to spaceflight, there is one thing to consider: Weather. The number of ‘clear’ and launchable days that a spaceport has is very important (Fun Fact: Kennedy is the “lightning capitol” of the US and about half of all launches are weather delayed). One way to measure the number of days that a launch can be performed is to find the number of days per month without cloud cover. Fortunately, this is straightforward to find:

The seasonality of Macapa is really easy to see! It’s really wet there in the spring months. Also, the Galapagos gets a lot of rain and is mostly cloudy year round. However, Kismayo and Kiribati are pretty clear places year round. This means that some places will switch their starshed during different parts of the year. In April and March, places that are kind of near the border of the Voronoi cell between Macapa and Kismayo with likely use the Kismayo starport instead of Macapa since the skies are more likely to be clear in Kismayo at that time. But in October, the two starports will be about as clear as one another and the Voronoi cell will be unmoved. So, when looking at the Voronoi cells, you have to adjust them based on the amount of ‘clear days’ each one has. Effectively, each cell must be puffed up or shrunk compared to the neighboring cells’ number of clear days. Since we are looking at long term trends, the month to month swings will just be averaged out over the year. Here is how the number of clear days per year looks like (I threw in Cape Canaveral for a comparison too):

This is a bit easier to look at over the course of the year. The clear winners are Kiribati and Kismayo. They have nearly 1.5 times as many clear days as Macapa, Singapore, and Nauru. However, each starport must adjust the border of the Voronoi cell with each partner, one by one. When you do that, the map changes to look something like this:

See that Kismayo now takes all of Africa and Europe and most of India, as it’s so clear there. The Galapagos gives up more of eastern North America; however, the rest of the Voronoi cells stay essentially the same relative to each other.

Still, take a close look at Western Africa, about where Togo and Cote d’Ivore are. The weather adjusted Voronoi cell says that it’s closer to Kismayo than to Macapa. But that doesn’t make a lot of sense. It would be very easy to sail a boat full of rocket parts, space food, or other goods straight east across the Atlantic. If you wanted to go to Kismayo, you’d have to sail around the Cape of Good Hope and up Eastern Africa or you’d have to put the goods on a railroad across Africa.

Look at Louisiana. It’s in the Galapagos starshed, but shipping parts to Macapa looks like it would be easier than waiting and going through the Panama Canal. Also look at Portugal. It’s in the Kismayo starshed, but it seems like shipping things to Macapa would be easier than going through the Mediterranean Sea and the Suez Canal.

So, let’s adjust the starsheds even further and talk about the shipping of goods and where the lines of the Voronoi cells should be based on that too. One thing to note here is that riverways and watersheds play a big part in this analysis. It turns out that shipping anything that weighs at least one metric ton or more is mostly done on boats. Since boats don’t easily move on land, where the mouth of a river is plays a BIG role in the shape of the starshed. It also turns out that if you want to calculate the time to ship something from any part of the Earth to any other, you can pretty easily look that up. The internet is a nice wonderful thing!

Now we have it all! These are the starsheds of planet Earth. If you want to ship something off of this world, and you are sitting somewhere on it, this is the map of where you will be sending your one metric ton of goods or services averaged over one year’s worth of weather. Look at West Africa. Due to the Niger River and the time it takes to go around the Cape of Good Hope, the Voronoi line is really jagged. This also happens in India with the Ganges River and up through Tibet and eastern Russia. The Galapagos starshed is also really strange and actually gets cut off in Northwestern Canada. The only real parts of the world it covers are the western US and the west coasts of Central and South America. The delay times through the Panama Canal really shrink it’s starshed.

The important thing to note is that the Singapore starshed just barely captures the mouth of the Ganges River and nearly all of eastern Asia except for Hokkaido, Japan. This is really important because it then captures a lot of the population of the world and their economic activity. If there were the other Indonesian starports competing in that starshed, then each of them would be taking slivers of the populations that live there. Also, Singapore is JUST west enough to get to capture the Ganges river. Kismayo, due to it’s clear skies, goes up the coast of India quite a ways.

However, the Kismayo starshed manages to just be close enough to the mouth of the Niger River that all the people and economic activity there go to the stars through Kismayo and not Macapa. Lagos, the Nigerian capitol, is just barely in the Kismayo starshed and may swish back and forth to the Macapa starshed as weather changes. Still, ALL of Europe, the Middle East, much of India and Africa, and most of Russia are in the Kismayo starshed.