Skip to main content

Stereographic Projections

Introduction

If you’ve compared a globe and a paper map, you’ve probably noticed that despite displaying the same information, the two have slight differences. For example, on most paper maps Greenland appears rather large, even though this is not the case on an actual globe. Such distortions are unavoidable when trying to make a spherical surface flat. Imagine trying to cut open an inflatable beach ball in such a way that it forms a perfect square or rectangle- it’s impossible! Therefore, in order to transform a globe into a map, we use a function called a stereographic projection. This post will explore how this is done and the distortions that result from it.

Stereographic Projections

The function to perform a stereographic projection is fairly simple. First, imagine we position our globe above the map we want to make. For any point \( a \) on our map, its resulting location on the globe, \( f(a) \), is found by drawing a line from \( a \) to the topmost point of the globe. The location where this line intersects the surface of the globe is \( f(a) \).

Let’s look at an example to get an idea of how this works. Imagine it is the year 2200 and you are one of many humans living in Martian settlements. One day, you are out driving on the planet’s surface when disaster strikes- you crash your space car in a crater. You have a limited amount of oxygen and need to get to the closest Martian settlement to avoid dying. Fortunately, you salvage a Martian map from the wreckage:

Since you really don’t want to run out of oxygen and die, you want to head toward the closest settlement. However, you are worried about map distortions and want to get an idea of what the actual Martian globe looks like. You happen to know (you are very smart) that this map was made from a stereographic projection in which the map was placed 50 units beneath the globe’s center. Consequently, you determine the stereographic projection looked as follows:

Great, you now have an idea of what the actual planet looks like. Your next step is to compare your map to the original globe to get an idea of what distortions have occurred.

Distortions

Your first priority is to find the closest settlement to walk to. Notice that on the original map the red path appears shorter, since the crater you’d have to walk around is smaller. However, on the original globe, the blue path actually appears to be slightly shorter!

Notice the same thing happens with the squares- the settlement at the end of the blue path appears to be larger in the paper map; however, on the globe we can see that it’s actually about the same size as the one at the end of the red path. This is because these shapes were stretched during the projection and appear bigger on the map than they actually are. This effect is much more pronounced at the edge of the map than in the middle. For example, consider these identical circles at the center of the map, which still appear identical when projected onto the globe:

Conversely, consider the three apparently identical circles at the edge of the map. Notice that on the globe their actual sizes are drastically different:

Let’s look at another example of distortion. Imagine you followed the blue path shown previously and arrived safely at the settlement. Looking again at the map, it seems you could travel in a straight line to arrive at the settlement in the lower right corner of the map. However, when you look at the actual globe, you realize the path isn’t straight at all!

In fact, if you tried to go in a straight line, you’d either miss the settlement or fall into a crater! This is another result of the stretching that happens during our projection. For example, look at the parallel lines (representing trenches) on the paper map that are much less parallel on the actual globe:

This same issue occurs if we are trying to get from the previous settlement to the home base. The trenches below seem to form a distinct angle on the paper map. However, on the globe, the ends of these lines get pulled downward, making the landmark seem much less angular:

Again, notice that all of these distortions are found at the edges of our projection- the center of our map is much more faithful to the original globe and, thus, much more trustworthy.

Why this example?

I chose this example primarily because I find maps to be the most interesting application of stereographic projections. It also gives us a meaningful way to discuss the resulting distortions, since it allowed me to discuss the way in which the projection could be misleading. I also thought the example of Mars would be interesting and engaging. Initially, I had intended the map to only consist of “craters”; however, I discovered that (other than comparing size differences) circles don’t stretch as notably as squares or lines. Thus, I added “trenches” and “settlements” in order to add more interesting features to my model.

Author: Sarah B

Comments

Popular posts from this blog

Do Over: Double Integrals over Regions

Introduction Over the semester we've looked at many topics and created 3D models. For this we are going to revisit an old topic, double integrals over a region. In this we found the volume of a surface in the xyz-plane bounded by two curves. From the many topics I chose to revisit this topic. I have a couple reason to why I chose to redo this. First, the model did not print correctly. The print added spaces between the rectangular prisms. Another reason was that I think the surface and curves did not represent the topic entirely. The surface I chose just increased between the curves. Improvements When making the model on Onshape there were no spaces between the rectangles, which can be seen on the right. However, when printing this spaces were being added. The second issue was with the surface I chose which was \(f(x,y)=xy+x\). This function only increased over the two curves I chose \begin{align*} f(x) &= \sqrt{x} ...

The Septoil Knot

Knots are a very interesting topic and a field that has not quite been fully discovered, so mathematicians are still discovering new ideas and invariances about knots even today. While it may seem like knots are a simple skill you learn at camp, they actually have a lot of mathematical properties and in this blog post we are going to look at just a few. By mathematical definition, a knot is a closed curve in three dimensional space that does not intersect itself. Since we are working with three dimensional space and you are reading this on a two dimensional screen, we need a way to look at knots in two dimensions and that is where knot projections come in. A knot projection is simply a picture of a knot in two dimensions and where a knot crosses itself in the projection is simply a crossing of that projection. The number of crossings of a knot is the smallest number of crossings among all projections of a knot. Since a knot is not necessarily solid, one...

Knot 10-84

Introduction In mathematics, a knot is simply a closed loop. The simplest version of this is the unknot, which is a just a closed circle (imagine a ponytail holder). Knots, however, quickly become more complicated than this more basic example. This post will examine a particular knot (knot 10-84) and a few of its knot invariants. Crossing Number Knots are often defined by their crossing number, which is the number of times the knot’s strands cross each other. As indicated in its name, knot 10-84 is a 10 crossing knot. In order to visualize the knot, we can look at its knot projection, in which the knot is represented by a line segment broken only at its undercrossings: Tricolorability Now that we’ve looked at knot crossings, we will examine a potential property of knots: tricolorability. In order to understand tricolorability, it is first important to know that one strand of a knot is defined as an unbroken line segment in the knot p...