0:23

To start with, if I was to tell you the location of something

as being 55 meters away in the x direction and

41 meters away in the y direction, where is this item located?

It all depends on where those coordinates are relative to.

Are they relative to me right now?

Or are they relative to some standardized location on the earth?

This distinction is very important.

Packed into that description I just gave you are two important concepts about

geographic data.

Number one, I gave you a unit of measure for our coordinates in meters.

Without units, distances are meaningless.

And number two, we need to have a reference point with a known location in

order to locate items with these coordinates.

This reference point is often called a datum and it's effectively a model

of the Earth's surface that coordinate systems can be built on.

Together, these two concepts form the basis for a coordinate system.

Not all data uses the same coordinate system.

Far from it.

And to be displayed on the same map,

your data doesn't need to be stored in the same coordinate system.

But the GIS software does need to convert your data

to the same coordinate system behind the scenes.

This will intuitively make sense because you have different reference points and

different coordinates.

How can you overlay them without some conversion to a common system.

Building on coordinate systems is the concept of projections.

The term projection is often used interchangeably with coordinate systems,

and you may hear me make that mistake occasionally.

But, in fact, they are different, and projections build on coordinate systems.

So to start with, what is a projection?

Projections help us display the earth on a flat surface like your screen or

a sheet of paper.

While it may not be intuitive at first,

we can't just flatten the Earth easily to fit on your screen.

To help illustrate this concept, let's try to imagine flattening a sphere.

Just like we would have to do to display the Earth on your screen.

Imagine an inflated spherical ball, just like a football, or a soccer ball for

you Americans.

Let's cut it down the side from top to bottom so

that the interior hollow part is exposed.

To make it even easier to visualize it we can cut it into completely into halves so

that we can set the cut side on the ground.

We now have half the ball sticking up off the ground but there's no

easy way to completely flatten it so the skin of the ball is against the ground.

If this ball was the Earth, we would need to stretch and distort it

in order to get it completely flat to display on your screen or on paper.

This set of stretches and distortions is what a projection is.

When working on a two-dimensional surface like computer screen,

we get some benefits out of working with projected data.

First, lengths and angles can be constant across the two dimensions

which we can't always say about our geographic coordinate system.

Think about the lines of longitude and how they converge at the poles.

The distance between the degree of longitude at the poles is very

different than the distance between the same degrees at the equator.

Projected coordinate systems have uniform distances and

map units regardless of location, as well.

This lets us identify locations by X, Y coordinates on a grid.

To bring this all together, to build a geographic coordinate system we

need an accurate model of the Earth's surface.

From there, coordinate systems are built.

And building on that are projected coordinate systems.

Now, we don't get this translation of our data to a 2D surface for

free, there are trade offs.

When we project geospatial data, you end up creating distortions.

Distortions can occur in the shape, the area, the distance, or

the direction of the data.

Different projections are created to optimize for

these distortions so some projections are good at preserving local shape.

These are called conformal projections.

Others preserve the area of the features.

These are called equal area projections and

still others preserve distances between points on the map.

These are called equidistant projections.

In practice a projection must restore at least one attribute.

Shape, area, distance, or direction.

Different projections also optimize these attributes for

different locations on the earth, and others do it for the entire earth.

The result is a large number of projections optimizing for

different attributes and different locations.

Depending on the work that you're doing,

you will find yourself needing different projections and coordinate systems.

So now let's take a look at some projections.

Before we do that, let's take a look at the earth on a sphere.

We'll start by looking at Africa, and

then we're going to compare it to the size of Greenland.

Since Africa is near the equator, it's shown closer to its true size in this

often undersized relative to the rest of the world on a world wide projection.

In contrast, Greenland is oversized by virtue of being

near the poles where more distortion occurs.

5:04

So keep these size differences in mind as we look at the following map projections.

The first projection here is the equirectangular projection.

It results from simply taking the angular coordinates of the globe and

plotting them as if they are linear coordinates on a sheet of paper.

So as you get closer to the poles, you have more distortion as the meridians,

your lines of longitude, stay the same with the part

on the sheet of paper instead of converging as they do on the globe.

It's a common choice in GIS software for

displaying data stored in a geographic coordinate system.

I want you to note right here the size of Greenland relative to Africa as well.

Next up is the Mercator projection.

You've probably seen quite a bit of this projection as it's very

common on the Internet.

A variant of the Mercator projection is used to most mapping applications online.

Note that in this projection,

Greenland is about as big as Africa, much, much larger than it should be.

The distortion is the result of this projection preserving angles

in order to aid a navigation and sacrificing sizes as a result.

It's an instance of a conformal projection.

Now, take a look at the Mollweide projection.

It is an equal area projection,

which means that area is preserved within the map.

Note that Greenland is it's appropriate size relative to Africa, but

we've distorted the shape of all of these locations in order to get

the appropriate areas.

So far the projections we've looked at have been very simple and optimized for

displaying things around the world.

As a result the distortions that we have in our maps are greater,

the Universal Transverse Mercator or UTM for

short, projection, tries to account for this by creating a series of

nearly identical projections that optimize for each area of the planet.

If you want to understand more details of exactly how it does this,

you'll need to take the rest of the courses in the specialization, but for

now we'll show you enough to use it.

The UTM projection divides the earth into 120 zones, 60 in the north and

60 in the south.

Each of these zones is 6 degrees of longitude wide.

Since this is just a rotated Mercator projection,

we're minimizing distortion by effectively making each location have the properties

that exist near the equator in the standard Mercator projection.

In doing this, we minimize the distortion of area that occurs in the Mercator

projection while getting the benefit of preserving angles.

You can still display locations that are outside of each zone, but

they become more distorted as they do in a Mercator map.

This projection is very useful, and I encourage you to look up

the zone that you are in as it could become a common part of your mapping.