The concepts you’ll learn here are not GeoPandas-specific. They apply just as well to other GIS tools like QGIS, ArcGIS, and pretty much any serious geospatial software you’ll encounter.
What Is a Coordinate Reference System (CRS)?
A Coordinate Reference System (CRS) (sometimes called a Spatial Reference System (SRS)) defines how locations on the Earth are represented using numbers (coordinates).
I like to think of a CRS as a spatial unit of measurement.
Just as temperature values only make sense when you know whether they’re in Celsius or Kelvin, coordinates only make sense when you know what system they belong to. Without a CRS, coordinates are just numbers with no spatial meaning.
Why Do We Need a CRS?
The Earth is not flat (controversial, I know 😄), yet most of our measurement tools and computational methods assume flat surfaces.
- Rulers, dividers, and grids work on flat planes
- Most measurement units (meters, miles, kilometers) are linear
- Even modern computer screens are flat
So in GIS, we need a way to represent a curved Earth on a flat surface in a consistent and measurable way.
That process is called map projection.
Projection is not unique to GIS. Rather, it’s a general mathematical process of converting a 3D surface to a 2D surface. GIS just happens to be one of its most visible applications.
How the Earth Is Represented in GIS
In GIS, representing the Earth happens in layers of abstraction:
- Datum: approximates the shape and position of the Earth
- Map Projection: flattens that curved surface
- Coordinate System: defines how we measure positions on the flat map
A Coordinate Reference System (CRS) simply bundles all of these decisions together.
Map Projections (The Big Picture)
At a high level, map projections fall into three broad families.
Conic Projections
These assume a cone placed over the Earth, onto which the surface is projected.
- Best suited for mid-latitude regions (roughly 30°–60° north or south)
- Commonly used for regional maps
Cylindrical Projections
Here, a cylinder is wrapped around the Earth.
- Excellent for direction and navigation
- Distortion increases dramatically toward the poles
- The famous Mercator projection is a very good example of cylindrical projection.
Planar (Azimuthal) Projections
These assume a flat plane touching the Earth at a point or along a line.
- It is usually centered on a specific location
- It can only depict part of the Earth at once
- It's often used for polar or localized maps
Map Distortion: The Trade-Off
Flattening the Earth always introduces a form of distortion. This is not negotiable. However, it can be mitigated.
Every projection preserves something, and distorts something else.
Common types of distortion include:
- Shape: Projections that preserve local shape are called conformal (e.g. Mercator).
- Direction: Projections that preserve true direction from a point are called azimuthal.
- Distance: Projections that preserve distance along certain lines or from specific points are called equidistant.
- Area: Projections that preserve area are called equal-area projections.
Most projections are designed to preserve one of these properties at the expense of the others.
This matters a lot for analysis.
- If you’re measuring distance to the nearest hospital, use a projection suited for distance.
- If you’re calculating census block areas, use an equal-area projection.
Coordinate Reference Systems: Geographic vs Projected
Now let’s talk about coordinate systems themselves.
Broadly speaking, there are two major types of CRS:
Geographic Coordinate Reference Systems (GCRS)
A geographic CRS represents locations using angular units i.e. latitude and longitude. Its unit of measurement in degrees.
- X-axis → longitude (ranging from -180° to 180°)
- Y-axis → latitude (ranging from -90° to 90°)
Due to their angular nature, geographic CRSs are not suitable for accurate distance or area calculations. For example, they may not be used for making buffers.
Projected Coordinate Reference Systems
Projected CRSs use linear units such as meters or feet.
- Coordinates are measured on a flat surface
- Values are often large (four to six digits)
- Designed for analysis and measurement
The raw numbers of projected coordinates are usually machine-friendly rather than human-friendly, but they’re essential for serious spatial analysis.
A Quick Rule of Thumb
Example: Same Place, Different CRS
Take Lagos as an example:
- In a geographic CRS (WGS84): Latitude: ~6.6° Longitude: ~3.3°
- In a projected CRS (UTM Zone 31N): Easting: ~533,163 m Northing: ~729,541 m
While they're the same place, the numbers that represent their coordinates are completely different. That’s why CRS information is not optional metadata instead it’s essential to every geospatial data.
Datum: The Earth Model Beneath Everything
The Earth is not a perfect sphere. It’s irregular due to:
- Mountains and valleys
- Gravity variations
- An uneven seabed
A datum defines how we approximate this messy reality using a smooth mathematical surface (an ellipsoid) and how that surface is positioned relative to the real Earth.
Different datums can place the same real-world location at slightly different coordinates.
That difference might be negligible for web maps, but it can be critical in:
- Land surveying
- Cadastral mapping
- Legal boundary work
Wrapping Up
I hope you now have a solid conceptual understanding of what a Coordinate Reference System is and why it sits at the heart of geospatial work.
The key takeaway from this post is this: CRS choices are not cosmetic. They directly affect how your data is displayed, how measurements are computed, and whether your analysis is even valid in the first place.
In this episode, I deliberately stayed on the theory side (i.e. projections, distortion, geographic vs projected systems, and datums) because these ideas form the mental model you’ll rely on every time you work with spatial data.
In the next episode of the GeoPandas 101 series, we’ll get practical by diving in to inspect CRS metadata, fix missing or incorrect reference systems, reproject datasets, and see how CRS choices affect distance and area calculations in GeoPandas.
See you in the next lesson.
Sources:
 Understanding Coordinate Reference Systems){kind=link}
Follow Me