Ever wondered why Greenland looks the size of Africa on Google Maps when it’s actually 14 times smaller? Or why your fiber-optic cable mapping starts going wrong as it crosses into a new region? The answer lies in map projections — the mathematical magic (and unavoidable trade-offs) behind every digital map you’ve ever used.
In this post, I’ll explain GIS projections in plain English, with real visuals built from actual world map data, and break down the three projections every GIS professional encounters: EPSG:4326, EPSG:3857, and UTM zones.
What is a GIS Projection?
Imagine the Earth as a round orange. A computer screen or a sheet of paper is flat. A GIS projection is the mathematical method used to flatten the curved surface of the Earth onto a flat map.
Here’s the catch: you cannot perfectly flatten a sphere. Try peeling an orange and laying the peel flat — you’ll either tear it, stretch it, or squish it. Every projection makes the same kind of compromise, distorting one or more of these properties:
- Shape — countries can look stretched or squashed
- Area — small countries can appear huge, and vice versa
- Distance — a centimeter on the map may not equal the same kilometers everywhere
- Direction — north may not point straight up
So projections aren’t just about making maps pretty — they’re about choosing which lie is most useful for your task.
Why Computers Need Flat Coordinates
The Earth is naturally described using latitude (north/south) and longitude (east/west). For example, Shimla sits at roughly 31.10°N, 77.17°E. These are angular measurements — degrees, not meters.
But GIS software needs to do real work: calculate distances, measure areas, draw buffers around points, render tiles on a screen. For that, it needs flat (x, y) coordinates, ideally in meters. The conversion from angular (lat, lon) to flat (x, y) is exactly what a map projection does.
1. EPSG:4326 (WGS84) — The Universal Standard
EPSG:4326, also known as WGS84, is the most common geographic coordinate system on Earth. It’s what your phone’s GPS uses, what GeoJSON files default to, and what most spatial databases store internally.
How It Works
EPSG:4326 doesn’t really “project” anything — it just plots latitude on the y-axis and longitude on the x-axis as if they were regular Cartesian coordinates. The result is a perfect 2:1 rectangle covering the entire globe.
Coordinate example for Shimla:
Longitude: 77.1734 Latitude: 31.1048 Units: degrees (NOT meters)
The Problem With Degrees
One degree of longitude is not a fixed distance:
- Near the equator: 1° longitude ≈ 111 km
- At Shimla’s latitude (31°N): 1° longitude ≈ 95 km
- Near the poles: 1° longitude shrinks toward 0 km
This is why measuring distances directly in EPSG:4326 gives wrong answers. Each row of pixels at the top of the map represents a tiny slice of real ground, while each row near the equator represents a huge one.
When to Use EPSG:4326
- ✅ Storing GPS coordinates in a database
- ✅ Sharing global datasets between systems
- ✅ GeoJSON, KML, and most file formats
- ❌ Don’t use it for distance, area, or buffer calculations
2. EPSG:3857 (Web Mercator) — The Internet’s Map
Open Google Maps, OpenStreetMap, Bing Maps, or any web map built with Leaflet or OpenLayers, and you’re looking at EPSG:3857, also known as Web Mercator or Pseudo-Mercator.
How It Works
Imagine wrapping a giant cylinder around the equator of the Earth, projecting each point on the surface outward onto the cylinder, and then unrolling the cylinder into a flat sheet. The result is the famous Mercator map you’ve seen in every classroom.
Coordinates are now in meters from a fake origin off the coast of Africa:
Shimla in EPSG:3857 X (easting): 8,591,500 m Y (northing): 3,632,200 m
Why Web Maps Love It
- Tiles are perfectly square at every zoom level
- Powers-of-two math makes zoom and pan extremely fast
- Local shapes stay almost correct — Italy still looks like a boot, the UK still looks like the UK
- Direction is preserved (compass bearings work correctly)
The Famous Greenland Problem
The cylinder trick comes at a cost: anything far from the equator gets dramatically inflated. Compare these two areas:
- Greenland: 2.16 million km² (real area)
- Africa: 30.37 million km² (real area)
Africa is 14 times larger than Greenland. But on a Web Mercator map, Greenland appears almost the same size as Africa. Russia balloons. Antarctica becomes infinite (which is why Web Mercator clips it off entirely above 85° latitude).
When to Use EPSG:3857
- ✅ Web mapping (Leaflet, OpenLayers, Mapbox GL)
- ✅ Tile servers and slippy maps
- ✅ Visualization where you don’t need accurate area
- ❌ Never for area calculations or precise measurements
- ❌ Engineering or surveying work
3. UTM Projection — The Surveyor’s Choice
UTM stands for Universal Transverse Mercator. Instead of trying to flatten the entire Earth in one go, UTM takes a much smarter approach.
How It Works
UTM divides the Earth into 60 narrow vertical strips, each 6° of longitude wide. Each strip is then projected separately, with the cylinder rotated 90° so it touches the Earth along a north-south line in the middle of that strip (the central meridian).
Because each strip is so narrow, the distortion within it is tiny — less than 0.1%. That’s accurate enough for engineering, surveying, drone mapping, utility infrastructure, and anything where being off by even a few meters matters.
Finding Your Zone
Zones are numbered 1 through 60 starting from 180°W moving east. Each zone has a north (N) and south (S) version above and below the equator.
- India spans Zones 42N, 43N, 44N, and 45N
- Shimla (77.17°E) falls in UTM Zone 43N (covers 72°E to 78°E)
- Delhi (77.21°E) is also in Zone 43N
- Kolkata (88.36°E) is in Zone 45N
Coordinate example for Shimla in UTM Zone 43N:
Easting: 707,500 m (distance from zone's central meridian, with offset) Northing: 3,442,800 m (distance from the equator) Zone: 43N
Same Region, Three Different Projections
To really see how each projection changes things, here’s the same area — India and surroundings — drawn in all three:
Notice the differences:
- In EPSG:4326, India looks slightly squashed vertically (degrees of latitude don’t equal degrees of longitude in real distance)
- In EPSG:3857, India looks taller and more like Google Maps
- In UTM 43N, the grid lines curve gently — and shapes here are the most ground-truth accurate
The UTM Boundary Problem (And Why Your Cable Map Broke)
Here’s a classic gotcha that catches almost every GIS engineer at some point. Suppose you’re mapping a fiber-optic cable that runs from Chandigarh (in Zone 43N) eastward to Lucknow (in Zone 44N).
If you keep using Zone 43N’s projection for the entire cable, here’s what happens:
- The portion inside Zone 43N: accurate to within a few centimeters
- The portion outside Zone 43N: distortion grows the further you go from 75°E (Zone 43’s central meridian)
- By the time you reach 84°E, errors can exceed several meters
This is why anything that crosses zones — long pipelines, transmission lines, fiber routes, highway corridors — needs either a country-wide projection (like Lambert Conformal Conic for India) or careful handling at zone boundaries.
When to Use UTM
- ✅ Surveying and engineering work
- ✅ Drone mapping and orthophotos
- ✅ Utility line mapping (electric, water, fiber) within a single zone
- ✅ Agricultural and forestry analysis
- ✅ Any local-scale measurement where accuracy matters
- ❌ Don’t use across multiple zones without careful boundary handling
- ❌ Don’t use for global datasets
Quick Comparison Table
| Projection | Units | Best For | Main Problem |
|---|---|---|---|
| EPSG:4326 | Degrees | GPS data, global storage, file exchange | Cannot measure distances directly |
| EPSG:3857 | Pseudo-meters | Web maps, tile-based visualization | Severe area distortion near poles |
| UTM (e.g. 43N) | True meters | Surveying, engineering, local accuracy | Only works within one 6°-wide zone |
The Real-World GIS Workflow
In professional GIS work, all three projections appear in the same project — each handling what it’s best at:
- Store raw GPS data and shared datasets in EPSG:4326. It’s universal, lossless, and what your data sources will give you.
- Visualize on the web in EPSG:3857. Every tile server expects it, and rendering is fast.
- Analyze in the appropriate UTM zone (or country projection) whenever you need to measure distance, area, or buffers. Switch back to 4326 or 3857 for storage and display.
Most modern GIS tools (PostGIS, QGIS, ArcGIS, GDAL) handle these conversions transparently — you just need to know which projection to ask for, and when.
The One-Sentence Summary
Every flat map lies — the question is just which lie is most useful for your task. Use 4326 to store, 3857 to show, and UTM to measure.
Got questions about projections, or hit a tricky boundary issue in your own GIS work? Drop a comment below — I’d love to hear what challenges you’re running into.
I hope this tutorial will create a good foundation for you. If you want tutorials on another GIS topic or you have any queries, please send an mail at contact@spatial-dev.guru.