Blog › What 30-Metre Resolution Actually Means on Screen and Paper

What 30-Metre Resolution Actually Means on Screen and Paper

November 6, 2026 · 3 min read

"30-metre resolution" appears in every description of global elevation data, including ours. It is also one of those numbers that means little until you convert it: 30 metres on a hiking map, on a monitor, on an A3 print, and on a phone screen are four different experiences. Here is the conversion, and the simple arithmetic behind it.

Where the number comes from

The elevation behind the terrain app comes from the AWS Terrain Tiles dataset — a global grid assembled from the SRTM space-shuttle radar mission, the USGS NED and other national elevation programs. Worldwide it resolves features down to roughly 30 m; countries with national LiDAR programs are considerably finer in the source data, and the dataset also includes bathymetry for the ocean floor.

Thirty metres means one grid cell — one sample of height — every 30 m of ground. A building (10–30 m) fits inside a fraction of one cell. A hill (500 m wide) spans about 16 cells. A mountain massif spans thousands.

The arithmetic every map reader should know

A rendered terrain map has a fixed pixel budget for its output grid — around 2,000 pixels on the long edge in the app's standard render. Metres per pixel is just division:

metres per pixel = selection width (m) ÷ output grid (px)

  a 20 km county   →  20,000 ÷ 2,000  ≈   10 m/px   (3 grid cells per pixel)
  a 100 km state   → 100,000 ÷ 2,000  ≈   50 m/px   (data oversampled)
  a 500 km country → 500,000 ÷ 2,000  ≈  244 m/px   (source far finer than image)

Read the table right and two things follow. First, shrinking the selection is the real zoom: a county view delivers near-source detail because each output pixel covers just a few elevation cells, while a country view cannot possibly show 30 m features — each pixel averages a quarter-kilometre. Second, past a point, more pixels cannot add information: a 4× render of a 500 km country makes smoother edges, not new mountains. The detail is bounded by the DEM, and always was.

On paper

Print raises the bar, which is why the poster export exists. An A3 sheet at 300 dpi is about 3,500 × 4,900 pixels — so the honest advice for a printed poster is the same as for the screen: keep the selection small enough that the terrain, not the interpolation, does the talking. A 30–60 km national park or valley system is the sweet spot where the 30 m grid still supplies genuine structure at print density.

One practical tell: if two prints of the same mountain at different sizes look equally sharp but the larger one has softer ridges, you are looking at interpolation. The DEM is a ceiling, not a promise.

How to check any claim — including ours

Hover anywhere on the map and the app reads out the exact elevation at that point — sample the same spot at different zooms and you are querying the same grid cell, getting the same number. That consistency is the whole point of a DEM: one dataset, one truth, rendered at whatever scale you ask. The heightmap export goes one further and hands you the grid itself, one pixel per cell, for your own project.

And when a country's national survey publishes finer data than 30 m, the gap between "the map" and "the ground" shrinks again — the pipeline updates from the same open sources, and the arithmetic above never changes.