S2 grid

Google’s S2 cells, derived from projecting the sphere onto a cube and ordering cells along a Hilbert curve for strong spatial locality. Good for global indexing.

The interactive explorer below is rendered with deck.gl and behaves like the h3geo.org map: the S2 level follows the zoom and cells are generated in the browser for whatever is in view. Two neighbouring levels are shown at once — the current level with a darker, heavier border and the next finer level with a lighter, thinner one — so the quadtree nesting stays visible. It is powered by the shared m3s_core Rust/WASM build so the cell ids and edges match M3S exactly. GIS-native (lon, lat) order is used throughout.

Why S2?

S2 is the global-scale workhorse: 30 exact-nesting levels from continents down to about a centimetre, no polar singularities (the sphere is projected onto a cube, not a cylinder), and Hilbert-curve ids that keep nearby cells numerically close — which is why it backs Google’s geo systems and BigQuery’s GEOGRAPHY type. Cells are spherical quadrilaterals: mostly near-square, more distorted near cube edges, and only approximately equal in area. Choose Geohash grid when you just need human-readable string prefixes, H3 grid for hexagonal analytics, or A5 grid when exact equal area is the requirement.

Usage

Encode a point and tile a small bounding box around Paris — same result in Python and JavaScript (both call the shared core):

Python

import m3s

cell = m3s.S2.from_geometry((2.35, 48.86))            # (lon, lat)
cells = m3s.S2.from_geometry((2.2, 48.8, 2.4, 48.9))  # bbox
print(cell.id, len(cells))

JavaScript

import * as m3s from "m3s";
await m3s.ready();

const cell = m3s.S2.fromPoint(2.35, 48.86);            // (lon, lat)
const cells = m3s.S2.fromBbox([2.2, 48.8, 2.4, 48.9]); // bbox
console.log(cell.id, cells.length);

from _deckmap import DeckExplorer, read_grid_js


DeckExplorer(
    center=(6.0, 48.5),
    zoom=5,
    grid_js=read_grid_js("s2"),
    hover="#CC6677",
    wasm=True,
)