// geographic relationships — aid corridors — fy 2024

Europe and North America Anchor a Hub-and-Spoke Aid Network That Converges on East Africa and the Levant

Lines connect coordination hubs ● to field operation sites ○. Arc weight encodes annual tonnage. Hover any node to reveal its corridors. Data: simulated FY 2024 supply network.

Connection map: global humanitarian aid corridors, FY 2024 World map on Natural Earth projection. Five coordination hubs — Geneva, New York, Washington DC, Brussels, London — connect via weighted arcs to field operation sites in East Africa, the Levant, South Asia, and elsewhere.
Node Type
Coordination hub
Field operation site
Annual Tonnage (MT)

What is a Connection Map?

A Connection Map (also called a Link Map or Ray Map) encodes geographic relationships by drawing lines between points placed at their real coordinates. The lines are the data — not the areas, not the polygons. What the viewer reads is the structure of connections: how many a node has, how far they reach, how concentrated or dispersed they are across space.

Arc stroke-width is proportional to flow volume using a square-root scale, so area — not raw pixel width — grows in proportion to tonnage. Color differentiates hub-to-field corridors from regional connections when the filter buttons are active. Shape redundantly encodes node type (filled circle = hub, outlined circle = field site) so the distinction is never color-dependent alone.

This example shows a simulated FY 2024 humanitarian aid supply network. Five coordination hubs — Geneva, New York, Washington DC, Brussels, and London — connect via great-circle arcs to field operation sites in East Africa, the Levant, South Asia, and West Africa. Nairobi, Amman, and Kyiv receive the highest combined inbound volume. Hover any node to see its full corridor profile — total outbound and inbound tonnage, number of corridors, and operating organization.

Chart typeConnection Map (Link / Ray Map)
FT Visual VocabularySpatial — connection
Best forOrigin-destination flows, network topology at geographic locations
Avoid whenMagnitude at a place (use bubble map); rates (use choropleth)
Arc routingGreat circle geodesics via d3.geoPath()
// learn — connection map (link map · ray map)

Why a Connection Map

What this chart is

A connection map encodes geographic relationships by drawing lines between points placed at their real coordinates. The lines are the data. Positions are geographic facts. What the viewer reads is not the magnitude of a place but the structure of its connections: how many it has, how far they reach, how concentrated they are. The perceptual channel is proximity and density of lines. A node with many thick arcs is obviously central; an isolated node with one thin arc is obviously peripheral. Unlike a choropleth or a bubble map, the connection map says nothing about a region's intrinsic attributes — it only speaks to what each place is connected to.

Why it was chosen here

The data is a logistics network: specific origin-destination pairs with a flow volume. The message is structural — the network converges on a small number of field sites from a distributed set of hubs. That is a connection story, not a magnitude story. A bubble map would encode aid volume at each location but would erase the corridor structure entirely — you would see which cities receive the most tonnage, but not which hubs supply them or how the network is wired. The connection map shows the wiring.

Secondary encoding: arc stroke-width is proportional to annual tonnage via a d3.scaleSqrt() mapping — the same sqrt-area principle as proportional symbol maps, here applied to perceived line weight. Color differentiates hub-to-hub corridors from regional connections when the filter buttons are active. Shape redundantly encodes node type (filled circle = hub, outlined circle = field site) so the distinction is not color-dependent alone.

What a flow map would break — and what it would add

A flow map is a specialised form of connection map that encodes directionality and often merges parallel routes into a single tapered ribbon. It would be correct here — aid flows from hubs to field sites, not the other way. The reason it was not chosen: the routes share many intermediate paths and a proper flow map would need to merge those paths, which requires Sankey-style path merging that adds significant rendering complexity for a marginal gain in this dataset. The connection map's arrowheads carry enough directionality for the message at hand.

The other honest alternative: a matrix/adjacency heatmap, which would make the hub-by-field-site combinations legible without any geography. Rejected because the geographic position of Nairobi relative to Amman relative to Dakar is informative — it reveals the spatial concentration in East Africa and the Levant, which would be invisible in a matrix.

// framework — FT Visual Vocabulary

The FT Visual Vocabulary places connection maps in its Spatial category alongside choropleths and bubble maps. Its note on the type: connection maps are best when the connection itself is the story — not the attributes of the connected places. The operative test: could you make this point without geography? If showing which cities are connected without their actual positions would lose the story, geography is earning its keep and a connection map is the tool.

// design decision — great circle arcs vs. projected straight lines

The arcs follow great circle routes — the shortest path between two points on a sphere. This is implemented by passing each connection as a GeoJSON LineString to d3.geoPath().projection(naturalEarth), which automatically interpolates the geodesic. On a Mercator or equirectangular projection, great circle arcs produce visible curves. On Natural Earth 1 (used here), the effect is subtler but remains geometrically honest. Straight lines in projected space would be shorter visually but incorrect physically. Geodesics were used because this is an aid-logistics dataset where the real route geometry is the subject.