Desire Line Analysis
- Introduced by
- Transportation planning tradition (urban transportation studies)
- Year
- 1955
- Type
- Mapping and analysis of origin–destination travel demand as straight flow lines
- Family
- Flow mapping and spatial interaction visualization
What does it do?
Desire line analysis reveals the underlying demand for travel between places by drawing straight lines that connect each origin to each destination, with line width or weight proportional to the volume of flow between them.
The term comes from transportation planning, where a 'desire line' represents the direct, idealized path a traveller would take if no network constrained them — capturing where people want to go, not how the roads make them go. Aggregating trips into an origin–destination matrix and rendering it as weighted lines exposes the dominant corridors of movement, making desire lines a foundational tool for visualizing and analysing travel demand.
The intuition behind the method
Strip away the roads and ask only: where does everybody actually want to get to? Draw an arrow from each starting point straight to each endpoint and make it thicker the more people make that trip, and a pattern leaps out of the tangle — fat lines pointing at the city centre, bundles converging on a campus or a station, faint threads showing trips few people make. The straightness is deliberate: it ignores the contingent geometry of the existing network and shows raw demand, so the analyst can see corridors that the current infrastructure may serve well, badly, or not at all. Desire lines are the demand-side complement to network analysis, which shows the supply side.
History
Desire lines entered planning practice through the great urban transportation studies of the 1950s — notably the Detroit Metropolitan Area Traffic Study (1955) and the Chicago Area Transportation Study (1959) — which collected origin–destination survey data and mapped travel demand as bundles of straight lines to justify expressway and transit investment. The technique grew alongside the four-step transport model, in which trip distribution produces the OD matrix that desire lines visualize, a lineage documented in Boyce and Williams' history of urban travel forecasting. With modern GIS, big mobility data, and edge-bundling algorithms, desire line analysis has been revived for cycling-potential tools and interactive flow maps that handle far larger matrices than the hand-drawn originals.
When to use it
Use desire line analysis to visualize and explore origin–destination travel demand — commuting patterns, transit ridership, cycling potential, migration, or freight — when you want to see the geography of where people move between places before worrying about how. It is ideal for communicating demand to stakeholders, identifying dominant corridors, and spotting mismatches between demand and existing infrastructure. It works whenever you have an OD matrix and zone locations. It is less suitable when the actual routed path matters (use network or flow-mapping methods that follow the network), when the number of zones is so large that even filtered lines overwhelm the map, or when directionality and asymmetry of flows must be preserved precisely, which straight undirected lines can obscure.
Limitations and considerations
- Straight lines ignore the real network, so they show demand geometry, not achievable routes.
- Dense matrices produce visual clutter that filtering can only partly resolve.
- Results are sensitive to zone definition and centroid placement (the modifiable areal unit problem).
- Overlapping and crossing lines can mislead about which flows are largest or where they concentrate.
Source conditions do not replace independent scientific review or validation by an analysis engine.
Strengths
- Exposes the dominant geography of travel demand at a glance, independent of the current network.
- Simple to construct from an OD matrix and zone centroids, with no routing required.
- Excellent for communicating demand patterns and corridor priorities to non-technical audiences.
- Reveals demand–supply mismatches where strong desire lines lack adequate infrastructure.
Applications
- Mapping commuting flows between residential and employment zones to plan transit and roads.
- Assessing cycling potential by visualizing short-distance desire lines suitable for bike trips.
- Visualizing public-transport ridership demand to prioritize routes and frequencies.
- Displaying migration or freight flows between regions for regional planning.
- Identifying corridors where strong demand is poorly served by existing infrastructure.
How it works
How it works
The raw material is individual trips, which are aggregated into an origin–destination matrix T_ij counting how many trips go from each origin zone i to each destination zone j. Trips come from travel surveys, ticketing or mobile-phone data, censuses of commuting, or model outputs. The zones are typically administrative areas or traffic-analysis zones, and the aggregation collapses thousands of individual movements into a compact, analysable flow table.
Formula
T_{ij} = \sum_{\text{trips}} \mathbb{1}\{o = i,\, d = j\}Each non-zero matrix cell T_ij becomes a straight line L_ij drawn between the centroid of origin zone i and that of destination zone j, deliberately ignoring the actual route taken. The line's width, colour, or opacity is scaled in proportion to the flow volume T_ij, so heavy movements appear as bold lines and minor ones as faint ones. The collection of these weighted straight lines is the desire line map, encoding the whole flow matrix geographically.
Formula
L_{ij} = \overline{c_i\, c_j},\qquad \text{width} \propto T_{ij}A full origin–destination matrix produces a dense, illegible hairball of lines, so flows are filtered and simplified. A threshold τ drops lines below a minimum volume; top-n filtering keeps only the strongest flows; and edge-bundling or aggregation merges nearby lines into thicker corridors. These steps reduce visual clutter so that the dominant structure — the corridors that carry most demand — becomes readable without discarding the analytically important flows.
Formula
\text{draw } L_{ij} \iff T_{ij} \ge \tauWith the clutter removed, the map is read for structure: the thickest lines mark the dominant travel corridors, radial bundles reveal centripetal demand toward employment or service hubs, and conspicuous gaps point to latent or poorly served demand where strong desire lines lack good infrastructure. Planners use this to prioritize transit routes, road investments, and cycling networks, aligning supply with the demonstrated geography of demand.
Formula
\text{dominant corridors} = \arg\max_{(i,j)} T_{ij}Common pitfalls
- Plotting every flow without thresholding or bundling, producing an unreadable hairball.
- Treating the straight desire line as a route and inferring infrastructure needs from its path.
- Letting large zones place centroids far from real trip ends, distorting line origins and directions.
- Ignoring flow direction and asymmetry when bidirectional volumes differ substantially.
Frequently asked questions
Why are desire lines drawn straight instead of along roads?
The straight line deliberately abstracts away the existing network to show pure demand — where people want to go regardless of how the current infrastructure forces them to travel. This makes desire line analysis the demand-side complement to network routing, and it is precisely the straightness that lets planners spot corridors where strong demand is not yet matched by good infrastructure. If you need the actual travelled path, you route the flows over the network instead, which is flow-mapping analysis.
How do you stop a desire line map from becoming an unreadable hairball?
Because a full OD matrix has thousands of cells, you must filter and simplify: apply a volume threshold to drop minor flows, keep only the top-n strongest lines, or use edge-bundling to merge nearby lines into corridors, and encode volume through width, opacity, or colour so dominant flows stand out. The goal is to preserve the analytically important structure — the heavy corridors and the gaps — while removing the clutter of countless small flows.
How does desire line analysis relate to gravity and spatial interaction models?
Desire line analysis visualizes an origin–destination flow matrix, while gravity and spatial interaction models predict the values in that matrix from zone sizes and the cost of travel between them. The two are complementary: a spatial interaction model generates or completes the OD matrix, and desire lines render it geographically so the predicted demand can be inspected, validated, and communicated. Desire lines are thus a natural display layer for the output of interaction models.
Related methods
Sources
Boyce, D. E., & Williams, H. C. W. L. (2015). Forecasting Urban Travel: Past, Present and Future. Edward Elgar, Cheltenham.
ISBN: 9781848440319Copied.