OpenStreetMap can provide useful building outlines and some height information for an early shadow study, but coverage varies by place. A reliable workflow distinguishes mapped height tags, floor-derived heights, explicit estimates and missing outlines instead of treating every building as equally certain.
An infill study before the survey arrives
In this fictional example, Sam is an architect preparing a feasibility study for a small urban infill project. The client wants to compare a compact six-storey block with a stepped five-storey alternative. The team has the site address and approximate proposal dimensions, but the measured survey and neighbour roof levels will arrive later.
Sam wants enough context to identify a promising massing direction before investing in detailed modelling. At the same time, the result will be shown to a client, so hidden assumptions would be risky. Sam needs to know which surrounding outlines were loaded, which heights come directly from map tags, which are derived from floor counts, and which are only defaults.
This guide is for Sam and similar readers: architects, urban designers, developers and technically curious homeowners who want the speed of map-based context while keeping its limits visible. The goal is not to prove that OpenStreetMap is accurate or inaccurate for a whole city. The goal is to decide whether the particular context controlling a particular shadow question is good enough for the current stage.
What the tool reads from OpenStreetMap
The tool requests mapped building ways within the radius you choose. It uses each returned outline as a plan footprint. For height, it applies a clear priority:
- Mapped building height (OSM field:
height). A usable positive height value is applied first. OpenStreetMap defines building height as the physical height of the feature; for buildings, the description includes the roof and excludes mounted equipment such as antennas. - Mapped number of storeys (OSM field:
building:levels). When height is absent, the tool multiplies the above-ground level count by 3 metres. The storey count excludes roof and underground levels. OpenStreetMap’s documentation discusses 3 metres per storey as a common default approximation. - Visible estimate. If neither value is usable, the tool applies 3 metres to certain low building types and 9 metres as its general default. These are tool assumptions, clearly counted as estimates.
The map tooltip shows the source used for each building. The summary counts mapped heights, floor-derived heights and estimated heights, so the user can see how much of the loaded context depends on assumptions.
What a confidence label can and cannot mean
| Tool label | Meaning within the loaded context | What still needs checking |
|---|---|---|
| High | At least 75% of loaded outlines use a mapped height | Outline completeness, currency, roof geometry, ground level and tag accuracy |
| Moderate | Below the High threshold, but at least 60% use mapped or floor-derived heights | Floor-to-floor assumptions and all items above |
| Limited | Context was not loaded, no outlines were found, or fewer than 60% have mapped or floor-derived heights | Whether critical surroundings are absent or poorly described |
A mapped height is not necessarily surveyed: the presence of a tag alone does not verify its source or accuracy. The label is deliberately narrow. It describes the height sources of the outlines returned to the tool. It is not a completeness score for OpenStreetMap and does not certify the real site. The statement “missing outlines remain unknown” matters in every confidence band.
A worked context audit for an infill option study
Sam locates the site and loads nearby buildings within a modest radius. The summary reports 28 outlines: 18 mapped heights, six floor-derived heights and four estimates. That distribution produces a Moderate label, but Sam does not stop at the percentage.
First, identify the sunward context
For the chosen winter afternoon, likely obstructions lie in the direction of the sun. Sam watches the plan-view shadows and identifies three neighbouring buildings capable of reaching the analysis point. Distant buildings behind the point do not deserve the same immediate attention, even if their data is incomplete.
Second, click the critical outlines
One neighbour has a mapped height, one is derived from four levels, and the closest uses the general estimate. The closest estimated object controls much of the blocked interval. The overall confidence label cannot remove that local weakness. Sam records the building as the first verification target.
Third, compare relative and absolute conclusions
Both proposals use the same OSM context. Proposal B preserves more afternoon sun than Proposal A because its stepped upper floor causes less added shade. That relative comparison may be useful now. The absolute claim—exactly how many hours of direct sun the point receives—remains provisional because the closest neighbour height is uncertain.
Fourth, identify the sensitivity check still needed
Sam cannot edit an OSM building inside the tool, so the team compares the map result with available site evidence and notes a plausible height range for later modelling. If a small height difference could reverse the option choice, the feasibility report does not select a final mass. It asks for the survey first. A later sensitivity test with editable, verified context is needed before claiming that Proposal B remains preferable across that range. The current tool cannot establish that by changing an imported OSM height.
The example shows why data quality is a design input rather than a footnote. A single critical estimate can matter more than many high-quality buildings that never shade the selected point.
Why outline completeness matters to a shadow result
A shadow calculation cannot include a building it does not know exists. If a tall neighbour is missing from the map, the tool can overstate direct sunlight. If a demolished structure remains mapped, it can understate it. Construction dates, recent extensions and detailed roof changes may also lag behind reality.
This is why the colour map alone is not evidence of complete 3D context. Compare important surroundings with a current survey, recent imagery you are permitted to use, planning documents, site photographs or other verified project records. Treat conflicts as open items.
A short OSM context checklist
- Check the radius. Low sun can produce long shadows. Make sure the requested area includes likely obstructions in the sunward direction.
- Count the outlines. Compare the loaded map with current knowledge of the site.
- Click critical buildings. Read the name, applied height and source quality in the tooltip.
- Look for estimates. An estimated building close to the analysis point deserves more attention than a distant estimate.
- Record missing context. Add the limitation to the report and obtain verified geometry when it could change the option choice.
How OSM context affects Existing and Proposal comparisons
The same loaded surroundings are used for Existing, Proposal A and Proposal B. That consistency is valuable: a proposal’s gained or lost sunlight can often be attributed to the planned objects that differ between scenarios. The shared timeline and cause note help identify that design effect.
Consistency does not remove source uncertainty. An absent neighbour affects all scenarios, and the absolute sunlight duration can still be wrong. Separate two conclusions:
- Relative conclusion: under the same context assumptions, Proposal B preserves more afternoon sun than Proposal A.
- Absolute conclusion: the selected point receives a particular total duration on the stated date.
The relative conclusion may remain helpful earlier in design. Both conclusions depend on critical obstructions: shared errors can also change the ranking of proposals when their shadows overlap the missing or uncertain context.
Use a confidence ladder tied to the decision
| Project stage | Reasonable use of OSM context | Minimum next action |
|---|---|---|
| Idea screening | Compare obviously different heights, setbacks or locations | Record that context is map-based and inspect visible gaps |
| Client option review | Explain relative differences between coherent proposals | Check every critical outline and source; list estimates in the report |
| Design commitment | Use the early study to direct verified modelling | Replace controlling estimates and missing objects with measured geometry |
| Formal or legal assessment | Retain OSM work only as background or option history | Follow the prescribed method with suitable survey data and professional review |
Use the decision to set the level of evidence. Map data can support early exploration; a design commitment needs verified information about the buildings that control the result.
What to write in the report
- The selected radius and whether nearby context was successfully loaded.
- The number of mapped-height, floor-derived and estimated outlines.
- The names or OSM identifiers of important surrounding buildings, identified from the map and supporting records.
- Any important outline visible in current evidence but absent from the loaded geometry.
- Whether the conclusion is relative between proposals or absolute for a duration.
- The geometry to verify next and the decision that depends on it.
A precise note might read: “Proposal B produces less added afternoon shade than Proposal A under the same loaded OSM context. The nearest southern building uses an estimated height and must be verified before relying on the absolute direct-sun duration.” That sentence is more useful than either “OSM is accurate” or “OSM is unreliable.”
When to replace map context with verified geometry
Move to surveyed or otherwise verified geometry when the result supports a permit, legal assessment, neighbour agreement, property decision or costly design commitment. Also escalate when the selected point sits close to a shadow boundary, because a small height or position error may change whether a 10-minute sample is counted as sun or shade.
Keep the OSM-based study in the project record. It can document the early question, the options tested and the assumptions that led to more detailed work.
Checks to make even when confidence is High
Even a High label needs a site check. Ask whether every building capable of reaching the point has an outline, whether the mapped height describes the present structure, and whether a simple footprint and maximum height represent the roof form closely enough for the decision.
Then ask where the point sits relative to the shadow edge. A result deep inside a long shadow may be insensitive to a small height error. A point that changes from sun to shade when the edge moves slightly needs stronger geometry. Confidence therefore depends on both the source and the sensitivity of the design question.
Finally, record when the context was loaded. A reproducible proposal link preserves the planned scenarios, but surrounding map data is requested again. If the study is reviewed later, note any change in the returned outlines or height sources before comparing reports.
Inspect the data behind the shadow
Load nearby OSM buildings, see mapped and estimated height counts, compare three scenarios, and export an early design report with the limitations attached.
Open Yamaken Shadow Study →Frequently asked questions
Does OpenStreetMap contain building heights everywhere?
No. Mapping detail varies. Some outlines have a height, some have a level count, and some have neither.
Why does the tool use 3 metres per level?
It is a visible approximation used only when a direct height is absent. Real floor-to-floor height can differ by building type, age and use, so verify important buildings.
Can I assume an empty area has no buildings?
No. “No mapped outlines found” describes the returned data, not the physical site. Check current evidence.
