Sources and limits
Measured
Pedestrian passages from City of Helsinki LIDO–TIKU, counters 100016673 (2_Forumin tunneli), 100016667 (4_Aleksanterinkatu) and 100040158 (Eteläesplanadi), source EcoCounter, CC BY 4.0. Embedded snapshot retrieved 23 August 2026; latest common observation 5 July 2026. Building footprints are mapped geometry from City of Helsinki open geographic data (CC BY 4.0). Heights are interpretive. The 624 footprints describe 621 buildings — three of them are mapped in two parts.
Source-geometry artefacts. The extract carries one water polygon twice, with identical rings (Opaskartta_alue.16235-0 and .16691-0); the duplicate is dropped when the page loads. Nine footprints repeat a vertex and six more contain collinear points, which makes them non-simple in the strict sense. None contains a proper self-crossing, so all of them fill and hit-test correctly here; they are named because a stricter renderer might not be so forgiving.
Re-verified against the live API on 1 September 2026. All three counters exist under those ids with those names, and their published coordinates place them within a few metres of where this map draws them. Every one of the 21 counter-year figures reconciles against the API's monthly aggregates: the January–June sums plus between two and a half and six days of July, measured against each July’s own daily mean — which is what a window ending 5 July should look like. The 2026 figures match the API total exactly, and that total has not moved since the August snapshot — these counters have recorded nothing since early July.
Where the measurement fails
The Eteläesplanadi counter is missing whole months inside the comparison window in three years: February and March 2021 (396 and 494 passages against month medians of about 8,000 and 10,000), May 2023 (6,320 against a median of 54,535), and January and February 2024 (545 and 197). Filling May 2023 from the other years puts that window at roughly 157,100 rather than the recorded 108,923 — level with 2022 and 2024 instead of a third below them. The 2023 dip on this counter is mostly a dead sensor, not a change in how many people walked there. The recorded figures above are left exactly as the API returns them and are marked gap in the table; the modelled Esplanadi density index for 2023 is the one number on this page corrected for it, set to follow its neighbours at 51 instead of the 35 the raw window implies. The Forum tunnel and Aleksanterinkatu counters have no gaps inside the window, though both lose most of September and October 2025 outside it.
Modelled — how the purpose split is built
A pedestrian counter records one number per hour. It carries no trip purpose and no origin, so every category on this map is a model, not a measurement. The split is built in three steps, from published sources:
- A city-wide baseline. Helsinkiläisten liikkumistottumukset 2023 (Kaupunkiympäristön julkaisuja 2024:9; 3,061 phone interviews, Sept–Oct 2023) gives the trip purposes of Helsinki residents: leisure 38%, work 23%, shopping 15%, education 10%. The remaining 14% covers errands, business trips and other purposes, which the report does not break out — this model allocates it rather than reading it off the page.
- A centre-specific reweighting. The 2026 centre vitality count (published 11 May 2026) counts 2,995 premises, of which 2,674 are occupied and classified — that is the denominator for its sector shares, and it is not the 2,663 the same report gives as palvelevat liiketilat, serving premises. Against those 2,674 it puts services at 64.0%, restaurants alone at 759 units or 28.4%, and retail at 790 units or 29.5%. Between the 2023 and 2026 counts restaurants gained 33 units (+4.5%) while retail lost 33 (−4.0%), and vacancy fell from 9.72% to 7.88%. That sets the drift between the leisure and shopping shares from 2023 onward; the 2020–22 shares are a pandemic assumption, not a measured series. Two categories a resident travel survey cannot contain are then added for the centre: tourism, and transfers and through-movement.
- A per-zone profile. Each density area gets a 2026 reference profile from what is actually there — the retail spine at Stockmann and Keskuskatu, the interchange load at Rautatientori and Kamppi (Helsinki Central Station is commonly cited at around 200,000 users a day), the university quarter at Senate Square, the market, ferry and cruise traffic at Kauppatori. Year factors then move each purpose across the window: tourism collapses in 2020–21 and recovers to a record, work falls with remote work and returns to near pre-pandemic, education follows remote teaching. The shape of each of those curves is asserted from the documented direction of the trend, not fitted to a published annual series — across the full 2020–26 window, only the three counters are measured.
Cross-check, and where it fails. Removing tourism and transfers from the modelled 2026 centre mix and renormalising gives work 27%, shopping 30%, leisure 32%, education 12%. The travel survey's four published categories, renormalised the same way, give 27 / 17 / 44 / 12. Work and education land within a point. Shopping runs about thirteen points high and leisure about thirteen points low — the direction a retail and hospitality core should move in, since city-wide leisure counts a great many neighbourhood trips that never reach the centre, but a large adjustment resting on a premises count rather than on any centre-specific travel survey. It is the least secure assumption in the model.
What is measured, and what is not. Of everything on this page, exactly three things are measurements: the building footprints, the three counter series, and the counters' positions. Building heights, the density field, the six purposes, the origin split and every year factor are interpretation.
Tourism and origin
Tourism weight and its 2020–26 trajectory follow Helsinki's 2025 tourism record: 4,924,251 registered overnight stays, up 8.2% on 2024, international stays up 17.4%, 175,000 international cruise passengers, and a 5–7% growth outlook for 2026. Origin shares are not measured at all: they are derived from the purpose mix through a fixed purpose-to-origin matrix, and the matrix itself is asserted. The only external check on it is Helsingin keskustan asiointiselvitys 2019, whose main sample of 1,694 respondents (1,206 online and 488 by telephone; a street poll of 300 was run separately and is not the source of this figure) put 73–74% of centre trips on public transport — a mode share, not an origin share, so it supports the direction (most presence in the centre arrives from outside it) without calibrating the numbers. This is the oldest source on the page and the weakest leg of the model.
A second measurement, and where the two disagree
The city publishes a separate centre footfall series (28 May 2026). Its headline — centre visitor numbers 3.6 percentage points above 2025 for January–April, still below 2019 — comes from mobile-subscription data. The same release separately reports the Aleksanterinkatu measuring point up 4.9% for January–April, without saying which instrument that figure came from.
Compared like with like, the two do not agree. This page's Aleksanterinkatu counter recorded 964,709 passages in January–April 2025 against 1,058,168 in the same months of 2026: +9.7%, not 4.9%. LIDO–TIKU lists only one EcoCounter station on Aleksanterinkatu, so this is not two different sensors, and since both figures cover exactly January–April it is not the length of the window either. The gap is unexplained. It is why nothing here treats the city's number as confirming this page's.
Pandemic restrictions, construction and missing observations affect comparability across the window. Density figures are a relative 0–100 index, not a headcount; each purpose layer is normalised against its own strongest zone-year, so 100 means "the busiest this purpose gets anywhere in the window", not a population. Selecting an origin scales the field down rather than rescaling it, so a dim map means that slice is genuinely a small part of the whole.
Colour
Measured counters are drawn in one ink family and separated by line pattern, with identity carried by direct labels rather than by hue. The six purposes use a categorical order validated for colour-vision deficiency separation against this white surface; three of the six sit below a 3:1 contrast ratio on white, which is why every purpose is labelled in text wherever its colour appears. Magnitude within a purpose is carried by lightness on that purpose's own single-hue ramp.