Sources and what is ours
This page lists every source a figure in Events mode rests on, what we took from it, and how well we were able to check it — and then every threshold and figure that is ours, so nothing we chose can pass for something a guide said.
How to read the status
- Checked — primary source read by us; the figure is reproduced as published, with the page it was read on.
- Reproduced — consistent across council reproductions of the guidance; the gated primary text has not been read by us.
- Assumption — no published UK figure; our own choice, labelled.
The sources
| Source | What rests on it | Status |
|---|---|---|
| Fire safety risk assessment: open-air events and venues (gov.uk, OGL) | Occupant densities (Table 3, printed p.66: standing 0.3, assembly 0.5, dining 1.0, camping 6.0 m² a person); rates of passage 109 / 73; escape times by fire-risk level (Table 4); travel distances to an exit (Table 5); 1.05 m exits, a single exit for up to 60 people, two otherwise; the exit-width table (160 / 200 / 240 plus 15 per 75 mm); exits 45° apart (Figure 18); fire-appliance access — 3.7 m route, 3.1 m gates, 3.7 m headroom, 12.5 t, 16.8 m turning circle; extinguishers — one per 200 m², at least two, one under 90 m², none more than 30 m away | Checked (read 17 and 18 Oct 2026; pages cited by printed folio) |
| SGSA — Guide to Safety at Sports Grounds (Green Guide, 6th ed.), as sgsa.org.uk summarises it | Rates of passage 82 / 66 (the conservative default); 660 persons per turnstile an hour; stewards one per 250 | Checked (the SGSA pages) |
| SGSA SG03 Event Safety Management (2021) | Stewards one per 100 where many children are expected (4.12.c) | Checked |
| Our planning assumptions — entry lane rates | Ticket scan 480 an hour, scan plus bag search 300, open gate 1,200: below the SGSA 660 ceiling. The SGSA ingress page names admission method and screening as lowering the rate but gives no figure for them, and no UK source publishes one | Assumption |
| HSG195 The Event Safety Guide (HSE, 1999; withdrawn in 2014 — HSE's event-safety FAQ, National Archives capture). Cited from The National Archives' copy of HSE's PDF | Hand-wash one per five toilets, never fewer than one per ten (para 519); water one outlet per 3,000 (para 472) and 1 per 750 (para 1009, all-night music events — applying it to camping and alcohol profiles is our extension); first aiders two per 1,000 for the first 3,000 (para 737); the per-female, per-male 50:50 split (para 515); up to 430 tents a hectare (para 665, corroborating the campsite density); stewards 1 per 250 found effective at classical concerts (para 956); entrances sized for the peak of the arrival profile (p.17 para 91) | Checked |
| The Purple Guide (Events Industry Forum) — Sanitation chapter public preview | Toilet baseline for a typical local event (under 6 hours): 1 per 100 female, 1 per 500 male, 1 urinal per 150; at least one accessible toilet | Checked (the public preview). Subscription guide: cited only — no text or table reproduced |
| Aberdeen City Council event guidance (a reproduction of the Purple Guide tiers) | The 6-hour-plus toilet tiers with and without alcohol and food, the campsite tier; the 50:50 split (p.2) | Reproduced |
| HSE L24 Workplace health, safety and welfare: Approved Code of Practice (2nd ed., 2013, OGL); approved for Northern Ireland by HSENI | Crew toilets and wash basins: Table 1 (para 197) and one more per further 25 above 100 (para 200), counted at the most crew at work at one time | Checked |
| DfT Inclusive Mobility (December 2021, OGL) | The accessible parking share to adopt: 6% of capacity plus one per disabled employee (s.8.2). Named on the car-park line; no accessible count is checked against it yet | Not yet adopted |
| DfT Traffic Signs Manual, Chapter 6 (2019), para 11.7.1 | The 1.2 m/s design walking speed route minutes are read at | Checked (council-hosted copy) |
| CAA Safety Sense Leaflet 17, Helicopter Airmanship (v2, April 2024), p.20 | A private landing site at least twice the helicopter's overall length — the 30 m default pad | Checked |
| CAA ORS4 No.1492 (General Exemption E5454) and ORS4 No.1643 | Landing within 1,000 m of an organised assembly of more than 1,000 people; the emergency-services exemption | Checked |
| Terrorism (Protection of Premises) Act 2025 (c. 10), on legislation.gov.uk; Home Office factsheets | The s.3(1) qualifying-event test (800 at one time, staff included, s.3(1)(d)) and the s.3(2) entry-controlled areas; the s.2 premises tiers (200 / 800), not assessed — flagged, not enforced; the duties in the readiness checklist: s.5 procedures, s.6 measures, s.7 security document (the SIA copy within 30 days of a revision, s.7(2)), s.8 co-operation, s.9 notification, s.10 senior individual; Schedule 2 Part 2 paras 5 and 6 (excluded events). Not in force: status as read 18 Oct 2026 and re-checked 19 Oct 2026 — only Commencement Regulations No. 1 (SI 2026/320) and No. 2 (SI 2026/622) exist, and neither appoints a day for the duties | Checked |
| Home Office statutory guidance, The Terrorism (Protection of Premises) Act 2025 (GOV.UK, published 15 Apr 2026, correction slip 6 Aug 2026, OGL) | The implementation period of at least 24 months from Royal Assent (para 2.10); the peak headcount and its methods (paras 5.16–5.20, Supplementary Document A); entry checks at a well-defined, secure perimeter (paras 5.22–5.26); entry-controlled areas (paras 5.40–5.41); the immediate vicinity has no fixed distance (paras 7.13–7.14); invacuation may be impossible outdoors, record why (paras 7.24–7.25); dispersal rather than a fixed assembly point (para 7.36); no security document for the standard tier (para 7.32); a removable bollard closing a service road (para 8.20); co-operation where a council or landowner also has control (para 8.23); a queue can itself be a target (para 8.37). Para 8.22 names the Purple Guide and the Green Guide; nothing from the Purple Guide's text is used | Checked |
| The Terrorism (Protection of Premises) (Notification Requirements) Regulations 2026 (SI 2026/793) | Notify within 14 days of first publicity, or of commencement day for an event publicised on or before it (reg. 3(2)(a)); the notice gives the most people reasonably expected at one time and how it was worked out (Schedule Part 1, paras 8–9). Made 9 Jul 2026; no effect until s.9(1) commences | Checked (not yet operative) |
| SIA, "Martyn's Law and the SIA's role as regulator" (GOV.UK collection, last updated 11 Sep 2026) | Commencement expected in spring 2027. The SIA's final s.12 guidance is not yet published — only the draft, whose consultation has closed (checked 19 Oct 2026) — nor are its templates or scope tool | Checked (expected date: the SIA's own) |
| GOV.UK Crowded places guidance: hostile vehicle mitigation | Stand-off and impact speed come from a site-specific vehicle dynamics assessment, not a fixed figure — which is why the 30 m vehicle-approach reach is ours (checked 19 Oct 2026) | Checked (no figure taken) |
| OpenStreetMap, through the Overpass API — Road data © OpenStreetMap contributors, ODbL (openstreetmap.org/copyright) | The public roads near the site that the vehicle-approach note checks: every road a motor vehicle can be driven along from the public network, from motorway to service road, including pedestrianised streets and busways; not footways, cycleways, paths, bridleways, steps, field tracks, private roads or driveways. Fetched on our server and kept for up to six hours | Open data, ODbL 1.0 — complete only as far as the map is; confirm the frontage on site |
| Licensing Act 2003, ss.100 and 107 | A TEN covers fewer than 500 people (499) including staff; at most 15 TENs and 21 TEN days a premises a year | Checked |
| Building Regulations 2010, Schedule 2 Class 4 | A temporary building intended to remain no more than 28 days is exempt — in the pack; no check reads it yet | Checked |
| HSE HSG175 Fairgrounds and amusement parks: guidance on safe practice (3rd ed., 2017, OGL) | A ride's footprint includes its barrier line (para 242); no fixed gap between rides (para 182); a powered ride's Declaration of Operational Compliance before use (paras 74, 142 and 187); no ride inspection for a stall (Table 2) | Checked |
| HSE GS6 Avoiding danger from overhead power lines (4th ed., 03/13, OGL) | The overhead-line check (Rides: what is checked): a safety zone 6 m horizontally from the nearest wire on either side, widened on the line owner's advice or to allow for a jib or other moving part (para 19, printed p.6); the 10 m at ground level from below the nearest wire within which the risks of any work must be managed (para 11, printed p.4); its application to temporary structures and the line owner's own separation under ENA TS 43-8 (paras 26–27). No voltage-varying figure is taken: GS6 para 25 quotes the ENA exclusion zones and that page has not been read, so no ladder appears anywhere | Checked (read 19 Oct 2026) |
| gov.uk Pleasure fairs: model byelaw 4 (first published 8 Feb 2013, updated 7 Sep 2018, OGL) | The hours band quoted in the licensed-hours prompt: it should not be necessary to allow a fair to open before 10am, a manager should not be required to close before 9pm nor be allowed to stay open after midnight (guidance notes para 8). Byelaw 3 itself leaves the hours blank for the council to insert. It is guidance to councils about what a byelaw should impose, never a limit on an event, and no check measures against it | Checked (read 19 Oct 2026) |
| Welsh Government model byelaws set 4, Pleasure fairs and roller skating rinks (2015) | Fire-appliance access to within 50 m of any structure (clause 5(i)) — the only figure in the second built-in rule pack. A model template, not law anywhere until a council adopts it; the English model set (2018) does not contain the clause, and we have not checked whether any adopted byelaw still carries it. The clause's 4 m route width, 4.5 m headroom and 12 t are deliberately not shipped: no rule reads them | Checked (read 19 Oct 2026) — a template, not in force |
| Public Health Act 1961 s.75; Civic Government (Scotland) Act 1982 s.41; Licensing Act 2003 Sch.1 Part 1 para 2; GPDO 2015 Sch.2 Part 4 Class B (legislation.gov.uk, OGL) | Who sets a funfair's hours (byelaws in England and Wales, public entertainment licence conditions in Scotland); that the rides themselves are not licensable entertainment; and that the permitted-development route limits the land to 28 days' use in a calendar year and says nothing about height — which is why there is no height check. Northern Ireland is not covered and says so | Checked (read 19 Oct 2026) |
| UK Civil Aviation Authority — crane notification, obstacle lighting (ANO 2016 art. 222) and aerodrome safeguarding | Not read — the pages were never opened, so nothing affirmative about the CAA is claimed anywhere: the height-check explanation says only that separate notification routes exist for tall structures and for cranes, that we have not read them, and that you should ask the CAA and your local aerodrome. © UK CAA with no Open Government Licence statement, so it is cited by existence and no figure from it is quoted anywhere. The legal point the explanation does make is carried by ANO 2016 art. 222, which was read. We hold no aerodrome or safeguarding dataset | Cite-only — nothing reproduced |
| NFDC DRG115:2019, Temporary Works Hoardings guidance notes (National Federation of Demolition Contractors), §5.6 p.13 and §7.2 p.14, read 20 Oct 2026 | The Privacy screen adds wind load note: netting, signs or an acoustic barrier on temporary fencing can raise its wind load enough to change its classification, and the fence may then need more ballast, stabilising supports or rake-back bracing; a 2 m fence with netting in an exposed location is a risk class higher than without it. Construction-site guidance: applying it to an event is ours. Paraphrased, never quoted at length (the NFDC asks to be approached to reproduce it), and no wind speed from it is repeated | Industry guidance · Assumption |
| Temporary Works Forum, Lightweight temporary open mesh fencing — design considerations (TWf2024:01) | Named in the privacy-screen note as the UK guidance written for this fencing. Behind a sign-in: not read, and nothing is taken from it | Cite-only — unread |
| BS 7909:2023+A1:2024, Temporary electrical systems for entertainment and related purposes | Named in the generator-capacity line as what temporary distribution follows. A paid standard: cited by its existence only, unread, and nothing is claimed about what it says | Cite-only — unread |
| HSE, Bouncy castles and other play inflatables: safety advice | A PIPA tag or ADIPS Declaration of Compliance on each play inflatable; no use in winds above 24 mph (38 km/h) | Checked |
| ADIPS (adips.co.uk) and PIPA (pipa.org.uk) | The two inspection schemes the ride table's reminder names. A ride's DOC and an inflatable's tag can be checked on their sites | Checked (industry) |
| gov.uk VAT rates | The 20 % standard rate on the budget total | Checked |
| AGF Festival Sustainability Insights 2024 (PDF), p.33 | 0.5 kg per person per day at a day event (range 0.1–1.6), 1.4 kg camping (range 0.8–3.4); 40 festivals in 16 countries | Benchmark · Checked |
| Council-filed traffic management plans (Sevenoaks DC) | The parking defaults — 2.5–3 people per car, 80 % by car for rural sites — and the 25 m² gross per car | Assumption: enter your own TMP figures |
| Lichfield DC event management plan, section 6 | Applies one steward per 100 where alcohol sales play a key role, crediting the Green Guide — the basis for our alcohol trigger | Assumption (secondary) |
| Hastings BC toilets guidance; SWAST / North Devon event medical guidance; Isle of Man fire service campsite slides and fire prevention note 7 | Hastings prints a mixed-use accessible figure of 1 per 40 with the unit unstated (not adopted). SWAST's tier table is a superseded edition: it copies the Purple Guide medical table removed in December 2024 (UKHSA scoping review, February 2026, pp.25–26). The Isle of Man campsite slides are where the campsite note's 3 m / 6 m / 4 m spacings come from. No figure rests on them | — |
| MUTA best-practice guide (marquees) | Marquee exits and spacing — not yet modelled | To verify: obtain and transcribe |
| Environment Agency, Flood Map for Planning — Flood Zones 2 and 3, rivers and sea (OGC API Features; OGL v3) | The flood zones of an English site (Site conditions): present day, ignoring defences. The service carries no date; our register records the revision (20 May 2026) and when we last checked it | Open data, read live by our server |
| Natural Resources Wales, Flood Map for Planning — rivers and sea (DataMapWales WFS; OGL) | The flood zones of a Welsh site, which include an allowance for climate change and assume no defences; dated by the features' own publication date | Open data, read live by our server |
| SEPA Flood Maps v3.0, river and coastal (OGL 3.0) | The High, Medium and Low likelihood bands of a Scottish site (10-, 200- and 1,000-year return periods); how they map onto the checks is ours | Open data, read live by our server |
| ONS Countries (December 2025) Boundaries UK BFE (OGL v3); postcodes.io (ONS Postcode Directory) as a fallback | Which nation each part of the site is in, so each part is judged by its own nation's map | Open data; the boundary is held on our server |
| LandIS Soilscapes — NATMAPsoilscapes, Cranfield University (LandIS Open Licence, Public Data; downloaded 2026) | The soil texture and drainage class in England and Wales, from our 100 m grid of the 1:250,000 map | Open Licence, Public Data only; the platform is never queried live |
| BGS Soil Parent Material Model (1 km) and peat layer (OGL v3) | The parent-material texture and peat in Scotland, and in England and Wales wherever Soilscapes gives no class | Open data, read live by our server |
| HSE event safety, "Transport" (updated 4 Sep 2025) | Hardcore, metal trackway or temporary surfaces for routes that deteriorate in bad weather — the trackway control on slow-draining ground | Checked (the control); applying it to a soil reading is ours |
| NPPF Annex F, flood risk vulnerability classification (England) | Holiday or short-let camping is "more vulnerable"; it says nothing about temporary events, so applying it to an event campsite is our reading | Checked (England only — TAN15 and NPF4 not read) |
| GOV.UK flood guidance for campsite and caravan park owners (updated 18 Sep 2023) | A flood plan with a site map, flood warnings and evacuation routes — the control the campsite warning asks for | Checked (aimed at campsites, not events) |
| Met Office UK and regional series (areal values from HadUK-Grid 1 km) — Met Office, OGL v3 (landing page) | The typical weather in the event's month (Site conditions): rainfall, rain days ≥ 1 mm, mean daily maximum and minimum temperature, mean temperature, sunshine and air-frost days, as a 1991–2020 monthly average — never a forecast. The files carry no wind, no 10 mm rain days and no summer days, and each is shown as "not in this source". The files are themselves HadUK-Grid 1 km averaged over a region, so the limitation is how local it is | Checked (read 20 Oct 2026); the 1991–2020 monthly mean, the units, the rounding and the region mapping are ours |
HadUK-Grid 5 km monthly 1991–2020 averages (v1.3.2.ceda) — Met Office / CEDA, OGL v3, DOI 10.5285/a7947ae607fd4be8994b940eebca8872 | Not in use. It would give a 5 km cell over the site instead of a regional mean, with wind, 10 mm rain days and summer days. The nine monthly files are confirmed by name and size, but reading them needs a registered CEDA account and an access token, and the licence requires the catalogue record's own citation, which has not been read. Nothing in the app uses it, and no citation is composed from the DOI | Licence read 20 Oct 2026 (the archive's own README); no data in hand |
| Met Office National Severe Weather Warning Service | Cited by existence only, in "Not assessed": live warnings are linked, never computed. This app builds no forecast feature | Cite only — nothing is read from it |
| HSE event safety, "Outdoor equipment to support speaker clusters, lights etc" (updated 12 Oct 2024, OGL) | The exclusion zone "in the order of 1.5 times the tower height", where no adverse-weather arrangements are in place; and the wording of the wind note — take account of foreseeable wind loading, ground strength and slope; monitor and measure local conditions where the structure is susceptible to the weather; in adverse conditions know what to do with the equipment. The page publishes no wind speed of any kind | Checked (read 20 Oct 2026); applying the multiplier as a drawn circle against live zones is ours |
| IStructE Temporary demountable structures (4th ed.); BS EN 13782:2015; the MUTA guide | Named in the wind note as where the design and operation of temporary structures is set out. Not read — paywalled — so no figure, wording or action level is taken from any of them, and none is attributed to HSE | To verify: obtain and transcribe |
Figures we could not check — medical cover above about 3,000 (current industry guidance is by subscription and no longer publishes numbers), the stewarding ratio of one per 100 for alcohol sales, the scan and search lane rates — are badged Assumption or noted as not modelled rather than filled in from memory.
What is ours
Every one of these is labelled Assumption where it appears, can only warn in the checks, and is yours to replace — with a loaded rule pack, a loaded rate card, or the figures in your own plans.
Rule-pack thresholds that are ours
| Threshold | Value | What it does |
|---|---|---|
| gate_zone_tol_m | 2 m | A gate within this of a zone edge is one of that zone's exits |
| fire_gate_tol_m | 3 m | A gate within this of where a fence crosses the emergency route serves that crossing |
| tent_fp_tol_m | 5 m | A fire point inside a tent's footprint or within this of it serves the tent |
| heli_clear_m | 10 m | No structure, obstacle or crowd zone within this of the pad edge |
| gate_fence_tol_m | 1 m | A gate within this of a fence run opens it in the walking graph |
| node_offset_m | 0.6 m | A shoulder's width: how far off a barrier the walker rounds its corner |
| node_cap_n | 600 nodes | The walking graph's budget on an iPad; a capped graph says so |
| travel_sample_m | 5 m | Sampling step along an enclosure boundary for its farthest point |
| queue_m2pp | 0.5 m² a person | The queue's holding density — the guide's assembly figure used as a holding-pen density |
| route_warn_carpark_entry_m | 1,000 m | Comfort length, car park to entry (IHT "Journeys on Foot" Table 3.3 gives 800–1,000 m for parking — a town-centre table) |
| route_warn_camp_arena_m | 800 m | Comfort length, campsite to arena |
| route_warn_entry_arena_m | 400 m | Comfort length, entry to arena |
| route_warn_arena_toilets_m | 250 m | Comfort length, arena to the nearest toilets |
| route_warn_arena_medical_m | 300 m | Comfort length, arena to the medical post |
| route_detour_warn | 1.8× | A walk this many times the straight line is worth a second look |
| route_detour_floor_m | 25 m | The detour ratio is judged only once the walk is this much longer than the straight line |
| hvm_review_m | 30 m | A vehicle line — a trackway, the site entrance, a car park or a public road from OpenStreetMap — within this of a crowd zone is an approach the Martyn's Law readiness check asks to be protected. The guidance gives no distance (para 7.13), and GOV.UK's hostile vehicle mitigation guidance sizes stand-off by a site-specific assessment |
| safer_m2pp | 0.3 m² a person | The most people a safer area can physically hold: the gov.uk open-air guide's standing occupant-load figure, used as a ceiling. The Martyn's Law guidance gives no shelter density, so this use is ours and is not a shelter standard |
| flood_samples_max | 4,000 samples | The most sample points laid over one zone to estimate its flood shares — the shares are a grid estimate (outlines simplified to about 2 m), printed to the nearest per cent |
| site_refetch_m | 1,000 m | How far the site may move before its point readings (the soil) are stale and the panel asks for a fresh fetch |
| ground_pad_m | 50 m | The margin added round the site boundary for the flood query box |
| ride_overlap_tol_m | 0.05 m | The inset applied to each ride footprint before the two are tested for interpenetration, so footprints that touch or share a barrier line are not reported as overlapping. No source publishes a figure of any kind here — this one exists only to keep a drawing's result stable, and the finding says it is ours. There is no ride-to-ride gap figure at all; see Rides: what is checked and what is not |
| safer_host_tol_m | 1 m | How far a corner of a safer area may sit off the building or hard-sided structure its centre lies on before it "runs off its structure" and does not count. The guidance gives no figure (paras 7.24–7.25); the tolerance is ours, for hand drawing |
Other figures that are ours
- The emergency route "reaches" a zone when it ends within 10 m of the zone boundary.
- The arrival-curve shapes (Peaked 60 % in the busiest two hours, Early, Even, Late) — the guidance publishes no profile.
- The parking estimate — share by car, people per car, 25 m² gross a car — and the grass car-park field defaults (2.5 × 5.0 m bays, 6 m aisles, 90°, nothing built). The accessible share of event parking is not checked yet: Inclusive Mobility's 6% is named as the basis to adopt.
- Stewards one per 100 for alcohol sales with no children (SG03 names children, not alcohol); the ticket scan, scan plus bag search and open gate lane rates (480, 300 and 1,200 an hour); the accessible-toilet placement guide of one per 20 units.
- Every hire rate on the built-in card (47 of them), the five stage presets, and every object's footprint (typical UK hire dimensions — no standard governs them).
- The clearance for a temporary helicopter pad (10 m); CAP 1264's hospital-site safety area is cited in the basis for comparison, not adopted.
- A safer area counts only inside an existing building on the site or a hard-sided structure, with every corner within 1 m of it, and which library objects count as hard-sided (cabins, containers, the ticket kiosk — never a tent or marquee, and never the control room or the lost-children cabin). Both rules are ours (user decision, 18 Oct 2026; the control room and lost-children cabin taken off 19 Oct 2026).
- Which OpenStreetMap ways count as public roads for the vehicle-approach note, and the suggested barrier line it offers for a road frontage — a sketch for you to confirm or redraw, never drawn for you.
- The SEPA band mapping: High or Medium likelihood in the higher role (with Flood Zone 3), Low in the lower — SEPA's bands are not flood zones, and every Scottish line says "(our mapping)".
- The texture and waterlogging mapping: reading LandIS drainage classes (impeded, slightly impeded, naturally wet, surface wetness) and BGS clay, silt or peat as a waterlogging risk, freely draining classes and sand or chalk as free-draining, and loam as intermediate; the Soilscapes texture words mapped to clay, sand, loam and peat.
- The peat warning — "low bearing capacity likely — structures and vehicles" — and the reading of an event campsite as NPPF's "more vulnerable" camping use.
- The 5 km limit on the flood check, and our 100 m grid of the LandIS map (a site near a soil-unit boundary may read the neighbouring class; the point is placed on that grid by the OS Helmert transform, to about 4 m).
- The Fairground / funfair event type's seven defaults (3 days, an 8-hour opening, alcohol and food on, no camping, children expected, open gate, 60 % by car) — no open UK source publishes any of them. "Alcohol & food" on is the stricter of two existing toilet rows, so no sanitation figure is new either way. See Rides: what is checked and what is not.
- In the ride checks: that the probe depth outside a ride's entrance or exit is the opening's own width (derived from the design, not a safety distance — no source publishes a depth); that no diversity factor is applied to the rides' connected load, because none is published for a fairground. There is no ride separation distance, no queue or viewing allowance and no height limit — none exists to adopt, so none is shipped.
- The Martyn's Law readiness checklist itself: which drawn objects show each duty on the plan is our reading of the Act and the guidance, and the checklist is a readiness aid, not the security document.
- The 1991–2020 monthly mean of the Met Office regional series, computed by us from the published yearly rows, with a month missing more than two of its thirty years shown as not available rather than averaged. The Met Office publishes the yearly table; the monthly normal in the app is ours.
- MO_REGION_OF — the table that maps the ONS region of the nearest postcode onto a Met Office series. The Met Office regions are not ONS regions and no boundary file for them is published, so the mapping is ours; two Met Office regions straddle the England–Wales border and a Welsh site is always given the Wales series. Every output says which series was used and how it was chosen.
- The units on every weather figure (mm, days, °C, hours). The Met Office files name the quantity and never the unit, and the series landing page states none, so the units are ours — as is the rounding of each figure.
- Taking the weather month from the event date, and saying so when the calendar straddles two months. With no date there is no reading at all, never a default month.
- How
tower_excl_xis applied. The 1.5 × height figure is HSE's, and so is its condition (where no adverse-weather arrangements are in place); drawing it as a circle, measuring it from the height of the object you placed, and testing it against the crowd zones that are live are ours. It is always a note, and it never states a wind speed — each structure's operating wind speed is its supplier's or manufacturer's figure.
Licensing. The Purple Guide's terms forbid commercial reuse and database creation, so no text or table from it appears in the app — only citations to the chapter and figures from its public preview. The gov.uk guide and the legislation are Open Government Licence; the SGSA pages, HSG195 and the council documents are cited as what they are. OpenStreetMap road data is under the Open Database Licence: wherever a road from it is named — the vehicle-approach note, the readiness check, the appendix — the credit "Road data © OpenStreetMap contributors, ODbL" goes with it. The Met Office series is Open Government Licence v3 and asks for its own credit, "Contains public sector information licensed under the Open Government Licence v3.0", which is printed wherever a weather figure shows.
Site data credits
The flood, ground and weather readings (Site conditions) rest on open data, each under a licence that asks for its own credit. They are printed here, in the Site conditions block whenever that source's reading shows, and in the appendix's provenance:
- LandIS Soilscapes: Soil data from the LandIS Portal (Cranfield University): NATMAPsoilscapes, 31 Mar 2026, Public Data 2026, Open Licence. Licence: https://cranfield-portal.maps.arcgis.com/sharing/rest/content/items/91f00fd749314faa8356f9e41b65ff8c/data LandIS and Cranfield University do not endorse this tool or its readings. Only LandIS Public Data is used, never Restricted Data, from an offline download from the LandIS Portal; nothing in the app queries the LandIS platform. If the licence were withdrawn, the Soilscapes reading stops and BGS parent material carries the ground in England and Wales too.
- Environment Agency (flood zones, England; OGL v3): © Environment Agency copyright and/or database right 2025. All rights reserved.
- Natural Resources Wales (flood zones, Wales; OGL): Contains Natural Resources Wales information © Natural Resources Wales and database right. All rights reserved. Some features of this information are based on digital spatial data licensed from the UK Centre for Ecology & Hydrology © UKCEH. Defra, Met Office and DARD Rivers Agency © Crown copyright. © Cranfield University. © James Hutton Institute. Contains OS data © Crown copyright and database right.
- SEPA (flood maps, Scotland): ©SEPA 2025; this SEPA product is licensed under the Open Government Licence version 3.0. SEPA asks that its Flood Map Version 3.0 data attribution statements (5 February 2025) be included on reuse; the river and coastal sets are printed in full in the appendix's provenance with every Scottish reading.
- ONS (countries boundary): Source: Office for National Statistics licensed under the Open Government Licence v.3.0. Contains OS data © Crown copyright and database right 2026.
- BGS (parent material and peat; OGL v3): Contains British Geological Survey materials © NERC 2026.
- Met Office (the typical weather for the month, from the Met Office UK and regional series (areal values from HadUK-Grid 1 km); OGL v3): Contains public sector information licensed under the Open Government Licence v3.0
- HadUK-Grid (DOI
10.5285/a7947ae607fd4be8994b940eebca8872; OGL v3, Permitted Use "Any"): not in use in this build. Its licence requires the catalogue record's own citation, which has not been read, so no reading uses the grid and no citation is written here.
Crowd simulation (Beta)
The figures behind Crowd simulation. Each one carries its own status in the panel and the report — Checked only once the page it was read on is recorded — and a paywalled work is cited by its existence only, never quoted.
| Source | What rests on it |
|---|---|
| Weidmann, Transporttechnik der Fussgänger (ETH Zürich IVT Schriftenreihe 90, 2nd ed., 1993) | The speed–density relation: free speed 1.34 m/s, jam density 5.4 persons per m², and its shape constant. The constants are cited as facts with the page they were read on; no table is copied. The scan carries no licence statement, and the ETH Research Collection record's rights statement has not been read (the page failed to load), so we claim no permission for it |
| RiMEA 4.1.1 guideline (CC BY-ND 4.0) | The pre-movement cases (rapid, speedy, slow), the congestion speed, and the test cases of the verification appendix. The 8-minute figure it quotes from EN 13200-1 is recorded as an alternative escape-time basis and is not shown anywhere in S1a |
| IMO MSC.1/Circ.1533, revised guidelines on evacuation analysis (2016) | The SFPE passage rate (78 persons per metre per minute) and test cases of the verification appendix |
| NIST Technical Note 1822 (Ronchi et al., 2013) | Test cases of the verification appendix (blocked exit, maximum flow) |
| Seyfried et al., "New insights into pedestrian flow through bottlenecks" (arXiv physics/0702004) | Why a gate fed from a wider area can reach the passage rate even though an open corridor cannot |
| SGSA Worked Example A (football and rugby) | The Green Guide's emergency evacuation time range, recorded as an alternative escape-time basis — not shown anywhere in S1a, and never set against a simulated time |
| Cabinet Office / Emergency Planning College, Understanding Crowd Behaviours (2009) | The published density reference lines, and the limits text's disclosure that the model has no crowd-pressure, crush, smoke or behavioural-panic mechanics |
Crowd simulation figures that are ours
| Figure | Value | What it does |
|---|---|---|
| crowd_cell_m | 1 m | The grid the Flow model runs on |
| crowd_dt_s | 0.5 s | The time step on the 1 m grid |
| crowd_dt_fine_s | 0.25 s | The time step of the half-size grid sensitivity check |
| crowd_queue_cap | 3.3 persons per m² (or 2, 4, 5.4) | The queue density: how tightly a waiting crowd packs in the model. A setting, not a prediction |
| crowd_route_alpha | 0.1 | How quickly congestion-aware routing reacts to a busy exit, so the whole crowd never switches exit at once |
| crowd_recompute_s | 5 s | How often congestion-aware routing is recomputed |
| crowd_recompute_big_s | 10 s | The same, on a large area |
| crowd_recompute_big_cells | 200,000 cells | What counts as a large area |
| crowd_clear_ppl | 0.5 person | An exit or an enclosure is clear when fewer than this are still to leave |
| crowd_clear_dens | 0.05 persons per m² | The Clearance map's threshold: when each cell last held this or more |
| crowd_cap_share | 95 % | An exit counts as running at capacity at this share of it or more |
| crowd_frame_q | 0.025 persons per m² | The step the timeline's density frames are stored in |
| crowd_keyframe_s | 30 s | How often the run's state is fingerprinted for the run log |
| crowd_cell_limit | 1,000,000 cells (desktop), 400,000 (iPad) | The largest area the model runs at 1 m |
| crowd_frame_budget | 48 MB (desktop), 12 MB (iPad) | The memory the timeline's frames may use; past it, frames are stored less often and the run log says so |
| crowd_window_max_s | 1 hour | The longest run; anyone still inside then is counted as not out |
| crowd_site_split | In proportion to capacity | A whole-site run shares the attendance among the crowd zones by their capacity, each zone at most full; people not in any crowd zone are not modelled, and the result says how many |
| crowd_arrival_linear | Piecewise-linear, if chosen | The day profile's alternative arrival shape within each hour, keeping each hour's total; the default holds each hour's rate constant |
| crowd_departure_end | Nobody leaves before the gates close | The day profile's departure shape ('end'): the people on site are everyone who has entered |
Licensing. The Purple Guide's terms forbid commercial reuse and database creation, so no text or table from it appears in the app — only citations to the chapter and figures from its public preview. The gov.uk guide and the legislation are Open Government Licence; the SGSA pages, HSG195 and the council documents are cited as what they are. OpenStreetMap road data is under the Open Database Licence: wherever a road from it is named — the vehicle-approach note, the readiness check, the appendix — the credit "Road data © OpenStreetMap contributors, ODbL" goes with it.
Try it on your own site
Turn a boundary into an optimised UK layout in minutes — free to start.
Open the app →