Pawston Annual 2026

In-person, single-day · The Conservatory, Royal Horticultural Halls, London, UK · methodology event-v1.2.0

Total footprint

35.7 tCO₂e

95% range 32 to 39.4

Per attendee

78 kgCO₂e

95% range 69.6 to 85.7

Attendees

460

312 of 450 surveyed

Survey response rate

69.3%

Non-responders extrapolated

Headline with 95% confidence interval

5,000-iteration correlated Monte Carlo

32.035.7 tCO₂e39.4

The band reflects structural uncertainty in the emission factors and the extrapolation for non-responding attendees. Bit-identical replay from the same inputs hash (see Reproducibility).

Open the methodology atlas

The 35.7 tCO₂e headline decomposes across 7 GHG-Protocol scope categories and 26 emission lines. Every line names its factor, source, version, vintage, and the person who recorded it.

· σ-banded on every line and on the total
· Factors with source, version + vintage in plain text; DEFRA 2024 where DEFRA publishes one, marked Estimate where it does not
· Override named to the person who recorded it, with their note
· Double-count guard results explicit on every check
· Replay command on every published figure

Open the methodology atlas →

Scope split

Scope 3 (categories 1, 5, 6, 7)31.8 tCO₂e · 89.2%

Scope 2 (venue energy, allocated)2.7 tCO₂e · 7.7%

Scope 1 (venue gas heating, allocated)1.1 tCO₂e · 3.2%

Per ISO 20121 boundary practice, venue energy is allocated to the event by booked-hours share rather than scoped to the venue. Scope 1 is the venue's gas heating for event hours, allocated the same way. Pawston owns no vehicles, so there is no mobile combustion.

Travel mode mix (top contributors)

Mode	Pax share	Pax	kgCO₂e
Short-haul flight	5.8%	26	10,241
Car (petrol/diesel)	31.3%	141	7,996
Rail	48.0%	216	2,892
Coach	12.2%	55	781

Short-haul flight is 5.8% of ticket holders and the largest single contributor; car (petrol/diesel) is 31.3% and second. The lever sits in catchment + invitation policy, not the day-of programme. Modes cover the attendee lines of Cat 6 (21.9 of 26.0 tCO₂e); speaker and crew travel are separate lines.

How we did this, in three steps

Step 1

Survey

Public attendee-travel survey collects mode, origin postcode and nights away. Unique slug per event, rate-limited.

Survey results →

Step 2

Calculator

DEFRA-pinned factors run through the activity ledger; the structural σ engine bands every record. Non-respondents extrapolated by banded mode-mix.

Methodology →

Step 3

Publication

Snapshot frozen with an inputs hash. Re-runs replay bit-identical. Rolls up into the parent's inventory, category by category, for the reporting year.

Allocation →

Action over accounting.

The footprint is the easy part. The work that moves the number is invitation policy, catchment design, and venue choice, decisions made before the event, not after.

What this demo claims

· Every line names its factor, source, version + vintage
· Every number σ-banded (central + 95% CI)
· Every line named to the person who recorded it
· Inputs hash + MC seed published with headline; bit-identical replay
· Non-responder treatment documented inline with the σ it adds
· Double-count guard runs across every line, with three checks each
· Methodology PDF mirrors the on-screen content from the same snapshot

What it doesn’t claim

· No review or assurance by Hemera: the organiser declares the figures and Hemera ships the method and the pack
· No sector-specific add-on protocols wired in v1
· No factor of its own: DEFRA 2024 where DEFRA publishes one, an estimate marked as such where it does not
· Demo numbers are illustrative for a fictional event, not a real footprint

Sample event is illustrative. Numbers and attendees are fictional. The methodology, σ engine, factor lineage and publication snapshot are the same ones HemeraScope ships in production.
Last synced with the production methodology (event-v1) on 2026-06-10.