How the numbers are made.
Every figure on this site and in Atlas carries its source and date. This page explains where the data comes from, how the models work, and what "planning-grade" means in practice.
Editorial standard
- A figure or date is published only after a person has opened the page it comes from. Search snippets, summaries and recollection are not sources.
- Primary sources first: regulators, government, company filings and press releases, then reputable trade press. Market-research vendors' headline forecasts are reported as marketing, not evidence.
- Author's estimates (pricing, revenue scenarios, indicative costs) are labelled as such and never mixed into a sourced table without a label.
- Corrections are logged on the page they affect. Readers who find an error are asked to write to hello@evtol.nz.
- AI agents draft; people verify. Nothing reaches a client or this site without human sign-off.
Data layers
New Zealand: Toitū Te Whenua LINZ (topography, elevation, buildings, aerodromes, cadastre; most layers CC BY 4.0), Airways and the CAA (airspace and aerodrome information, reused under agreement; Atlas publishes derived screens, not the AIP itself), the CAA aircraft register, Stats NZ (population, dwellings, commuting, tourism), NIWA, MetService and ERA5 reanalysis through Google Earth Engine (climate), lines-company network maps and Transpower (grid), councils and the Department of Conservation (plans, estate), Te Puni Kōkiri and LINZ (Māori land, under data-governance terms agreed with iwi partners). Australia: Geoscience Australia, Airservices, CASA and the ABS.
The five models
1. Site suitability
Multi-criteria screening of candidate vertipads, vertiports and charging sites. Hard constraints (controlled airspace classes and aerodrome protection surfaces, obstacle surfaces, conservation estate, exclusion zones) remove sites; weighted criteria (population and tourist flows within access thresholds, hospital and heliport proximity, zoning compatibility, flood, wind and fog climatology, distance to grid capacity) rank the rest. Geometry follows CASA's AC 139.V-01 vertiport guidance as the regional reference because New Zealand has no vertiport standard yet. Weights are shown and editable in every run.
2. Demand and accessibility
Origin-destination demand is built from population, tourism and patient-flow data, then compared with road travel times to value time savings. Accessibility uses two-step floating catchment methods from the health-services literature, which the founder has applied in peer-reviewed geospatial work. Mode-share assumptions follow published elasticities rather than vendor projections: the Illinois DOT study found fare three to four times more influential than access time and mode share under 3% at US$6 to 9 a mile.
3. Economics and cost-benefit
Capex and opex templates by site type and aircraft, revenue scenarios, standard cost-benefit and equity tables with sensitivity on fare, utilisation, charge time and energy price. For aeromedical cases, cost-effectiveness is expressed in the terms health funders use, with explicit assumptions on time-to-care and outcomes.
4. Energy and grid load
Energy per sector from a performance library (published range, battery and charging figures by aircraft type, flagged where unverified), charger power and sorties per day converted to aerodrome load profiles, then screened against lines-company capacity where available. The only measured New Zealand figures so far come from Air New Zealand's ALIA trial (65 kW chargers; 40 to 60 minutes to recharge; about NZ$20 of electricity for a Wellington to Blenheim sector).
5. Noise and consent
Route noise contours from published aircraft noise profiles and standard propagation, counted against dwellings and sensitive sites, with the district-plan and designation pathway mapped for each site. Outputs are consultation evidence, not noise-compliance determinations.
What planning-grade means
Atlas supports siting, consenting, investment and policy decisions. It does not produce airworthiness findings, obstacle-limitation-surface assessments, instrument flight procedures, noise-compliance certification or legal advice. Where a decision needs those, the report says so and names who provides them. Every model run records its inputs, weights and data versions so a reviewer can reproduce it.
Reproducibility and publication
Model methods are being written up for peer review so that councils and funders can cite them. Method notes, changelogs and data versions are published on this site as they are released.