Data sources & Methodology

Plot summaries

Light / heavy vehicle counts show the 7-day rolling median change in light / heavy vehicle counts in 2026 relative to the equivalent day in 2025, aligned by day of week and week number to preserve weekly patterns (e.g. weekdays vs. weekends). Each day’s vehicle count is normalised by the 9–22 February median which provides a baseline for monitoring the impact of the fuel supply disruption. Only those counters with data in the baseline period for 2026 are used in the analysis. Due to data latency there is a delay of 7 days to 1 month in updates for individual sites.

Heavy commercial VKT shows the daily total vehicle kilometres travelled by heavy commercial vehicles if the trip goes through any of 5 key freight areas: Auckland, Wellington City, Christchurch City, Taupo District, Horowhenua District territorial authorities.

PT boardings - Auckland shows the total daily boardings on Auckland public transport for bus, rail and ferry. Totals are calculated from payment tag on with either HOP or contactless bank cards.

PT boardings - Wellington shows the total daily boardings on Wellington regional public transport for bus and rail. Totals are calculated from payment tag on with Snapper cards as well as contactless payments.

PT boardings - Canterbury shows the total daily boardings on Canterbury regional public transport for bus and ferry. The daily boardings include all ticket types including Motu Move, SGC, transfers etc.

Aviation departures - scheduled services (inc. freight) shows the scheduled flights by airport and includes both domestic and international flights.

Detailed methodology

Light / Heavy vehicle counts

These plots show the rolling median of the regional weighted-mean change in Traffic Monitoring for State Highways (TMS) counts relative to the pre-conflict baseline (9–22 February 2026). TMS data is retrieved regularly from NZTA’s public API. The data processing methodology is as follows:

  • Only Continuous and National monitor types are used. The input data is provided by vehicle weight and direction, and this is retained throughout the processing.
  • For each site-direction and vehicle weight, compute the daily traffic ratio (2026 / 2025) by matching days on week-aligned day-of-year, so that e.g. Monday week 5 in 2026 is compared to Monday week 5 in 2025.
    • The week-day alignment does not account for holidays occuring at different times/days of the year (e.g. Easter)
  • Normalise each site’s ratio by its median ratio during the baseline window (9–22 February 2026).
  • Remove outlier sites whose baseline ratio falls outside the 5th–95th percentile nationally within each vehicle weight.
  • Aggregate normalised ratios across sites by region using a count-weighted mean.
  • Apply a 7-day trailing rolling median to smooth day-of-week variation.
    • Any increases or decreases can be seen up to 7 days after the event in the smoothed line due to the seven day look back period (e.g 2026 Easter period was from 3-6 April, but the point for 6th April is included in the 7-day smoothing until April 12th)

Heavy commercial VKT

These plots show the rolling mean of total vehicle kilometres travelled (VKT) of heavy commercial vehicles as represented in the CompassIoT data (around 90% of the heavy fleet). Full trip VKT tallied if it includes travel within any of 5 key freight areas: Auckland, Wellington City, Christchurch City, Taupo District, Horowhenua District territorial authorities.

Auckland - bus + rail + ferry

  • Boardings for bus, rail and ferry.
  • Boardings represent the number of individual passenger trips, i.e. a bus to bus transfer counts as 2 boardings.
  • Does not include untagged rail (i.e. Cash/Supergold). Untagged bus boardings are included.
  • Counts for 23-26 March 2026 are estimated due to lack of data availability.

Wellington - bus + rail

  • Boardings for bus and rail only. Ferry trips are not included.
  • Boardings represent the number of individual passenger trips, i.e. a bus to bus transfer counts as 2 boardings.
  • Does not include untagged rail (i.e. Cash/Supergold). Untagged bus boardings are included.
  • Does not include trips where buses replace trains. The Metlink website provides current (and next month) days of bus replacements by line e.g. https://www.metlink.org.nz/assets/Buses-replacing-trains/Buses-replace-trains-HVL-2026-Apr-May.pdf.

Canterbury - bus + ferry

  • Boardings for bus and ferry.

PT boardings - year on year analysis

  • Sum the total boardings across modes for each region
  • Compute the boarding ratio by matching days on week-aligned day-of-year, so that e.g. Monday week 5 in 2026 is compared to Monday week 5 in 2025.
    • The week-day aligning does not account for holidays occuring at different times/days of the year (e.g. Easter)
  • Normalise each region’s ratio by its median ratio during the baseline window (9–22 February 2026).
    • This adjusts for pre-existing trends. e.g. if Wellington boardings were already running below 2025 levels before the crisis, dividing by a baseline ratio of less than 1 scales the post-crisis numbers back up, isolating the fuel crisis’s effect.
  • Apply a 7-day trailing rolling mean to smooth day-of-week variation.
    • Any increases or decreases can be seen up to 7 days after the event in the smoothed line due to the seven day look back period (e.g 2026 Easter period was from 3-6 April, but the point for 6th April is included in the 7-day smoothing until April 12th)

Aviation departures - scheduled services (inc. freight)

  • Departures only for both international and domestic.
  • Airports: AKL (Auckland), CHC (Christchurch), WLG (Wellington), ZQN (Queenstown).
  • For future dates - number of scheduled flights including freight, schedule goes about 6 months ahead
  • This data is provided by Airport Coordination Limited (ACL).

Notable revisions

We have made revisions to the dashboard as it evolves for improved relevance to the wider transport sector and general public. Notable revisions include:

  • Removal of Light VKT due to issues with representative sample size.
  • Underlying data change which increased values of Heavy VKT due to updates from data provider.