FreighterLog The record of Great Lakes shipping

Archive: January 2024 to December 2025, 28,207 passages. Seaway transits checked hourly. Last checked 2026-09-23 18:00 UTC — 14:00 EDT. NOAA publishes in quarterly batches, and the archive currently runs nine months behind.

Archive to December 2025 · Seaway checked 2026-09-23 18:00 UTC — 14:00 EDT

Methodology

How the passage record is made

Everything on this site is derived from public records. This page says which ones, what we do to them, and where the numbers stop being trustworthy.

Where it comes from

Every passage here is computed from the Nationwide Automatic Identification System, the Coast Guard's AIS record, published by NOAA as CC0 public-domain monthly track files. We publish derived products only: passage events, counts, season aggregates. Raw AIS is not mirrored here and not offered for download. NOAA's copy is free. The near-current layer is the Seaway's public vessel-transit and order-of-turn tables, read once an hour.

Vessel positions

Passage events are computed from the Nationwide Automatic Identification System (NAIS), published by the NOAA Office for Coastal Management at marinecadastre.gov under CC0. The archive uses the monthly vessel-track files, which sample each vessel's position at roughly one-minute intervals.

That archive arrives in quarterly batches rather than month by month, so the gap between today and the newest AIS month is not a fixed figure. As this page was built the archive was running nine months behind: FreighterLog's passage record runs January 2024 to December 2025 and stops there. It is not delayed by choice and it is not going to be fresher than the source.

Seaway transits

The transit board on the home page is an hourly snapshot of the St. Lawrence Seaway's own public vessel-transit table at seaway-greatlakes.com. It is reproduced with its fetch timestamp attached and is not merged into the AIS-derived passage record. Most recent snapshot: 2026-09-23 18:00 UTC.

How a passage is computed

A chokepoint on this site is a line drawn across a channel, not a lock. A passage is one vessel track crossing that line, timed at the sample where it crosses. Direction comes from which way the boat went over it, checked against the same boats arriving at the next chokepoint downstream rather than read off a map: the Detroit River labels were inverted at first, because Windsor sits south of Detroit. Two crossings by the same vessel inside three hours count once: a boat waiting her turn drifts back and forth across the line.

That cross-check is the main way this record catches its own mistakes, and it is the reason the direction column can be wrong in a way the counts are not. Over a full archive a chokepoint has to run close to even, because a boat that goes one way has to come back; a lopsided split means the gate line or its labels need work, not that the traffic is one-way. It has caught the inverted Detroit River labels and, later, a gate line at the head of the lakes that was cutting the harbor rather than the channel. Where a chokepoint still reads lopsided, its own page says so at the top of the direction split rather than reporting the skew as a finding.

Step by step

  1. A gate line is drawn across each chokepoint: a single line, positioned on the observed traffic corridor rather than on a chart label.
  2. Each month of AIS tracks is narrowed to roster vessels whose track touches a gate's envelope.
  3. The crossing point of the track against the gate line is found geometrically, and the nearest track vertex supplies the timestamp. Vertices are roughly one-minute samples, so the timing survives lock queues, where assuming constant speed would not.
  4. Direction is inferred from the heading of the track across the line, and is expressed in the vocabulary that chokepoint actually uses: upbound and downbound at the Soo, eastbound and westbound at Mackinac, inbound and outbound at the Duluth and Superior entries.
  5. Re-crossings within three hours are collapsed into one event, so a vessel maneuvering in a lock approach is counted once.

How precise the times are

AIS tracks sample about once a minute, so the sampling is fine and the meaning is coarse. Crossing our line at the Soo is not the same event as entering the Poe, and a boat can sit in the approach for hours. Times are stored in UTC, shown in the chokepoint's local time, and quoted as good to about fifteen minutes. A passage here is not an official lock record.

Passage times are interpolated from roughly one-minute AIS track samples and are good to about ±15 minutes. They are not official lock records. Times are held in UTC and shown in the local time of the chokepoint the vessel crossed.

  • ±15 minutes. Treat every displayed time as a window, not a stamp. These are not lock records and no official lock schedule was consulted.
  • Time zones. Timestamps are stored in UTC and displayed in the local time of the chokepoint where the passage happened. A single vessel's log can therefore move between Central and Eastern time as she moves west and east.
  • Bucketing. Daily counts use the local calendar date at the chokepoint. Monthly and season totals use UTC, matching the monthly files the archive is built from.
  • AIS is self-reported. Vessel names, dimensions and identifiers come from the transponder, entered by people aboard. Spellings vary, dimensions are sometimes the tug-and-barge unit rather than the hull, and the odd figure is simply wrong. Those values are reproduced as broadcast rather than quietly corrected.
  • A passage is a crossing, not a voyage. One round trip generates several passages. These counts describe line crossings, not cargoes or tonnage.

How current it is

NOAA publishes the AIS months in quarterly batches; the archive currently runs nine months behind, and ends at December 2025. The batches are irregular, so the gap is not a fixed number of months and this page recomputes it every time the site is built. The record runs from January 2024 through December 2025. Between the end of the archive and now, the Seaway table is the only live thing we hold, and it covers the Seaway alone: the Welland Canal and the St. Lawrence locks, not the Soo or Duluth.

What “not observed” means

Not observed is not the same as did not happen. A crossing can be missing because its month is not in the archive yet, because the boat's AIS was off or her track had a gap, or because she is not a hull we carry. The lay-up list works the same way: it holds fleet boats with no AIS match in the one month used to build the roster, a fact about that month, not about today. Names and dimensions come from AIS messages typed in by crews, so spellings wander and a few hulls report impossible dimensions; we keep the names as broadcast.

Why our count can differ from a community log

BoatNerd's passage lists are compiled by people watching boats and reading lock information. Ours is a line on a chart and a computer. On what happened at one lockage, believe the humans. On how many times that boat came past Port Huron in 2025, we can count it.

Season aggregates, as a CSV

season-aggregates.csv — every season on this site in one tidy table: the year's totals, then its passages and distinct vessels broken out by chokepoint, by fleet, and by the ten busiest hulls. 2 seasons (2025, 2024), generated from the same export that builds these pages. Free, no sign-up, served from this domain.

If you publish anything from it, carry the citation. These are derived aggregates; the records underneath them are NOAA's and come with terms of their own. Vessel data derived from the Nationwide Automatic Identification System (NOAA Office for Coastal Management, marinecadastre.gov), CC0.

Following the record by RSS

Three kinds of feed, all static files on this domain, all free of accounts, addresses and trackers. Items appear when the next batch of the public AIS archive is processed, not when a boat passes — NOAA publishes in quarterly batches, and the archive currently runs nine months behind.

  • One vessel. /vessels/<slug>/feed.xml — her last 40 recorded crossings. Linked from every vessel page, and the reason most people subscribe: nobody else publishes a per-hull passage feed for the lakes.
  • One chokepoint. /locks/<id>/feed.xml — the last 50 crossings of that line, newest first.
  • Everything. /feed.xml — the newest 60 crossings anywhere in the archive.

Each item is one passage. Its identifier is the vessel's MMSI and the UTC instant of the crossing, so a passage shows up once in your reader however many feeds carry it, and stays put when the window slides. Where a chokepoint's direction labels are currently known-bad, the item title says the boat crossed and does not say which way — that decision is computed from the archive on every build, not maintained by hand. The NOAA citation travels in each feed's copyright field.

What is deliberately not here

  • No raw AIS. FreighterLog publishes derived products only: passage events and statistics computed from public-domain AIS. Raw AIS records are never republished here and are not offered for bulk download. What is published is passage events, aggregates and histories, and nothing that would let the underlying position records be reassembled from them.
  • No live map, and no one else's embedded. Third-party live maps are linked to, never framed into these pages.
  • No charge for the data. The underlying records are public domain and stay that way.
  • No tracking. This site makes no external requests: no fonts, no analytics, no third-party scripts of any kind. Nothing loads except pages from this domain.