Strait of Hormuz: Tracking the Ships

The daily backbone

PortWatch is the only public source with an unbroken daily count for the whole period, published to 30 August. It recorded 1,237 qualifying transits in 170 days, both directions combined. The second half of March averaged three a day. April rose to 7.1, May fell back to 3.9, and June brought the one real recovery at 12.9 a day, peaking at 51 on 24 June. July slipped to 10.2 and August to 4.8.

The gold bars are PortWatch's tanker class, 520 of the 1,237 transits. It is an aggregate class with no crude, product or LNG split, and the series does not publish direction.

More ships left than arrived

PortWatch does not publish direction. The open strait-of-strait project does, from 14 March to 23 July, counting every AIS-visible vessel type. It logged 1,063 outbound and 820 inbound transits, an outbound share of 56.5%: a strait emptying faster than it refilled. The raw counts are shown here. The project's own illustrative uplift of five vessels a day per direction for dark traffic is left out.

On the 132 days both series cover, this one averages 14.3 transits a day against PortWatch's 8.0, a ratio of 1.78. That gap comes from different geofences and vessel definitions, not from more ships, which is why the two are not combined. The series ends on 23 July and nothing is spliced in after it.

Twenty-one published days

From mid-July, Reuters began quoting Kpler's daily count of visible commodity vessels crossing the strait. Only 21 such days were published between 14 July and 3 September, and they are shown as isolated points rather than a line: the gaps are real, not missing data. The count excludes vessels with transponders off, so it sits below what AXSMarine records for tankers alone in the same weeks.

On the 17 dates that overlap with PortWatch, Kpler averages 7.2 vessels a day against PortWatch's 4.4. Because Reuters reports on newsworthy days, that sample leans towards busy ones. Several early-September figures were revised between reports; the latest value is used. Select a point or a card for the underlying report.

Tankers only, dark ships included

The only tanker-specific public series comes from AXSMarine crossing records published by Signal Group. Unlike the Kpler counts, these include tanker passages identified despite switched-off transponders. Across three non-overlapping phases from 27 February to 15 July they logged 759 tanker crossings: 329 in the war phase at 3.0 a day, 326 in the 19-day MoU reopening at 17.2 a day, and 104 in the ten days of renewed escalation at 10.4 a day. Crude's share of the traffic rose from 25% to 49% across the three.

A fourth window, 14 July to 3 August, overlaps the escalation phase and is published with more detail: 102 tanker crossings, 55 inbound and 47 outbound, 42 of them crude. Around 62% could not be traced on AIS.

Different levels, same story

Matched to AXSMarine's phase windows, PortWatch's all-vessel daily average moves in step with the tanker-only series: low in the war phase, a sharp jump in the reopening, a fall in the escalation, and a further drop in the blockade window, where the two land on the same 4.9 a day by coincidence of definition. The disagreement between providers is in level, and it is structural. Averaging them would hide it.

We have not been able to find a source that gives a complete six-month, day-by-day count of energy tankers by direction, cargo and load state. PortWatch has the daily backbone but no direction or cargo detail. The independent AIS series has direction but covers all vessels and stops on 23 July. Kpler's visible counts are public on 21 days only. AXSMarine publishes tanker direction and cargo mix as period aggregates. A full vessel-level record requires commercial data. Every daily series here is an AIS-visible lower bound, and no gap in one provider has been filled from another.