# The Bosphorus: two water layers beneath a narrow shipping route

At the surface, water flows south; below, it returns north. Ships pass through the same strait, while maps, measurements, rules and accounts of 1453 come from different sources.

## How narrow is the Bosphorus?

### About 31 km between two seas
Between the Black Sea and the Sea of Marmara, the Bosphorus runs about 31 km by conventional geographic description. Follow its bends on the map.
This is a conventional description, not a length recalculated from the OpenStreetMap extract.

### The reported average is 1.6 km
Alpar and Yüce report an average width of 1.6 km, with descriptions ranging from 0.7 to 3.5 km. These are the study’s geographic descriptions, not a new survey of every point.
The average and range belong to one study description; they do not define every bend.

### 36 m on average; 110 m maximum
Alpar and Yüce give an average depth of 36 m and a maximum of 110 m. These are descriptions in the 1997 study, not a main-channel minimum or a recent point-by-point survey.
The average and maximum do not set the depth at every stretch.

### About 700 m
Türkiye’s Ministry of Transport characterizes the Aşiyan–Kandilli stretch as about 700 m wide. Keep the attribution: it is the ministry’s characterization, published in 2025.
The ministry characterizes this stretch; the figure is not a new measurement series.

The map shows where these reference points sit along the full strait.

Sources: [M10](https://dergipark.org.tr/en/download/article-file/103910), [M11](https://mdpi-res.com/d_attachment/jmse/jmse-10-00659/article_deploy/jmse-10-00659.pdf), [M07](https://www.uab.gov.tr/haberler/istanbul-bogazi-ndan-8-ayda-26-binin-uezerinde-gemi-gecti/)

## Water moves in two opposing layers

### Two layers, two directions
In the same channel, two layers move in opposite directions: less-salty surface water heads south toward Marmara; saltier water returns north below it. The study observes that exchange in instrumented sections.
The drawing shows the directions; it is not a measured profile to scale.

### Speed changes from south to north
In 2008–09 observations, southward surface flow often exceeded 2.0 m/s in the southern section; in the north it was generally below 0.30 m/s. These are local readings, not constant speeds across the Bosphorus.
Keep the period and place together: north and south do not give the same value.

### Below it, water returns north
The same study finds northward flow usually below 0.70 m/s in the south and often reaching 1.00 m/s in the north. Speeds vary with section and depth; these observations are from 2008–09.
A reading from one section does not become the usual speed of the whole strait.

The directions can be placed on a map; the speeds stay tied to the sections where they were measured.

Sources: [M08](https://archive.org/details/DTIC_ADA553775), [M09](https://open.metu.edu.tr/handle/11511/65560)

## The Bosphorus between the Black Sea and Marmara

### The waterway and coasts on one map
On the OpenStreetMap map dated 6 October 2026, follow the Bosphorus line between its two coasts; Aşiyan and Kandilli appear as named points. This is dated geometry, not a historical route or a track of a particular ship.
This is OSM; the 1453 boats’ route remains a separate schematic.

### The waterway opens to the Black Sea
Follow the line north to the Black Sea. The coast and waterway come from the same dated extract; their position here says nothing about depth or speed.
This map locates the shores; current speeds come from separate observations.

### Aşiyan and Kandilli
The named Aşiyan and Kandilli points anchor the ministry’s characterization of a stretch about 700 m wide. The points are from OpenStreetMap; the width figure comes from a 2025 ministry release.
Two sources, two roles: mapped locations and a width characterization.

On the European shore, Rumeli Hisarı offers another clue to control of access.

Sources: [OSM](https://www.openstreetmap.org/copyright), [M07](https://www.uab.gov.tr/haberler/istanbul-bogazi-ndan-8-ayda-26-binin-uezerinde-gemi-gecti/), [M10](https://dergipark.org.tr/en/download/article-file/103910), [M04](https://www.kiyiemniyeti.gov.tr/vessel_traffic_and_pilotage_services)

## Rumeli Hisarı and control of access

### Rumeli Hisarı: 1452
On the European shore, the 2020 photograph shows Rumeli Hisarı above the channel. Türkiye’s Ministry of Culture and Tourism dates the fortress to 1452; the image cannot establish its full role in the siege.
The date is attributed to the ministry page; the photograph is a 2020 view.

### Aid from the north
The ministry says the fortress was built to prevent aid reaching Byzantium from the north. Keep this as the institution’s explanation; the fortress alone does not explain the course of the siege.
An attributed purpose does not establish the military outcome on its own.

### Spring 1453; 29 May
A separate ministry page places the siege in spring 1453 and the city’s entry on 29 May. These are calendar reference points; they do not compress the siege’s voices and episodes into one account.
Keep the fortress date, siege period and entry date distinct.

A century later, the accounts move from the fortress to a night of overland transport.

Sources: [C03](https://commons.wikimedia.org/wiki/File:Fortress_Rumeli_hisar%C4%B1_as_seen_from_a_Bosphorus_boat.jpg), [CULT-RUMELI](https://www.ktb.gov.tr/EN-113757/rumeli-fortress.html), [CULT-ENTRY](https://ekitap.ktb.gov.tr/TR-80057/turkler-istanbula-giriyor.html)

## What do the sources say about the night of 21–22 April 1453?

### Boats move overland
On the night of 21–22 April 1453, the boats’ route leaves the water: the Turkish Historical Society recounts Ottoman boats hauled from Tophane/Boğazkesen, across Galata toward Kasımpaşa and the Golden Horn, around the harbour chain.
This is a portage toward the Golden Horn, not a Bosphorus crossing; the diagram does not use OSM.

### 67–72 boats
The Turkish Historical Society’s account passes on different totals, from 67 to 72, depending on the chronicle cited. Keep the reported range: it is not an independently verified count.
The variation between accounts is part of the number.

### Ottoman court, Venetian witness, manuscript
Tursun Beg writes from a court setting shaped by dynastic praise; Nicolò Barbaro offers a Venetian eyewitness perspective. The BnF record identifies the miniature in manuscript Français 9087, folio 207 verso; it is a manuscript object, not a neutral image.
Keep the narrative, eyewitness account and manuscript record distinct.

The Convention and modern navigation rules ask for the same care: keep documents with different roles apart.

Sources: [M12](https://ttk.gov.tr/istanbulun-fethi/), [M13](https://ottomanhistorians.uchicago.edu/sites/default/files/tursunbeg_en.pdf), [M13-BNF](https://mandragore.bnf.fr/ark:/12148/cgfbt45374m), [M13-BAR](https://www.deremilitari.org/RESOURCES/SOURCES/constantinople3.htm), [C04](https://commons.wikimedia.org/wiki/File%3ASiege_of_Constantinople_BnF_MS_Fr_9087.jpg)

## The Convention and navigation safety are distinct layers

### One system, three waters
Open the Convention’s timeline at its three dates: signed 20 July 1936, provisionally in force from 15 August, definitive from 9 November. Its scope is the Bosphorus, Sea of Marmara and Dardanelles as one system.
Here we keep to the dates and Articles 2, 13–14 and 18–21; we do not paraphrase their text.

### The traffic scheme has a different timeline
Türkiye’s traffic separation scheme began in 1994 and received IMO approval in 1995. This is a navigation-safety measure, kept separate from the Convention text.
A separate regime does not become a Convention article because it uses the same water.

### Operational from 30 December 2003
The vessel traffic service became operational on 30 December 2003. Its operational page lists a 17-nautical-mile VTS area for the Istanbul Strait; this is not the measured length of the Bosphorus.
The operating date and the service area are different measures.

Now read the traffic figures as reported passages, each with its year and definition.

Sources: [M01](https://www.mfa.gov.tr/site_media/html/BOGAZLAR-RAPORU-2024-EN.pdf), [M02](https://www.mfa.gov.tr/implementation-of-the-montreux-convention.en.mfa), [M03](https://www.mfa.gov.tr/the-turkish-straits.en.mfa), [M04](https://www.kiyiemniyeti.gov.tr/vessel_traffic_and_pilotage_services)

## What do the 2024 and 2025 figures count?

### 40,172 passages in 2025
Open the ministry’s 2025 tally: 40,172 reported passages through the Istanbul Strait in the calendar year. Each mark is a movement, not a unique ship.
Keep the year and count passages, not distinct vessels.

### Tankers carrying hazardous cargo
A separate Türkiye Ministry of Foreign Affairs table reports 9,554 tankers and 169,782,040 tonnes in the hazardous-cargo category in 2025. The table does not cover every vessel or all cargo.
Do not compare these tonnes with totals defined more broadly.

### 41,363 passages reported for 2024
The ministry release dated 7 September 2025 reports 41,363 passages through the Istanbul Strait in 2024.
“2024 value and 2025 partial-year figures are dated records, not a current forecast.”

The grain corridor uses different ports, a different period and a different denominator.

Sources: [M06](https://www.uab.gov.tr/haberler/tuerk-bogazlari-2025-te-84-bin-640-gemilik-trafigi-yoenetti/), [M07](https://www.uab.gov.tr/haberler/istanbul-bogazi-ndan-8-ayda-26-binin-uezerinde-gemi-gecti/), [M03](https://www.mfa.gov.tr/the-turkish-straits.en.mfa), [C02](https://commons.wikimedia.org/wiki/File%3AWHITE_TRADER_Crude_Oil_Tanker%2C_Bosphorus_Strait%2C_Istanbul%2C_Turkiye_%28Ank_Kumar%2C_Infosys_Limited%29_01.jpg)

## The grain initiative is not current Bosphorus traffic

### 32.8 million tonnes; 1,003 voyages
Set the Bosphorus tally aside. Through 15 July 2023, the UN reports 32.8 million metric tonnes of food over 1,003 voyages under the Black Sea Grain Initiative, for three Ukrainian ports.
These are initiative exports and voyages, not a Bosphorus passage total.

### The initiative ended on 17 July 2023
The UN dates the end of the Black Sea Grain Initiative to 17 July 2023. The 2025 traffic figures belong to a different period and geography.
This is historical corridor context, not a current Bosphorus transit route.

### Keep each count in its own frame
Water exchange, navigation and history meet in one strait. Each figure holds its meaning only beside the place, year and category it describes.
Currents are measured by section; passages and voyages count different things.

Leave every number beside the period, route and institution that produced it.

Sources: [UN-JCC](https://www.un.org/en/node/207702), [UN-FAQ](https://www.un.org/en/black-sea-grain-initiative/faq), [M08](https://archive.org/details/DTIC_ADA553775), [M06](https://www.uab.gov.tr/haberler/tuerk-bogazlari-2025-te-84-bin-640-gemilik-trafigi-yoenetti/)