Civilian harm
Civilian deaths, and the counts of them
Strike casualties and protest deaths, and why this module publishes four irreconcilable counts rather than one figure.
Coverage · Iran–Gulf
AssessmentModerate confidence
Broad regional level only. This surface never maps aid routes, facility locations or operational detail. Casualty figures from every party to this war are treated as claims and carried with their attribution. The module publishes no total for either side, and attacker battle-damage assessment is never carried as established damage.
Current state
What civilian pressure looks like now
Moderate confidence · Data as of 2026-09
Three distinct civilian effects run through this system, and they are measured in three incompatible ways. A strike with an acknowledged cause and a known toll; a maritime workforce stranded by a commercial decision rather than by any attack; and a domestic casualty count on which no two sources agree about what is being counted. The module publishes each with its own method attached, and does not add them together.
- Killed in the Minab school strike168As of 2026-02-28High confidenceA United States strike, on the opening day of the war. A preliminary investigation found the United States responsible and attributed it to outdated targeting information — an acknowledged responsibility rather than an allegation, which is why the figure carries a high grade.
- Seafarers stranded aboard vessels unable to move~20,000 at peakAs of 2026Moderate confidenceThe International Maritime Organization's figure, which has fallen as evacuation proceeded. These crews were immobilised by the withdrawal of commercial cover rather than by any attack on their ships.
- Seafarer deaths confirmed by the maritime regulator17As of 2026-08High confidence
- Iranian protest deaths — government figure3,117As of 2026Low confidenceThe Iranian government's own total, of which it describes 2,447 as civilians and security forces and the remainder as "terrorists". A party's count of deaths it is accused of causing, carried as its stated position.
- Iranian protest deaths — individually documented7,007As of 2026-02-23Moderate confidenceHRANA's count of deaths documented case by case: 6,488 protesters, 236 children, 76 non-protester civilians and 207 military or government personnel. DOCUMENTED means each death has a record behind it.
- Iranian deaths listed as still under investigation17,091As of 2026-02-23Low confidenceA separate HRANA category and NOT a death toll. These are cases the organisation has not established, and adding them to the documented figure would present unverified reports as verified ones.
- Iranian protest deaths — higher media estimates16,500 – 20,000As of 2026Low confidenceCirculating in media reporting citing activist groups. The methodology behind them is not published, which is why they are carried as a separately labelled class rather than as the upper end of a range.
- Wounded in the 8 September attack wave on southern Saudi Arabia — Saudi figure73As of 2026-09-08Moderate confidenceThe Saudi-led coalition spokesman's count for Abha, Khamis Mushait, Jazan and Najran, including women and children. A party's figure.
- Killed in the attacks on Riyadh's airport of 10–11 October — Security Council figure13As of 2026-10-15High confidenceWith more than 300 wounded, in presidential statement S/PRST/2026/4. Saudi officials put the 10 October toll at at least twelve killed and 309 wounded the same evening; the two counts are not reconciled, and the three killed in the 8 October attacks are a separate count.
- Killed by strikes in the Eastern Province on 21 October4As of 2026-10-21High confidenceTwo at Dammam's airport on the civil aviation authority's count, which the Houthi movement said it targeted, and two foreign workers killed by debris from an intercepted drone at Shaybah on the energy ministry's account, which did not say who launched it.
Civilian pressure, as analytical objects The 4 pressures this module publishes on Iran–Gulf, as selectable objects. Each carries its present condition where the module states one, its own figures, the module's analysis of it and — where the corpus establishes one — the mechanism by which security conditions reach it.Interrogate Iran–Gulf in Explore
Civilian harm
Civilian harm in Saudi Arabia from the Red Sea front
Deaths and injuries from missile and drone attacks on Saudi cities and airports since September 2026, each count with its own source and none summed.
Protection
Seafarers held by the situation
Crews aboard vessels that cannot move, and deaths with no attack behind them — a civilian effect produced by risk pricing.
Civilian infrastructure
Power and water as one system
Electricity and water supply that fail together, because in much of the region they are the same system.
The full picture5 blocks of analysis and 10 dated figures, filed under the pressures above — the module’s own words, with their own vintages
Everything below is authored on this module’s humanitarian record. Figures carry the vintage of the release that produced them, not the date of this page: the section is graded Moderate confidence and its data is stated as of 2026-09.
Civilian deaths, and the counts of themCivilian harm
Strike casualties and protest deaths, and why this module publishes four irreconcilable counts rather than one figure.
No assessed-as-of date on the objectnever reviewed as an object
| Indicator | Value | As of | Confidence |
|---|---|---|---|
| Killed in the Minab school strikeA United States strike, on the opening day of the war. A preliminary investigation found the United States responsible and attributed it to outdated targeting information — an acknowledged responsibility rather than an allegation, which is why the figure carries a high grade. Sources: Congressional Research Service — Iran, the Gulf and the Strait of Hormuz | 168 | 2026-02-28 | High confidence |
| Iranian protest deaths — government figureThe Iranian government's own total, of which it describes 2,447 as civilians and security forces and the remainder as "terrorists". A party's count of deaths it is accused of causing, carried as its stated position. Sources: Official government statements | 3,117 | 2026 | Low confidence |
| Iranian protest deaths — individually documentedHRANA's count of deaths documented case by case: 6,488 protesters, 236 children, 76 non-protester civilians and 207 military or government personnel. DOCUMENTED means each death has a record behind it. | 7,007 | 2026-02-23 | Moderate confidence |
| Iranian deaths listed as still under investigationA separate HRANA category and NOT a death toll. These are cases the organisation has not established, and adding them to the documented figure would present unverified reports as verified ones. | 17,091 | 2026-02-23 | Low confidence |
| Iranian protest deaths — higher media estimatesCirculating in media reporting citing activist groups. The methodology behind them is not published, which is why they are carried as a separately labelled class rather than as the upper end of a range. | 16,500 – 20,000 | 2026 | Low confidence |
Why there is no single protest casualty figure here
The four figures above measure four different things, and the range a reader would naturally form from them — roughly three thousand to twenty thousand — would be a statement about nothing. A government total, a set of individually documented deaths, a list of unestablished cases and an unpublished estimate are not four points on one distribution. Vigil publishes each with its method attached and adds no midpoint, no range and no total. An earlier draft of this module did carry such a range and it was withdrawn during verification for exactly this reason. There is also no independent verification available, and the sources do not share a definition of a protest death.
Sources: UN Security Council and Panel of Experts reporting↗ · Established international media reporting
Recorded events:A United States strike on a school in Minab kills 168 people28 Feb 2026United States and Israeli strikes on Iran open the war28 Feb 2026
Parties:IranIsraelUnited States
Civilian harm in Saudi Arabia from the Red Sea frontCivilian harm
Deaths and injuries from missile and drone attacks on Saudi cities and airports since September 2026, each count with its own source and none summed.
Assessed as of 2026-10-21never reviewed as an object
| Indicator | Value | As of | Confidence |
|---|---|---|---|
| Wounded in the 8 September attack wave on southern Saudi Arabia — Saudi figureThe Saudi-led coalition spokesman's count for Abha, Khamis Mushait, Jazan and Najran, including women and children. A party's figure. Sources: Gulf state official government sources | 73 | 2026-09-08 | Moderate confidence |
| Killed in the attacks on Riyadh's airport of 10–11 October — Security Council figureWith more than 300 wounded, in presidential statement S/PRST/2026/4. Saudi officials put the 10 October toll at at least twelve killed and 309 wounded the same evening; the two counts are not reconciled, and the three killed in the 8 October attacks are a separate count. Sources: UN Security Council and Panel of Experts reporting↗ · Gulf state official government sources | 13 | 2026-10-15 | High confidence |
| Killed by strikes in the Eastern Province on 21 OctoberTwo at Dammam's airport on the civil aviation authority's count, which the Houthi movement said it targeted, and two foreign workers killed by debris from an intercepted drone at Shaybah on the energy ministry's account, which did not say who launched it. Sources: Gulf state official government sources · Established international media reporting | 4 | 2026-10-21 | High confidence |
Civilian harm in Saudi Arabia, counted occasion by occasion
As of 2026-10-21
The Red Sea front has killed and wounded civilians in Saudi Arabia since September 2026, and the counts come from different bodies on different occasions: the coalition spokesman for the 8 September wave, the civil aviation authority and Saudi officials for the attacks on Riyadh's airport, the Security Council for the attacks of 10 and 11 October, and the energy ministry for Shaybah. The first civilian death of the round was reported by Saudi state television on 17 September, from drone debris over Taif. The module publishes each figure with its source and adds no total, because the bodies count different occasions and, for Riyadh, give different figures for the same one. Attribution stays with its author: the movement claimed some attacks, Washington attributed one it did not claim, and the Shaybah drone was not attributed by the ministry. Civilian harm in Yemen from the coalition's campaign is outside this module's scope and is not authored here.
Sources: Gulf state official government sources · UN Security Council and Panel of Experts reporting↗ · US Department of State bureau fact sheets↗ · Established international media reporting
Recorded events:The heaviest Houthi attack wave against Saudi territory of this round8 Sep 2026, continuing to 3 OctAttacks on Riyadh's airport kill civilians in the Saudi capital8 – 10 Oct 2026A Security Council presidential statement condemns the Houthi strikes on Saudi Arabia and affirms its right of self-defence15 Oct 2026
Parties:Saudi ArabiaHouthi movement
Seafarers held by the situationProtection
Current conditionStranded numbers falling as evacuation proceedsAs of 2026-08
About 20,000 seafarers were stranded at the peak; 17 seafarer deaths were confirmed by the maritime regulator by August 2026.
Crews aboard vessels that cannot move, and deaths with no attack behind them — a civilian effect produced by risk pricing.
Assessed as of 2026-08never reviewed as an object
How conditions reach this
Crews aboard vessels that could not move
The human cost of an access constraint, counted by the regulator.
- Affected area or system Confirmed
- Access or service disruption Confirmed
- Civilian condition Confirmed
- Broader humanitarian significance Confirmed
Dependencies
Physical systems:Hormuz-dependent crude and products export system
| Indicator | Value | As of | Confidence |
|---|---|---|---|
| Seafarers stranded aboard vessels unable to moveThe International Maritime Organization's figure, which has fallen as evacuation proceeded. These crews were immobilised by the withdrawal of commercial cover rather than by any attack on their ships. | ~20,000 at peak | 2026 | Moderate confidence |
| Seafarer deaths confirmed by the maritime regulator | 17 | 2026-08 | High confidence |
A civilian effect with no attack behind it
The seafarers stranded in the Gulf were not stopped by anyone's weapons. They were stopped because the cover their voyages depend on stopped being written, and a vessel without war-risk cover does not sail. It is the clearest illustration in the module of how a commercial mechanism produces a physical humanitarian outcome, and it is why the module treats insurance as a system rather than as a market detail.
Sources: UN Security Council and Panel of Experts reporting↗ · Established international media reporting
Recorded events:Marine war-risk cover for the Gulf is terminated and repriced5 Mar 2026The IMO states there is no safe transit anywhere in the Strait of Hormuz24 Apr 2026The IMO records 65 incidents and 17 seafarer fatalities11 Aug 2026
Analytical Notes:The strait that closed without being blocked
Parties:IranUnited States
Power and water as one systemCivilian infrastructure
Current conditionGulf drinking water made by power-linked desalinationAs of 2026-03
A structural, national-level dependency; no plant-level dependency is published here, by rule.
Electricity and water supply that fail together, because in much of the region they are the same system.
Assessed as of 2026-03never reviewed as an object
Dated changes
- 2026-03-02 · DamageA strike damaged a water tank at Qatar's Mesaieed power complex — a water event and an electricity event at once.Moderate confidence · Sources: Gulf state official government sources · Official government statements
How conditions reach this
Where power and water are the same system
Why a strike on generation is also a strike on drinking water.
- Affected area or system Confirmed
- Access or service disruption Confirmed
- Civilian condition Plausible exposure
- Broader humanitarian significance Plausible exposure
Dependencies
Physical systems:Qatari LNG export systemSaudi east–west crude transit
Runs on:energy system
Desalinated water and the fuel it runs on — a national-level dependency
As of 2026-03
In most of the Gulf, drinking water is manufactured. Desalination supplies on the order of 99 per cent of Qatar's potable water, more than 90 per cent of Bahrain's and Kuwait's, about 86 per cent of Oman's, roughly 70 per cent of Saudi Arabia's and about 42 per cent of the UAE's, on compilations published during the 2026 war; the six states together operate roughly a third of the world's desalination capacity. Desalination and power generation are in large part the same installations, fed by the same gas and fuel, which is why the module files them as one system: the strike that damaged a water tank at Qatar's Mesaieed power complex on 2 March 2026 was a water event and an electricity event at once. Strategic storage buys days rather than weeks, and that is the whole of what this module states about it. Iran's exposure is the inverse. It draws most of its water from rivers, reservoirs and depleted aquifers rather than from the sea, and after a fifth consecutive drought year Tehran's reservoirs stood near a tenth of capacity in late 2025, with the president warning publicly of the capital's water supply; Iran's foreign ministry separately accused the United States of striking a desalination plant on Qeshm island during the war, a claim carried here as the party's position. The two dependencies do not meet: one is a system that can be struck, the other a climate that cannot be negotiated with, and neither is described here below the national level.
Sources: Gulf state official government sources · Congressional Research Service — Iran, the Gulf and the Strait of Hormuz · Official government statements · Established international media reporting
What this does not establish
- National desalination shares are compiled from different base years and definitions and are carried as orders of magnitude, not as a comparable series.
- Strategic water storage is stated by governments in days and is not independently verifiable; this module publishes no figure for any state.
Recorded events:Energy and shipping targets struck across four Gulf states2 – 3 Mar 2026
About this sectionScope and standing observables
Analysis and observables the module authors about the civilian picture as a whole rather than about any one pressure. They are filed under no object because they belong to none.
What this section deliberately does not contain
Gulf electricity and drinking water depend on a small number of fixed installations, and that dependence is real, structural and documented at national level. This module states it at that level and no further. It publishes no plant-level dependencies, no infrastructure vulnerability assessment and no single-point-of-failure analysis, because a public record of which installation a population cannot survive the loss of is a target list whatever it is called.
What to watch
- Does any body Vigil can cite publish a consolidated civilian casualty count for the 2026 war?The section carries four irreconcilable protest-death figures and one strike count because no consolidated count exists. A UN, ICRC or national-regulator series that covered strikes and protests on one method would replace them; a fifth party figure would not.
- Are the stranded crews repatriated, and does the regulator's seafarer death count stop rising?The peak of roughly 20,000 seafarers stranded and 17 confirmed deaths are the section's two maritime figures. The observable is a published repatriation count from a flag state, the IMO or a P&I club, and a regulator series that reaches a final figure rather than a running one.
Evidence and grading
- Figures
- Agency estimates on their own revision cycles. Each figure carries the date it is for; nothing on this surface averages, totals or compares across two vintages.
- Confidence
- How well established Vigil considers this assessment to be. It is not a measure of how severe the situation is.