All investigationsless ~/investigations/four-days-in-may-operation-sindoor.md 14 MAY 2025

Operation Sindoor, 7-10 May 2025

Four Days in May

What Operation Sindoor actually destroyed

A message to Pakistan, built from Pakistan’s own evidence: the air bases India cratered, the radars it blinded, the militant headquarters it flattened, and the gap in strength that Islamabad won’t show you. It’s also about why the “we won” story is the most expensive thing you’ll ever be told.

Theatre
Pakistan, PoJK and the northern Arabian Sea
Mostly on
Rawalpindi, Bahawalpur and Muridke
Reading time
34 minutes

Congratulations, Pakistan, on winning the four-day war of May 2025. It was such a clean win that your air force has spent every month since putting out contracts to rebuild the air bases that were never hit, more than thirty of them so far. This piece celebrates that victory using your own paperwork, because the best witness against the Pakistani military turns out to be its own filing cabinet.

It started at 1:44 a.m. on 7 May 2025. Over the next four days India struck deeper into Pakistan than at any time since 1971: past the border districts, past Lahore, down to the approaches to Karachi. Officially, none of it touched the Pakistan Air Force. Its air bases, radars and command posts are all fine. That’s the official position, so hold on to it while you read. There’s a short quiz at the end, and a satellite marks it.

The evidence here comes from Pakistan, not New Delhi: a list of repair tenders from your own air force, a cabinet minister who went off script, funerals your own officers attended, and satellite frames your own channels put on air. You don’t have to trust India for any of it. You only have to trust the Pakistan Air Force’s procurement office, which for some reason has just had its busiest quarter in living memory.

Act I Nine buildings that fell down by themselves

The first night went after nine sites that India called “terrorist infrastructure”. Pakistan calls that a slander against nine peaceful religious retreats, which happened to have a training schedule, a resident commander and, at Muridke, a big front gate that reads MARKAZ TAIBA. Two of them mattered most: Markaz Taiba, the Lashkar-e-Taiba campus near Lahore, and Markaz Subhan Allah, the Jaish-e-Mohammed headquarters at Bahawalpur. Both were mosques, we’re told. Neither is standing.

Maxar’s satellites photographed the result within hours, and Reuters showed it to the world. At Bahawalpur three domes were punched through and 2,100 square metres flattened. At Muridke the main block was simply gone, as if it had popped out for cigarettes and decided not to come back.

Ground-level photo of the destroyed Markaz Taiba building at Muridke, roof collapsed, rubble across the courtyard
Muridke, Punjab. The 200-acre Markaz Taiba complex at ground level, shortly after its roof came down on its own. Lashkar founder Hafiz Saeed built it; officially it’s a mosque. The men in the courtyard are standing where the inside used to be. Photo circulated in Pakistani and Indian press coverage

Going by the official line, nothing hit Muridke, so the four aim points marked inside it and checked against post-strike imagery must have been weather. Strange weather, though. It arrived at 600 knots, carried a guidance package and only hit buildings owned by Lashkar-e-Taiba.

Muridke targets, points 1 and 2, before and after satellite imagery
Muridke, points 1 and 2, before and after. The point 1 building is gone. Point 2 is now a debris field, circled in yellow.
Muridke targets, points 3 and 4, before and after satellite imagery
Muridke, points 3 and 4: the same compound, and the same clean-hit signature at both.

Pakistan called them mosques. Pakistanis with phones disagreed. Sky News in London geolocated videos filmed inside the Muridke compound and posted to TikTok, YouTube and Google. They show Lashkar-e-Taiba and “Brigade 313” flags, and in one clip children do martial-arts drills under a jihad hashtag. The congregation also identified itself. Jaish chief Masood Azhar publicly mourned ten relatives and four associates killed at Bahawalpur, which is a lot of deaths for a building that wasn’t hit. India’s government says more than a hundred militants died across the nine sites. One of them was Abdul Rauf Azhar, the man behind the IC-814 hijacking and the murder of Daniel Pearl.

A compound that trains armed men and flies a militant flag doesn’t become a house of worship because someone stands in the rubble afterwards and calls it one. India hit the headquarters of the groups behind Mumbai and Pulwama, and the “mosque” kept holding funerals for fighters.

Bahawalpur, building by building

Markaz Subhan Allah got less international airtime than Muridke, but its record is more detailed: four separate aim points, each confirmed on its own.

Bahawalpur target, point 1, before and after satellite imagery
Bahawalpur, point 1. An intact building complex before; a bare, cleared footprint after. No forwarding address.
Bahawalpur targets, points 2 and 3, before and after satellite imagery: an intact JeM training compound before, damaged after
Bahawalpur, points 2 and 3. A multi-building compound before, two tidy damage circles after.
Bahawalpur target, point 4, before and after satellite imagery
Bahawalpur, point 4. The same row of buildings and the same result: intact, then gone. Four out of four, at a place nothing hit.
NDTV broadcast graphic showing Maxar satellite images of Bahawalpur dated 2 May and 7 May 2025, before and after the strike, two separate buildings
The televised version, dated 2 May and 7 May, for anyone who wants the dates on screen. Satellite image © 2025 Maxar Technologies, via NDTV

The other seven

Muridke and Bahawalpur got the coverage. The other seven rarely do, so here is the full first-wave list. Pakistan disputes two of them, and they’re marked.

#FacilityLocationCoordinatesStatus
1Markaz Taiba (LeT HQ)Nangal Sahdan, Muridke31°50′10″N 74°15′47″EStruck; imagery above
2Markaz Subhan Allah (JeM HQ)Bahawalpur, Punjab29°23′44″N 71°41′01″EStruck; imagery above
3Tehra Kalan training campSialkot, Punjab32°29′24″N 74°25′12″EStruck
4Gulpur camp (Maskar Raheel Shahid)Kotli district, PoJK33°26′05″N 73°51′40″EStruck. Pakistan disputes it; dated Maxar imagery below says otherwise
5Masjid-e-Abbas (JeM)Kotli, PoJK33°31′06″N 73°54′08″EStruck
6Mehmoona Joya facility (Hizbul Mujahideen)Sialkot, Punjab32°30′00″N 74°31′48″EStruck
7Markaz Usman-o-Ali (Barnala/Ahle Hadith)Bhimber, PoJK32°58′29″N 74°04′23″EStruck. Pakistan disputes this one
8Shawai Nala camp (LeT) and Masjid Syedna Bilal (JeM)Muzaffarabad, PoJK34°21′48″N 73°28′05″EStruck: two facilities, one strike
9Training campNeelum Valley, PoJK34°35′00″N 73°55′00″EStruck

Rows 4 and 7, Gulpur and Barnala, are the two sites the government has repeatedly denied were struck. The satellite record on Gulpur, below, disagrees.

Maxar satellite imagery of the Kotli-Gulpur camp, dated 3 October 2024 and 14 May 2025, showing the camp damaged
Kotli-Gulpur in dated Maxar imagery, 3 October 2024 and 14 May 2025. The main building’s roof has caved in, a second structure is damaged, and the frame is captioned “Terror Camp Damaged”. This is a site the government swears was never touched. Someone should tell the roof.
Maxar satellite imagery of Syedna Bilal Camp, Muzaffarabad, dated 29 May 2025, showing structural damage with measurements
Syedna Bilal Camp, Muzaffarabad, with the damage measured on the frame: an 80-by-92-foot section marked “structural damage”. Precise numbers for something that didn’t happen.

Act II The air-defence network, now available as gravel

Pakistan answered with numbers. On the night of 7-8 May it sent drones and missiles at fifteen Indian locations: Awantipura, Srinagar, Jammu, Pathankot, Amritsar, Kapurthala, Jalandhar, Ludhiana, Adampur, Bathinda, Chandigarh, Nal, Phalodi, Uttarlai and Bhuj. Over the following nights that came to something like 600 drones and 800 munitions. They flew into an Indian air-defence net made up of the Akashteer system feeding the IACCS, the indigenous Akash, the Israeli SPYDER, and the S-400 held in reserve. According to the military historian Tom Cooper, in his study 88-Hours War, it shot them down at a rate close to 100 percent.

Pakistan also reached for its best weapons. It fired the CM-400AKG, a “carrier-killer” it had bought to sink Indian aircraft carriers, at parked radar dishes. It fired the nuclear-capable Ra’ad cruise missile too, and India reports catching several of those in the air. None of it changed the arithmetic. The Chinese-built shield over Pakistan worked well right up until it was asked to defend something.

Then it really did turn into gravel. Across 7 and 8 May, Indian Harop and Harpy loitering munitions, backed by jamming that blinded the radars, went after the HQ-9 batteries around Lahore and Sialkot. Before-and-after imagery confirms six radar heads destroyed, at Chunian, Pasrur, Sukkur, Arifwala, Jacobabad and Gujranwala. Each shows the sequence satellite analysts know well: dish present, dish gone, men standing around the spot where the dish used to be.

Cooper’s day-by-day reconstruction isn’t kind to Pakistan. On the evening of 7 May it emptied its magazines, firing three to four hundred drones and Fateh-1 rockets at some thirty locations, for next to no effect. India’s reply the next morning was roughly 160 Harop loitering munitions into three HQ-9 batteries. Cooper judges the Walton battery inside Lahore “largely destroyed”, and the munitions reached as far as the army’s own headquarters district in Rawalpindi. He also spots the flaw in Pakistan’s plan: it had two or three days’ worth of drones and rockets and no money to buy more. India could keep going for as long as it liked.

A Harop loitering munition detonating on the HQ-9 SAM site at Walton, inside Lahore, with debris thrown into the air
Walton, inside Lahore, around midday on 8 May: the moment an Indian Harop hits an HQ-9 battery, with debris going up into the air. Circulated footage
Sukkur airfield radar site, before and after satellite imagery showing the radar destroyed
Sukkur’s radar site: an intact dish before, a dark, collapsed footprint after.
Chunian air-defence radar, before and after satellite imagery, radar destroyed
Chunian air-defence radar. Same signature, and this time the frame says it outright: “radar destroyed”.
Pasrur air-defence radar, before and after satellite imagery, radar head damaged
Pasrur air-defence radar, located on the base plan and then in close-up, labelled “radar head damaged”.

One Harop was filmed taking apart a Swiss-made Oerlikon gun. Ground photographs show Pakistani soldiers posing next to their own burnt-out HQ-16 surface-to-air system, and other footage caught a Pakistani SAM battery being towed through the streets of Lahore to keep it safe, which is an odd look for a weapon whose whole job is to stay put and fight. By the night of the ninth the shield had holes in it, and India had a map of every one.

None of this was improvised. Months later the Indian Air Force chief, Air Chief Marshal AP Singh, gave a memorial lecture in Bengaluru and went through the targeting base by base, on the record, with before-and-after frames. A briefing slide that matches the lecture lays out the order: a 02:00 to 05:00 window on 10 May against Chaklala, Rahwali, Rafiqui, Rahim Yar Khan, Sukkur, Murid and Nayachor, then a 10:00 to 12:00 window against Sargodha, Bholari and Jacobabad. Each target is tagged with what was hit: SAM, runway, hangar, or “UG C2”, which is the underground command bunker you’re not supposed to know Pakistan has.

02:00-05:00Before dawn: seven airfields

  1. Nur KhanChaklala33.62°N 73.10°EUG C2
  2. MuridChakwal32.91°N 72.77°EUG C2
  3. Rahim Yar Khan28.38°N 70.28°ERWY
  4. RafiquiShorkot30.76°N 72.28°ERWY UAV
  5. Sukkur27.72°N 68.79°ERADAR
  6. Rahwali32.29°N 74.14°EAFLD
  7. Nayachor25.75°N 69.77°EAFLD

10:00-12:00The second wave: three airfields

  1. SargodhaMushaf32.05°N 72.67°ERWY
  2. Bholari25.24°N 68.04°EHGR
  3. JacobabadShahbaz28.28°N 68.45°EHGR RADAR
The two strike windows on 10 May, target by target, with the ceasefire at 17:00 for scale. The order comes from an Indian Air Force briefing slide and matches Air Chief Marshal AP Singh’s on-record account. RWY is runway, HGR hangar, AFLD airfield, UAV the drone complex, and UG C2 an underground command-and-control bunker.

The slide is also the clearest confirmation that the target at Murid was an underground command-and-control bunker, a hardened complex next to the base’s drone hangars, marked as a direct hit. Once your buried bunker turns up, correctly labelled, in the other side’s PowerPoint, it isn’t much of a secret any more.

Murid air base before and after, Maxar satellite imagery
Murid Air Base, before and after. The underground operations centre that India Today identified sat under the building shown here, next to the drone hangars. Satellite image © 2025 Maxar Technologies
Arifwala air-defence radar, before and after satellite imagery, radar head damaged
Arifwala air-defence radar, the fifth confirmed radar kill, with the same signature and the same “radar head damaged” label. Gujranwala makes six.

A longer list, with caveats

A wider tally of Pakistani radar and SAM sites has been going around: thirteen entries, with named equipment and grid coordinates. The exact model identifications and arc-second coordinates go further than I can independently confirm. Where a row matches the imagery above, it’s solid. Where it doesn’t, treat it as a rumour.

SiteSystem (as reported)CoordinatesStatus (as reported)
Arifwala radar siteAN/TPS-77 or YLC-8B30°07′36″N 72°59′26″EStruck and destroyed; matches the imagery above
Chunian radar siteAN/TPS-77 or YLC-8B30°57′25″N 73°58′12″EStruck and destroyed; matches the finding above
Pasrur radar siteAN/TPS-7732°14′19″N 74°41′21″EStruck and destroyed; matches the imagery above
Sukkur radar siteAN/TPS-43G/J27°43′00″N 68°47′20″EMatches the confirmed Sukkur radar strike above
Sargodha radar siteAN/TPS-7732°03′20″N 72°40′05″EReported, near PAF Mushaf, Sargodha
Jacobabad, second radar siteGiraffe 40 (mobile)28°17′08″N 68°27′42″EReported struck and destroyed, separate from the Shahbaz radar below
Shahbaz (Jacobabad) radar siteAN/TPS-77 (radome)28°16′15″N 68°28′00″EReported struck, near PAF Shahbaz
Bholari radar siteLikely AN/TPS-7725°14′49″N 68°01′02″EReported struck, near PAF Bholari
Murid SAM siteLikely FM-9032°55′09″N 72°45′54″EConsistent with the underground-complex strike confirmed above
Nur Khan SAM siteHQ-9BE33°37′16″N 73°05′03″EReported ineffective against the Indian strike
Malir Cantonment SAM siteHQ-9BE24°56′11″N 67°12′37″EMatches the confirmed strike from the sea; see below
Walton Airport SAM siteHQ-9P/BE31°31′N 74°24′EReported: C2 vehicle destroyed
Chor radar siteAN/TPS-77 (reported under construction)25°31′32″N 69°46′27″EUnconfirmed. Reported only as “possibly targeted”

Six of these thirteen rows back up findings already established through named imagery or on-record statements. The rest come from a wider compilation I can’t verify independently. They’re here for completeness, not as settled fact.

Act III Eleven air bases that are completely fine

Then came the blitz, in two rounds. The Observer Research Foundation counts four air bases confirmed damaged on 9 May and four more with limited damage on the 10th. The heavy strikes landed between 02:00 and 05:00 that morning, when India’s air force visited the Pakistan Air Force where it lives. Eleven air bases are confirmed by name, and Sukkur and Skardu, which are harder to confirm from imagery, bring the map to thirteen. Roughly a fifth of PAF infrastructure was destroyed in that three-hour window alone. A Swiss think tank later concluded that the IAF had won air superiority and pushed Pakistan to the ceasefire table. Runways and hangars are about the hardest things there are to hide afterwards. You can’t say a runway is fine while it has a crater in it, and you can’t frown a hangar off a satellite image, though the Ministry has tried. Here is what the satellites keep finding, base by base.

Jacobabad: the hangar that opened

This is the clearest single frame of the whole war. Zoom in on the apron at Jacobabad: an intact hangar in 2024, its roof blown open in 2025, and the aircraft that used to be parked outside now somewhere else.

Satellite image, August 2024: the hangar at PAF Base Shahbaz with its roof intact and two aircraft on the apron Satellite image, September 2025: the same hangar with its roof torn open AUG 2024 SEP 2025 HANGAR ROOF PAF BASE SHAHBAZ, 28.28°N 68.45°E
PAF Base Shahbaz at Jacobabad. Drag to compare. The hangar is whole in 2024. In 2025 the roof is torn open to the sky, in a district that reaches 50°C in summer. Six months later the hole was still there, still unrepaired and still making headlines, which is a long time to leave the roof off a building nothing hit.Esri World Imagery, dated captures, Aug 2024 and Sep 2025
Reference photo of the same Jacobabad hangar with an F-16 inside, next to the KawaSpace satellite damage image, matched to a PAF Base Shahbaz ceremony photo
How do we know it was that hangar, and what was in it? OSINT analysts matched the damaged roofline to an older ceremony photo taken inside the same building. It shows the “PAF BASE SHAHBAZ” sign, and an F-16 parked exactly where the roof later came down.
Jacobabad airfield briefing slide showing the aircraft hangar before and after the strike
A fourth angle on the same hangar, before and after: intact, then punctured and scorched. By now it’s better documented than most royal weddings.
Ground-level view of PAF Base Shahbaz at Jacobabad, the control tower lettered SHAHBAZ with the main hangar's collapsed roof behind it
From the ground: the control tower still lettered SHAHBAZ, and right behind it the main hangar’s roof folded down into a tangle of steel. Circulated footage

Nur Khan: next door to army headquarters

Maxar satellite images of Nur Khan dated 25 April and 10 May 2025, with the struck building boxed
PAF Base Nur Khan (Chaklala), Rawalpindi, a short drive from army headquarters. NDTV’s Maxar comparison, 25 April and 10 May, boxes the building that stopped existing. The target was the base’s underground air-defence command bunker. The operations complex above it was gutted, including two Türkiye-built decision-support centres. Pakistan then demolished the wreckage instead of repairing it, which is not what you do with a building that’s fine. Satellite image © 2025 Maxar Technologies, via NDTV
Chaklala (Nur Khan) briefing slide showing the operations centre destroyed, before and after satellite imagery
A third confirmation, in official-briefing style: the operations centre at Chaklala, boxed in the before frame and captioned “Operations Centre Destroyed After Attack” in the after.

Hitting Nur Khan was the boldest single move of the operation. A precision weapon went into a base minutes from the centre of the Pakistani military, and the message about reach was hard to miss. Cooper’s reconstruction adds another loss on the same apron: a C-130 Hercules transport, set on fire.

Night silhouette of a C-130 Hercules transport against a fireball at Nur Khan air base
Nur Khan at night: the nose and high wing of a C-130 Hercules, silhouetted by the fire burning it, minutes from army headquarters. Circulated footage

Bholari: a hangar, a $300 million aircraft and a name

Bholari airfield briefing slide showing the aircraft hangar before and after, roof scorched and collapsed
PAF Base Bholari, Sindh. The hangar before and after: intact, then scorched black and collapsed inward. India put a crater through it, and inside was a Saab 2000 Erieye airborne early-warning aircraft, one of a very small fleet. A retired Pakistani air marshal, doing sums nobody asked for, said the hangar took four BrahMos and put the loss at about US$300 million. He, at least, was off script.
OSINT geolocation matching the Bholari hangar hit by the IAF to a formation photograph of Pakistan Air Force personnel and aircraft
OSINT geolocation matched the Bholari hangar to a PAF ceremony photograph, with the same apron marking and the same hangar, and labelled it “Hangar Hit By IAF”.

Bholari is also where the denials run out, and where I stop joking. Pakistan’s own military conceded that five airmen were killed here, and Pakistani sources named one of them: Squadron Leader Usman Yusuf. A state doesn’t concede dead men by base, number, rank and name at a place it says was never touched. If anything at Bholari is absurd, it’s the officials who still tell his countrymen it didn’t happen.

Sargodha: craters on the master jet base

Sargodha airfield briefing slide showing two damaged runway areas, before and after, with the surface under repair
PAF Base Mushaf, Sargodha, home of the F-16 fleet. Two crater sites on the airfield plan: area 1, where a repair crew is already resurfacing, and area 2, where the crater scar still shows in the grass. OSINT analyst Damien Symon published matching imagery. One of the six airmen Pakistan admits losing died here.

Beyond those four, India also hit Rahim Yar Khan (runway cratered, field closed). A footnoted monograph from the Centre for Air Power Studies (CAPS) names all eleven bases and adds Rafiqui (runway and drone-control complex), Sialkot, and Murid, where the target was another underground bunker next to a hangar full of Turkish and Pakistani drones. Masroor on the Karachi coast and two bases that barely make the English-language news, Rahwali and Nayachor, complete the eleven. At Skardu the CAPS study admits it can’t confirm the exact target because there’s no imagery, which is more honesty than Pakistan’s information ministry managed all year. India says it downed as many as thirteen Pakistani aircraft across the operation. For a while that was only New Delhi’s word, but the air-to-air picture has since cleared, and it cleared in India’s direction, largely through footage Pakistanis filmed themselves (see below). Pakistan claimed six Indian jets and produced wreckage from none of them.

Rahim Yar Khan airfield briefing slide, area 1 runway craters and area 2 damaged building, before and after
Rahim Yar Khan. Two aim points on the airfield plan: area 1, a pair of fresh craters straight across the runway, and area 2, a support building captioned “Building Damaged”, with two neat impacts on its roof. The field was closed by NOTAM, which is aviation’s official way of saying “please don’t try to land in the holes”.

While the air force scrambled, the navy stuck to a doctrine strategists might call staying exactly where the boats were. India had moved its Western Fleet (a carrier, destroyers, frigates and anti-submarine ships) into the northern Arabian Sea, within strike range of Karachi. Its Director-General of Naval Operations said the fleet was in position to hit the city “at a time and place of our choosing”. You say that when you’ve already checked and nobody is coming out. The Pakistan Navy studied the threat and answered with the one manoeuvre it had practised to perfection: mooring.

It wasn’t a bluff. Air Marshal AK Bharti confirmed on the record that a missile site at Malir Cantonment, an army garrison inside Karachi, had been struck, and a Pakistani journalist added that it “was carried out from the sea, using the warship INS Vikrant”. So the air force and the navy reached the same city from two directions, and once again it was a Pakistani commentator who confirmed it.

You could call a fleet that sits at anchor for four days within reach of an enemy carrier “preserving its strength”. You could also call it hiding. Either way, a fleet that never sails can’t be sunk, and the Pakistan Navy kept that record intact from the pier.

What fell out of the sky

For weeks the air-to-air score was the murkiest part of the war. Both sides claimed kills and neither showed the wreckage. When that fog lifted, it lifted the wrong way for Islamabad. The footage that surfaced after May 2025, geolocated and cross-checked frame by frame, shows Pakistani aircraft coming down over Pakistani soil, including the two things an air force can least afford to lose: its flying radar and its pilots.

The Erieye over Dinga

The most valuable aircraft Pakistan has isn’t a fighter. It’s the Saab 2000 Erieye, the flying radar that tells every other jet where the enemy is. On 10 May one was tracked and destroyed in the air over Dinga by an Indian S-400 firing its 40N6E interceptor, the 400-kilometre missile. It was that missile’s first combat use anywhere, and the kill is credited as one of the longest-range air kills on record. Circulated video caught the interception and the fireball. Other clips caught debris falling toward Gujranwala, and residents who took the two tumbling halves of the aircraft for two falling jets. For once, Pakistan chose not to adopt the bigger number.

The S-400 engagement over Dinga, frame by frame: the interception, the fireball, and the interceptor’s own casing recovered in a field, its stencilled serial confirming the 400-kilometre 40N6E round. Another Erieye was reported down the same day over Kohat. Compiled from circulated footage

Pakistan can’t absorb this loss. It flies only a handful of Erieye aircraft, and India had already burned one on the ground at Bholari. Take a second one out of the sky (a third, if Kohat holds up) and you’ve blinded every fighter that relied on them. An air force without its eyes is feeling its way around a dark room while announcing on national television that it can see perfectly, and then walking into the furniture.

A JF-17 over Dinga

Fighters came down too. Over the same fields at Dinga, a JF-17, the Sino-Pakistani mainstay of the fleet, hit the ground. The debris included the scorched exhaust nozzle of its Russian RD-93 engine and a blackened avionics box, which a villager carried off. He now owns more of a JF-17 than he expected to this year.

Recovered JF-17 debris: a blackened avionics box and the RD-93 engine nozzle
Debris recovered near Dinga: a charred avionics box in a villager’s hands, and the nozzle of a JF-17’s RD-93 engine. Neither comes off an aircraft that’s fine. Stills from circulated footage

Then came the PAF’s finest piece of accounting. Instead of admitting it lost a jet at Dinga, it held a press conference to announce that it had “intercepted a missile”, and showed off the wreckage of a 40N6E as a trophy. That’s the long-range interceptor built to kill aircraft, and it doesn’t fall on open farmland unless it has just been fired at something in the sky. What came down at Dinga burned in a fireball far too big for the “drone” that local police first reported. If India had really lost a drone there, Islamabad would have paraded it from one end of the country to the other instead of waving a spent missile casing. Showing off the interceptor that just shot you down, as proof you won, is like getting mugged and holding up the mugger’s business card to show you came out ahead.

The Mirage argument

Then there’s the Mirage. A field of wreckage, including a wing panel stencilled 24-045, a shattered engine and an ejection seat, went round the internet with competing captions. Indian accounts called it a downed PAF Mirage III/V. Pakistani accounts said, with the confidence of a losing hand, that it was an Indian Mirage 2000, or maybe a MiG-29, and the story changed by the hour. The argument lasted exactly as long as it took someone to look at the engine.

Wreckage engine compared to reference photos: an Atar-9 engine (Mirage III/V) and an RD-93 engine (JF-17), neither of which powers an Indian jet
The wreckage engines matched against reference photos: a SNECMA Atar-9, which powers the PAF’s Mirage III/V, and a Klimov RD-93, the JF-17’s engine. No Indian aircraft uses an Atar-9, and the attempt to pass it off as a French Snecma M53 or a Russian RD-33 fell apart once the parts were laid side by side. It was Pakistan’s Mirage. OSINT engine comparison, circulated

The neatest confirmation came from the most restrained source available. When India’s Directors-General of Military Operations held their formal briefing, the one that took pains to say the fight was “with terrorists, not the Pakistani military”, the first exhibit on screen was video of the debris of a downed Pakistani Mirage. India spent that press conference playing down what it had hit and still opened with a Pakistani fighter in pieces on the ground. When even the polite version of events starts with your jet in a field, the argument is over.

Collage of Pakistani Mirage wreckage panels, engine parts and structural fragments, one panel stencilled 24-045, captioned The Pakistani Mirage
The compiled debris: the airframe panel numbered 24-045, the Atar-9 engine and structural fragments. It’s the same wreckage India showed at its DGMO briefing, labelled “The Pakistani Mirage”. Debris frames circulated after the strikes; the Mirage was also shown in India’s official DGMO briefing

The counter nobody told to stop

The funniest piece of evidence in the whole war belongs to neither government. Martin-Baker, the British firm that builds most of the world’s ejection seats, keeps a public “lives saved” counter on its website. It goes up by one every time a pilot pulls the handle and lives. It’s a nice bit of engineering pride, and, unluckily for Rawalpindi, a tamper-proof count of how many aircrew had to abandon an aircraft.

Martin-Baker public lives saved counter showing a Mirage silhouette and the total 7784
Martin-Baker’s public “lives saved” counter, shown here at an earlier reading. It stood at 7,789 when the war opened on 7 May and had reached 7,792 by the ceasefire on the 10th. Of those saves, five were logged with no crash details attached at all. Martin-Baker public ejection record (martin-baker.com)

Here’s why that matters. Martin-Baker only logs successful ejections, where a pilot leaves a doomed aircraft and survives, and under standard confidentiality agreements with air forces it withholds the where, the when and the what for any loss in active combat. Across the 88 hours, the counter moved by several ticks with the details left blank. Which aircraft in this fight sit their pilots on Martin-Baker seats? Pakistan’s F-16s, JF-17s and Mirage III/Vs, for a start. The blank ticks line up with Pakistani pilots getting out of aircraft that Islamabad insists it never lost. Pakistan’s government issued denial after denial, while an ejection-seat maker in Buckinghamshire, updating a web page, kept the receipts.

Pilots in the fields

Then there were the crews. In the days around 10 May, video after video showed Pakistani villagers combing fields and canals for their own downed airmen: a pilot filmed parachuting down and then searched for on the ground, another confronted by villagers in Sialkot after being found at night, and civilians in Bhimber hunting for aircrew near where a missile booster had come down. In the same Sialkot fields the booster of an Indian MR-SAM (Barak-8) interceptor turned up with its serial number still legible, which shows how hard, and how close to Pakistan’s own towns, the air battle was fought.

A man crouches beside the fin section and serial-numbered booster of an Indian MR-SAM interceptor recovered in a Sialkot field
Sialkot, 10 May: a resident looks over the fin and serial-numbered booster of an Indian MR-SAM (Barak-8), found in open ground near where villagers were searching for a downed pilot. Still from circulated footage, geolocated to Sialkot

The home videos

None of this came with an official press release. It’s footage Pakistanis filmed themselves, the fireballs, the debris and the searches, and posted before anyone in uniform could suggest they hadn’t seen anything. Every clip plays.

JF-17 debris, Dinga
Fireball, Dinga
Strike explainer, Dinga
Night impact, Gujranwala Cantt
Erieye wreckage, Kohat
Villagers on the crash ridge
Aircrew search, Sialkot
Pilot found at night, Sialkot
Rangers searching in fog
Missile-booster debris on wet groundBooster debris, Kotla
Indian MR-SAM interceptor booster, second viewMR-SAM booster, second view

By the last day, according to Cooper, the PAF was stretched so thin that it pushed its K-8 Karakorum jet trainers into the fight. These are the aircraft of its own “Sherdils” aerobatic display team. It bolted 23-millimetre gun pods underneath and sent them up on combat patrols to swat Indian loitering munitions, with what Cooper drily calls “unknown success”. Picture it: the airshow squadron, the one that draws smoke trails in the national colours, handed live gun pods and scrambled to chase robots out of the Punjab sky. An air force that had won the war doesn’t end it flying its display team at drones, watching its own citizens film the wreckage and searching its own fields for its own pilots. Cooper’s verdict is that India called the nuclear bluff, gutted Pakistan’s ability to retaliate and pushed the PAF back from the border, and that by the ceasefire the balance between the two states had changed for good. You don’t have to take his word for it. Take the tenders.

The reckoning, in Pakistan’s own paperwork

For weeks the military held its line: minimal damage, a few stray hits, Indian jets falling by the dozen. Then the evidence started arriving, and it came from Islamabad, not Delhi.

  1. 30+ tenders

    Within weeks the air force was putting out more than thirty urgent contracts to rebuild runways, hangars and buildings at Nur Khan, Jacobabad, Bholari, Sukkur, Sargodha, Masroor and Rafiqui, and at Rawalpindi, Kallar Kahar and Risalpur. If you’re keeping count, that’s thirty-one contracts to repair damage that didn’t happen, and none of the forms explain what the money is for.

  2. December

    Foreign Minister and Deputy Prime Minister Ishaq Dar confirmed on the record that Indian strikes had hit Nur Khan, injuring personnel and damaging the base. That’s a cabinet-level admission from the second most senior man in the government, after seven months of everyone else saying the opposite.

  3. 6 airmen

    The military admitted that six of its airmen were killed, five at Bholari and one at Sargodha, and one of them was named. States don’t concede dead men, by base and by rank, at installations they say were never touched. This one isn’t a joke, and it never was.

  4. $300M

    One of Pakistan’s own former air marshals described the Bholari hangar taking four BrahMos and put the destroyed early-warning aircraft at about $300 million, with wider losses “well over a billion dollars”. That’s a very specific figure for damage that officially rounds to zero.

  5. 14 dead

    The Jaish chief’s public grief for fourteen people killed at Bahawalpur is, at the very least, an admission that the “mosque” was full of his people.

Put together, it’s the most thorough self-incrimination a losing side has ever filed in a South Asian war. Its ministers, its procurement office and its funerals read the losses out loud, and in the same breath the state insists there weren’t any. A prosecutor writing Pakistan’s press releases himself couldn’t have got a fuller confession.

One more claim I can’t confirm

Indian outlets citing what they say is a leaked internal PAF assessment go further: total losses of roughly $7.6 billion across seven air bases, adding Minhas (Kamra), Faisal (Karachi) and MM Alam (Lodhran) to the confirmed six, plus a critically damaged electronic-warfare jet at Nur Khan. None of this is in the CAPS or ORF assessments, and “leaked internal report” tells you where a claim came from, not whether it’s true. I’m keeping it apart from the confirmed record because, unlike a Pakistani press briefing, I’d rather tell you which parts I can’t stand behind.

The scoreboard Pakistan would rather you didn’t see

  • 6Airmen
    admitted lost
  • 11Air bases
    all fine
  • 6Radars
    lying down
  • 9“Mosques”
    wx damage
  • 2Militant HQs
    at ground level
  • ~600Drones sent
    few arrived
  • 30+Repair tenders
    nil damage
  • $300MAEW jet
    hangar fine
  • 23 minAir defence
    transit time
  • 0Navy
    sorties
Every lamp on the panel is lit. Officially, nothing is wrong. Red is a warning, amber a caution, blue an advisory. WX is aviation shorthand for weather.

Every site in this piece

TargetTypeCoordinatesWhat India destroyed
Nur Khan / ChaklalaAir base33.62°N 73.10°EOperations and C2 complex gutted; later demolished
BholariAir base25.24°N 68.04°EHangar cratered; Saab Erieye AEW&C destroyed; Sqn Ldr Usman Yusuf and 4 airmen killed
Shahbaz / JacobabadAir base28.28°N 68.45°EF-16 hangar roof torn open (visible in imagery); radar; ATC damage
Mushaf / SargodhaAir base32.05°N 72.67°ERunway craters; 1 airman killed
Rahim Yar KhanAir base28.38°N 70.28°EHangar, radar and runway cratered; field closed
Murid (Chakwal)Air base32.91°N 72.77°EHDBT underground C2 node and TB2/Shahpar-I UAV complex destroyed (CAPS, Maxar, India Today)
SukkurAir base27.72°N 68.79°ERadar destroyed; specialised hangar and shelter destroyed
Rafiqui (Shorkot)Air base30.76°N 72.28°ERunway and UAV control complex targeted (CAPS)
SialkotAir base32.52°N 74.36°EMaintenance infrastructure targeted (CAPS)
Skardu (Qadri)Air base35.34°N 75.54°EHosts F-16 and JF-17 squadrons; exact target not confirmed because no imagery is available (CAPS, which says so)
RahwaliAir base32.29°N 74.14°EAirfield infrastructure struck
NayachorAir base25.75°N 69.77°EStruck in the 02:00-05:00 wave (IAF briefing)
MasroorAir base24.90°N 66.94°EOn Pakistan’s own repair-tender list
Malir Cantonment, KarachiSAM site~24.88°N 67.18°EStruck from the sea by INS Vikrant (Air Marshal Bharti, on record)
HQ-9 SAM systems, Lahore and SialkotAir defence~31.6°N 74.3°EChinese-built SAM batteries neutralised
HQ-16 (LY-80) SAM batteryAir defencen/aDestroyed; wreckage photographed with PAF personnel present
Radars at Chunian, Pasrur, Sukkur, Arifwala, Jacobabad and GujranwalaRadarvariousSix radar heads confirmed, each by before-and-after imagery or ORF confirmation; more reported in the longer list in Act II
Markaz Taiba, MuridkeMilitant HQ31.80°N 74.25°ELeT headquarters destroyed (Maxar, Reuters); ground photos show total collapse
Markaz Subhan Allah, BahawalpurMilitant HQ29.35°N 71.62°EJeM headquarters flattened; domes pierced
The other seven camps: Sialkot, Kotli, Bhimber, Muzaffarabad, Neelum ValleyMilitantPoJK and PunjabAll nine first-wave camps are listed one by one in Act I
Minhas (Kamra), Faisal (Karachi), MM Alam (Lodhran)Air bases (unconfirmed)variousReported only in a “leaked internal PAF report”; see the reckoning above. Not independently corroborated

Every entry is graded by source strength. On-record statements and press-credited satellite imagery outrank think-tank monographs, which outrank OSINT compilations, which outrank a single unverified clip. That’s a higher standard of evidence than any statement from a Pakistani podium during the entire war.

How this keeps happening

The whole thing turned on reach: hitting the targets without flying into them. Rafales fired SCALP cruise missiles. Su-30MKIs carried BrahMos, fifteen to eighteen of them, on an extended booster that takes them well past the advertised range. HAMMER bombs, the RAMPAGE missile, and Israeli-origin SkyStriker and Harop loitering munitions did the close work. According to Cooper, though, India’s real advantage wasn’t any one weapon. It was the wiring. Since 2019 India’s Su-30MKI fleet has flown plugged into a national air-defence network, so its fighters, radars and missile batteries all fought from one shared picture, while Pakistan’s mix of Chinese and Western equipment never talked to itself at all. In a fast air war the side that sees first and shares fastest wins, and the side whose systems are barely on speaking terms spends the next year issuing repair tenders. India’s own summary is that it jammed and bypassed the whole Chinese-built shield and finished the job in 23 minutes. You can’t get a pizza delivered in Rawalpindi in that time.

The wreckage shows who fired what. India recovered fragments of Chinese PL-15 air-to-air missiles, the kind Pakistan’s J-10Cs fire, and a captured, largely intact Turkish-Pakistani YIHA-III drone from the wave of hundreds of cheap drones Rawalpindi sent across the border. That’s hardware recovered on Indian soil, not a press release. India’s own indigenous loitering munitions went the other way in numbers New Delhi has said almost nothing about, but which Cooper judges “certain to have caused heavy losses to the Pakistan Army”. Pakistan says it’s fine about those too.

The bottom line

In four days India flattened two militant headquarters, one of them photographed as rubble both from the ground and from orbit. It blinded a ring of Chinese-built air defences: six radars, a national shield and a garrison inside Karachi. It cratered eleven air bases, two of them hiding underground bunkers that are now a lot less underground. An early-warning jet was turned to scrap, an aerobatic team was sent to war, and the navy spent the whole thing at anchor. Within weeks the air force that says none of this happened was commissioning thirty-one contracts to rebuild it. When the ceasefire was agreed at 17:00 on 10 May, it wasn’t because India had run out of targets. Pakistan can patch the craters, but it can’t take back the paperwork.

So, once more, from one side of the border to the other, and this time without the jokes: your government told you that you won, or that it was close. Then it floated thirty tenders to rebuild its air bases, admitted the Nur Khan strike at cabinet level and buried named airmen. Those aren’t India’s claims. They’re your own receipts. A country that had won a four-day war wouldn’t spend the year after it re-roofing eleven air bases and burying squadron leaders while telling the people who pay for both that nothing was hit.

Believing the victory story doesn’t make you loyal. It lets the same men lose again, hand you the same bill and dare you to notice. Laugh at the spin, which has earned every laugh here, but the number under the joke is real: the money spent on this war, and on the next one, is money not spent on your schools and hospitals. The men who died deserve the truth, and the men still telling you nothing happened don’t deserve your applause. This time the truth is in their own handwriting. All you have to do is read the tenders.

About the images and sources

This piece jokes a lot, but the facts under the jokes aren’t invented. That’s the point. The satellite views are Esri World Imagery, genuine dated captures from before (2024) and after (2025). Crater close-ups belong to the firms that shot them (Maxar, Planet, KawaSpace) and are credited and linked rather than reproduced. The official-style before-and-after slides of airfields and camps, and the Maxar/NDTV comparisons, come from battle-damage graphics circulated after the strikes. The map, timeline and strike-window graphic are my own work. Coordinates are public facility and town locations, the Western Fleet’s position on the map is a schematic, and borders follow India’s official map, drawn from Natural Earth’s India edition. India’s air-to-air claims are New Delhi’s own and are flagged as such. The Martin-Baker “lives saved” counter is that company’s own public web record. The Pakistani Mirage debris was shown in India’s official DGMO briefing and in circulated OSINT compilations, and the engine comparison is an OSINT identification, not a lab finding. The jokes are mine; the receipts are Pakistan’s. The reporting draws on:

Back to all investigations Write to me about this piece