Consider the horizon this morning from Ambala Air Force Station — the oldest airbase the Republic inherited, Arjan Singh’s first command, home to the Hunters of the Golden Arrows in 1957, to two squadrons of Jaguars since 1979, and to the Rafales of that same resurrected No. 17 Squadron since 2020.
The Strait of Hormuz has been effectively shut since 28 February 2026, when the American and Israeli air conflict with Iran began: six transits on 6 September against a pre-crisis norm of about eighty-five a day, with Brent near $107. On 11 and 12 September the Houthis completed their sweep of Yemen’s Red Sea coast, taking Mocha and overrunning Mayun island, leaving the alternative route out of the Gulf under artillery on both approaches. On 7 August, Saudi Arabia, Pakistan and Turkey signed the Mecca Joint Defence Agreement (an attack on one to be treated as an attack on all three), building on the Riyadh pact of September 2025 that the commentariat christened an Islamic NATO. And on 12 and 13 September, at Bharat Mandapam, the eighteenth BRICS Summit adopted a 140-paragraph New Delhi Declaration by consensus, with President Pezeshkian and the Abu Dhabi Crown Prince in the same room and much photographed warmth between our Prime Minister and President Xi.
It is a spectacular horizon.
A horizon is not a front
None of it materially alters the relative strength of the Indian and Pakistani air forces.
Weather is not terrain. Weather is dramatic, it changes daily, it fills the bulletins. The ground you will actually fight over was surveyed long before the weather arrived, and it will still be there after the weather clears. Every item on that horizon is weather.
Take the Mecca pact, the likeliest candidate for a genuine structural change. A Pakistan Air Force contingent of fighter and support aircraft has sat at King Abdulaziz Air Base since 11 April 2026. In the narrowest sense that is a subtraction from Pakistan’s order of battle. But it is small, it is rotational, it is four hours’ ferry from Sargodha, and no Saudi-Pakistani instrument has ever produced anything resembling an Article 5 response. Islamabad did not fight in the opening fortnight of the Iran conflict, preferring the broker’s role: hosting American and Iranian delegations while its aircraft sat on Saudi tarmac. A pact designed never to be invoked does not redistribute combat power.
The straits tell the same story. A closed Hormuz and a hostile Bab al-Mandeb are a first-order problem for our energy bill, our shipping insurance and our naval tasking, and close to a non-event for what happens between Bathinda and Bahawalpur.
As for the BRICS bonhomie: a declaration achieved by not naming either the American strikes on Iran or the Iranian strikes on the Gulf Arabs is an accomplishment of drafting, which I admire professionally. It is not an accomplishment of security. Nothing in those 140 paragraphs shortens a missile’s no-escape zone or repositions a single early-warning aircraft.
Set the horizon aside, then. The question was always decided by three columns, and I would title them aircraft, airmen and avionics.
Aircraft: the tables that flatter us
The published inventories give India roughly 588 combat aircraft against about 469 for Pakistan, or roughly 600 against 421 on a different compilation. The disagreement is itself the lesson: publications count trainers, strike aircraft and stored airframes differently, and mission-capable rates are classified. The honest reading is that India holds a real but unspectacular advantage of perhaps a third, not the three-to-one asserted in television studios.
That advantage is then taxed twice. Once by geography, since a large share of our fleet must face China, which is why one Rafale squadron sits at Ambala and the other at Hasimara, in the northern extremity of West Bengal, watching over the Siliguri corridor and the approaches to the North-East. The basing chart is our two-front problem rendered in geography. And once by availability: after the MiG-21’s retirement, the Air Force holds twenty-nine fighter squadrons against a sanctioned forty-two. The Multi-Role Fighter Aircraft programme, cleared for 114 Rafales in February 2026, is a process, not a fleet; it will land in the 2030s.
Pakistan has done something quietly impressive with less money: a J-10CE force at the high end, F-16s in the middle, and a JF-17 Block III that now carries an AESA radar, a tactical data link and modern beyond-visual-range weapons, some 156 of them, built and supported at Kamra, giving Rawalpindi one high-volume type it can sustain at home without a foreign clearance for every modification. Cheap is not the same as ineffective.
Airmen: the briefing is a weapon
Our crews were reportedly briefed that the export PL-15 was effective to around 150 kilometres, and I cannot get past what followed. The missile that brought one of our aircraft down was, on Pakistani accounts, launched from about 200, and Indian officials put it further still. New Delhi has acknowledged losses without ever confirming the type or number. That it was a Rafale rests on American officials speaking to Reuters, a later French assessment and wreckage analysis; Indian procurement documents have since been read as pointing the other way.
The aircraft was not defeated. The assumption was. A number in a briefing folder, carried in good faith into a cockpit, placed an experienced crew inside an envelope they believed they were outside of.
This is precisely where the human column then performed. The Chief of Defence Staff said publicly that mistakes had been made and rectified, and they were: within days the tactics changed, stand-off weapons went into Nur Khan and Rahim Yar Khan, and the opening night’s result was not allowed to define the week.
Avionics: whoever sees first
Which brings us to the column that decided the engagement, and that in my assessment will decide the next.
On the night of 7 May, roughly 110 aircraft manoeuvred without either side crossing the boundary in strength. That was two sensor networks hunting each other. Pakistan’s worked well: a Saab Erieye feeding tracks through a national system the PAF calls Data Link 17, allowing its J-10s to shoot while keeping their own radars quiet. Ours faced the harder problem of making French, Russian, Israeli and indigenous equipment talk to one another cleanly. That is the price of a diversified fleet, paid in milliseconds.
The Rafale’s own kit is superb: an AESA radar that Thales says supports precision targeting beyond 100 kilometres, the SPECTRA self-protection suite, Meteor with a ramjet that keeps thrusting into the endgame. None of it substitutes for seeing first.
Nor does the network stop at the airframe. That same week our Integrated Air Command and Control System was working with Akash, Pechora and older batteries against waves of drones, and the Air Force has since claimed a very long-range S-400 engagement, a claim I record as a claim. The contest is now fought by sensors and shooters on the ground as much as by fighters above, and all must share one picture.
Aircraft we can buy. Airmen we already have, and they are very good. Avionics (meaning the network rather than the box) is where the gap is, and the only column the horizon will not fill for us.
I am a civil servant and not an airman. The nearest I have come to aviation is as manager of the flying clubs at Patiala and at Amritsar, an assignment that arrives with the district and teaches you about licences and hangar bills, not air combat. I did read electronics and telecommunication engineering, and taught microwave techniques for a year while awaiting the Civil Services result, so the physics beneath a radar is not foreign. Put me before a cockpit screen, however, and I would not decipher a thing on it.
One conclusion, though, needs no technical training.
What we hold today is very good. The Rafale is among the finest non-stealth fighters flying; our ground-based defences held under waves of drones and missiles; and the officer corps showed in May 2025 that it can take a blow on Wednesday and re-plan by Friday.
But very good is a verdict on today, and today is the easy part. The horizon I began with does not change the equation between Ambala and Sargodha this morning. Sustained across a decade, it changes the climate in which that equation is eventually settled: a Pakistan with Chinese munitions and Gulf money, a northern frontier that never relaxes, two of the world’s chokepoints contested at once. Weather that persists long enough becomes climate.
For that ecosphere, very good is not enough. Twice our present network capability, and nearer three times if the two-front assumption is to mean anything at all: not for dominance, which we neither seek nor could afford, but for deterrence, the bare minimum a country of our size ought to take for granted.
As to when that should begin, the honest answer is the oldest one. The right time was yesterday. The next best is this morning.