The Third Pole — the Hindu Kush Himalayan system — contains more freshwater ice than anywhere on Earth outside the Arctic and Antarctic. It feeds ten major Asian river systems. More than 1.5 billion people depend on those rivers. The glaciers are retreating. The treaty governing the most contested of those rivers is suspended. And both countries holding that treaty are nuclear-armed states whose per capita water availability has been in structural decline for three generations.
T1 The mean annual retreat rate of Hindu Kush Himalayan glaciers is 14.9 metres per year, verified across studies by India's National Centre for Polar and Ocean Research, the Geological Survey of India, and the Wadia Institute of Himalayan Geology. Rates vary by basin: 12.7 metres per year in the Indus, 15.5 in the Ganga, and 20.2 in the Brahmaputra. The WMO's 2024 State of Global Water Resources confirmed 2024 as the third consecutive year of massive global glacier mass loss — 450 billion tonnes melted that year alone. 23 of 24 assessed Himalayan glaciers showed continued mass loss.
The critical variable is not just total volume but timing. As glaciers retreat, they initially produce more runoff — a period scientists call "peak water" — before entering long-term decline. The Indus basin, where glacier and snow melt comprises up to 72 percent of river runoff in the upper reaches, is closest to this inflection point. The Ganga and Brahmaputra, depending more on monsoon rainfall, face different but equally destabilising trajectories: increased flash flooding in the near term, unpredictable seasonal flows thereafter.
T1 Temperature projections for the Himalayan region range from 1.4°C to 6.4°C of warming by end-of-century. Even the conservative end produces 30–50 percent decline in annual Indus basin snowfall by 2071–2100. The 2025 iScience study specifically modelling the Jhelum sub-basin — the artery of Pakistani Punjab — found glacier area losses of 34.7 percent under moderate emissions and 55.3 percent under high emissions by the 2080s, with significant decline in glacier streamflow in both scenarios.
The Karakoram Anomaly deserves a caveat. Parts of the western Karakoram have shown glacier stability or slight advance, likely related to westerly winter precipitation. The mechanism remains imperfectly understood. It does not reverse the regional trend. The Indus tributaries flowing through Pakistani Punjab do not benefit from it.
The Indus basin is structurally more vulnerable to glacial loss than the Ganga or Brahmaputra because it depends on melt rather than monsoon. Pakistan is structurally more vulnerable than India because it is the downstream state — and because the Indus basin irrigates approximately 80 percent of Pakistan's arable land. Per capita water availability in Pakistan has dropped from 5,260 cubic metres in 1951 to approximately 900 cubic metres today. The UN defines water scarcity as below 1,000 cubic metres per person per year. Pakistan is already there.
T1 The Indus Waters Treaty was signed in September 1960 after nine years of World Bank-mediated negotiations. Its architecture divides the rivers themselves rather than the volume: India received the three eastern rivers (Ravi, Beas, Sutlej), Pakistan the three western rivers (Indus, Jhelum, Chenab) — accounting for roughly 80 percent of total system flow. The treaty survived three declared wars, multiple undeclared conflicts, and sixty years of diplomatic hostility. It was the world's most-cited example of two nuclear rivals subordinating military rivalry to shared resource logic.
On April 23, 2025, following the Pahalgam terrorist attack in which 26 tourists were killed, India placed the treaty "in abeyance" — temporarily suspending it, citing national security and Pakistan's alleged sponsorship of cross-border terrorism. Within days, India released water from the Uri Dam unannounced (causing flooding in Muzaffarabad), closed the Baglihar Dam on the Chenab, and the Chenab showed a sharp level drop at Sialkot. Pakistan's Foreign Minister called the move an act of war. Home Minister Amit Shah declared India would "never" restore the treaty.
T3 India's "abeyance" framing requires dissection. The treaty has no abeyance provision. Article XII states the treaty "shall continue in force until terminated by a duly ratified treaty concluded for that purpose between the two governments." India has not terminated the treaty — which requires mutual agreement — but has stopped acting bound by it. This is legally novel and strategically deliberate: it retains restoration as leverage while denying Pakistan the institutional certainty the treaty was designed to provide. The World Bank, designated as facilitator, has confirmed it will not intervene. International legal experts from both countries agree the suspension has no valid treaty basis.
The Himalayan water problem is not bilateral. It is a five-nuclear-power problem wearing a bilateral costume. China controls the upper Brahmaputra (the Yarlung Tsangpo). India is downstream relative to China on the Brahmaputra — a structural vulnerability Beijing has made intermittently explicit through dam construction and data-sharing restrictions. China suspended hydrological data-sharing with India during the 2017 Doklam standoff and again during the 2020 Galwan confrontation. India has no treaty with China governing Brahmaputra water use. No World Bank mediator, no permanent commission, no dispute-resolution mechanism.
T1 China's announced Motuo dam on the Yarlung Tsangpo — projected capacity 60 gigawatts, nearly three times the Three Gorges Dam — will alter the Brahmaputra's seasonal flow, sedimentation, and temperature at a scale India has no legal recourse to contest. India uses the Indus suspension to signal resolve to Pakistan. China uses the absence of any Brahmaputra treaty to build the world's largest dam. These are structurally identical actions — an upstream state exploiting legal ambiguity over a transboundary river — but India condemns one and has no effective response to the other. The asymmetry is not lost on Pakistan, Bangladesh, or Nepal.
The IWT was negotiated in 1960 and never updated for climate change. Its technical annexures assume relatively stable seasonal flows. They do not anticipate peak water, post-peak decline, or glacial lake outburst floods. Pakistan's Gilgit-Baltistan has over 3,000 glacial lakes, 33 classified as potentially dangerous. The 2021 GLOF in Uttarakhand destroyed two hydropower projects and killed over 200 people. The treaty has no provision for any of this. It was not designed for the climate system that now governs the rivers it regulates.
Managed decline with cooperation: India and Pakistan eventually re-engage — on the IWT or a successor framework — incorporating climate projections and joint GLOF monitoring. The most rational outcome and the least politically likely. It requires India to concede the suspension was leverage, and Pakistan to concede the original 80-20 water allocation may need revision. Both involve domestic political costs neither government is currently willing to pay.
Contested scarcity: The treaty remains suspended, India builds hydropower on western rivers, Pakistan accelerates its own dam construction, each side accuses the other of violations, and the glacial timeline makes the fight increasingly zero-sum. The most likely near-term trajectory. Its endpoint is not war in the short term — but it progressively closes the space for the cooperative framework the system ultimately requires.
Cascade failure: A combination of GLOF events, a severe drought year, and the suspended treaty creates a humanitarian crisis in Pakistan's Punjab that overwhelms state capacity. Low-probability in any given year. Rising in probability across a ten-to-twenty-year horizon, given the structural trends. The physics does not negotiate.
The Indus Waters Treaty survived four wars. It did not survive the combination of a terrorist attack, a domestic political calculation, and sixty years without a climate update.
The suspension is legally novel, strategically calculated, and geopolitically reckless in that order. It is legally novel because the treaty has no suspension mechanism and India knows this. Strategically calculated because it gives India leverage over Pakistan's existential anxiety in a way military action cannot. Geopolitically reckless because it weaponises a shared resource in a system that is already shrinking — and will require cooperation to manage at scales neither country has yet contemplated.
The glaciers do not care about the ceasefire. The Jhelum will lose 55 percent of its glacier-fed flow by the 2080s regardless of what India and Pakistan decide about the treaty. The question is not whether the two states will eventually need a new water-sharing framework. They will. The question is whether they build it before or after the first water-driven humanitarian crisis forces them to. The current trajectory is: after.