Four letters have gone from Pakistan's Commissioner for Indus Waters to his Indian counterpart since April 2025 concerning fluctuations in the Chenab, a river that reaches Pakistan through Indian-administered territory. None, he says, has been answered. Syed Muhammad Mehar Ali Shah has called the fluctuations a "strategic hazard”: Pakistan is trying to forecast a river whose upstream reaches lie outside its control, without the operational data the 1960 treaty entitles it to request from India, including at daily or other requested intervals where necessary for operational purposes.

The stakes are rising as the mountains feeding those rivers change. The Indus reaches Pakistan after flowing through Ladakh; the Jhelum also crosses from Indian-administered Kashmir. About %40 of Indus flow comes from snowmelt, according to ICIMOD, the Kathmandu-based mountain research center. Its reports in March found that glacier loss across the Hindu Kush Himalaya has doubled since 2000, with monitored glaciers thinning by as much as 27 meters since 1975.

Meltwater is not the only thing those mountains send down. As glaciers retreat, they leave unstable lakes that can burst and release millions of cubic meters of water and debris in a single surge. Pakistan's GLOF-II programme was designed against a 2013 Pakistan Meteorological Department inventory of 3,044 glacial lakes in Gilgit-Baltistan and Khyber Pakhtunkhwa, 33 of them identified as potentially hazardous, and an estimated 7.1 million people living in the most vulnerable districts. Those are project-baseline figures, not a current national risk assessment; nor does the installation or monitoring of equipment by itself establish complete operational warning coverage, which also depends on commissioning, calibration, integration and maintenance. The hazard does not stop at the border. India runs its own glacial-lake risk assessment across the Himalaya, including Jammu and Kashmir and Ladakh, and neither country's early-warning coverage extends to the others slopes.

Part of that information-sharing is already a treaty obligation. The 1960 Indus Waters Treaty reserves the Indus, Jhelum and Chenab largely for Pakistan while permitting specified Indian uses, and allots the Ravi, Beas and Sutlej to India. It requires regular river-flow and reservoir data, lets either side request operational readings more often, and obliges advance notice of extraordinary discharges and flood flows.

For more than six decades, the Permanent Indus Commission served as the treaty's regular technical channel. The treaty itself survived the wars of 1965 and 1971 and the 1999 Kargil conflict. The mechanism did not fail; regular engagement through it was interrupted when India placed the treaty in abeyance.

India placed the treaty in abeyance on April 23, 2025, the day after gunmen killed 26 people at Pahalgam in Indian-administered Kashmir, and says implementation will not resume until Pakistan "credibly and irrevocably" ends what New Delhi calls support for cross-border terrorism. Islamabad denies involvement in the attack and holds that no party may suspend the treaty alone.

Abeyance is not a status the treaty recognizes. Signed by India, Pakistan and the World Bank, it contains no provision expressly allowing one party to place it in abeyance. In its June 2025 Supplemental Award on Competence, the Court of Arbitration read Article XII(4) as providing that the treaty continues in force until terminated by a duly ratified treaty between the two governments, observed no unilateral abeyance mechanism capable of halting the ongoing dispute-settlement proceedings, and held that India's posture did not affect the court's competence. It considered it unnecessary at that stage to decide broader questions about the meaning or possible legal justification of India's abeyance decision. India rejects the court as illegally constituted.

One exception is on the record. In August 2025, India warned Pakistan of possible flooding on the Tawi, a Chenab tributary the treaty covers, through diplomatic channels and described the gesture as humanitarian. Pakistani authorities issued warnings for riverine areas as the flooding developed. But the underlying exchange of flood information is not a discretionary concession; it forms part of the treaty's information-sharing framework.

A warning sent this way depends on someone in New Delhi choosing to send it. The treaty obliges the exchange regardless, requiring flood-flow information as far in advance as practicable. Mohsin Leghari, a former parliamentarian and minister for irrigation, puts the difference in practical terms.

Leghari said the Permanent Indus Commission had provided standardized, engineer-to-engineer data with an audit trail. Without regular engagement through it, alerts now flow through diplomatic channels, which, he argued, lack technical credibility and allow both sides to question timing or completeness.

Resuming India's full performance and engagement through the existing Commission would restore routine river and reservoir exchange. Existing duties under Articles IV(8) and VI, operationalized through the Permanent Indus Commission under Article VIII, are distinct from additional cooperation under Article VII(1), which requires mutual agreement and formal arrangements. Such cooperation could cover snow and rainfall observations, glacier-lake monitoring and shared satellite data. It would supplement, not substitute for, treaty performance, and should be framed without prejudice to either side's rights, objections and contentions under the treaty, the 1978 Salal Agreement, and pending Court of Arbitration or Neutral Expert proceedings.

"Data capacity is adequate, not binding," Leghari said. "Both countries operate forecasting agencies with telemetry, satellite coverage and gauge networks. Pakistan's gaps are in real-time upstream Indian data, a channel problem, not a capacity problem." That describes the transboundary data gap, but not Pakistan's full forecasting challenge: documented domestic weaknesses in commissioning, calibration, integration and maintenance of early-warning systems also affect the reliability of the warning chain.

The economics showed themselves on both sides of Punjab last year. By early September, floods had affected about 1,900 villages across all 23 districts of Indian Punjab. Across the border, more than two million people had been affected and over 2,000 villages inundated in Pakistani Punjab. Reuters reported that India released water from rain-filled dams and warned Pakistan of cross-border flooding; the extent to which those releases affected downstream flood levels cannot be established without catchment rainfall, reservoir inflow and outflow, and routing analysis.

Disasters on that scale are the argument for spending on warning. Pakistan's 2022 floods, driven by extraordinary monsoon rain over the southern provinces, cost more than $30 billion in damage and losses. WMO puts the average return on early-warning systems at about nine dollars for every dollar invested, and says 24 hours' notice of an approaching storm can cut potential damage by %30. These are global benefit-cost estimates, not Indus- or GLOF-specific measures of warning-system performance.

Pakistan's glacier-risk work already runs on international money: GLOF-II was launched with $37 million from the Green Climate Fund through UNDP and has expanded early-warning and protective infrastructure across vulnerable northern areas. Transboundary cooperation opens another route. The World Bank's Global Facility for Transboundary Waters provides technical assistance, institution-building and investment preparation for shared basins, and a jointly agreed Indus warning arrangement developed through the existing Permanent Indus Commission fits the kind of cooperation it was created to support.

What such cooperation might prevent matters more than what it might fund. Pakistani officials warned at an Islamabad seminar in June that any attempt to divert or reduce their allocated flows would be treated as an act of war, while India has tied restoration of the treaty to its demands on terrorism. Operational protocols supplementing existing treaty obligations are easier to agree on while both governments still treat the issue as an engineering matter. Once a release is routinely read as pressure, there is no technical ground left to negotiate on.

Pakistan has been asking. Leghari argues that regular technical engagement through the existing Commission could resume within months if both governments wanted it: the freeze is a political choice, not a technical impossibility.

Strategic sanity does not require the political disputes between the two countries to be settled first. Those unresolved disputes lie outside the scope of a data-sharing protocol. What technical channels can do is keep working while those disputes remain unresolved, as they did for more than six decades. Resuming regular engagement through the Commission is not a concession either government makes to the other. It is what the treaty already requires.