Water as leverage

In the closing days of 2024, the Chinese cabinet approved what is set to become the largest hydropower project ever built: a cascade of stations on the great bend of the Yarlung Tsangpo, just before the river drops into Arunachal Pradesh and becomes the Brahmaputra. Its stated capacity is roughly three times that of the biggest plant now operating anywhere. India responded as it usually does, with a demarche and a request for consultations. The approval came without notice to any downstream state, and that was the point. The dam is a message long before it is a machine.

What India knows about the river above that bend, it mostly buys. Under memoranda renewed every few years, China supplies flood-season readings from a handful of upstream stations, for a fee. When relations soured during a border stand-off in 2017, the data simply stopped, even as the Brahmaputra ran high through Assam. It resumed once the politics did. That episode settled what upstream information actually is: not a technical courtesy between neighbours but a currency, priced, rationed and withheld like any other strategic good.

The value of that currency rises with every turbine. Flood forecasting for Assam and for Bangladesh further down depends on knowing what is happening upstream days, not hours, in advance. Reservoir operators on India's own cascade must anticipate releases they cannot see. Planners weighing new storage in Arunachal Pradesh are, in effect, modelling a river whose upper half is a black box. Roughly a third of the Brahmaputra's dry-season flow originates above the border; once that share sits behind concrete, the black box acquires a shutter that someone else controls.

The gauge is the weapon

Engineers will argue, correctly, that a run-of-river cascade cannot steal a river, and that much of the Brahmaputra's volume is generated by rainfall inside India. But diversion was never the sharpest edge of the project. Storage confers control over timing: when water arrives, how fast, with what warning. Uncertainty about timing is enough to force a downstream state to over-build its own buffers, discount its own forecasts and negotiate everything else, from boundary questions to trade, under a water shadow. Information asymmetry does the coercive work long before any sluice gate moves. That is what it means to hold a river as an instrument rather than an asset.

Whoever holds the upstream gauge sets the price of certainty for everyone living below.

Against this, India's standard repertoire is thin. China has not signed the United Nations watercourses convention, recognises no binding duty to consult, and treats hydrological data as a state secret in law. There is no basin treaty on the Brahmaputra and no prospect of one soon. Protest notes are therefore filed into a void, and each new demarche quietly acknowledges the asymmetry it objects to. Bilateral pleading with an upstream power that faces no audience beyond the bilateral is not really diplomacy; it is a request for patience, renewed annually.

The Mekong offers the useful counter-example. For years, downstream governments protested quietly while upstream reservoirs filled in silence. The conversation changed when independent teams, working from open satellite imagery and altimetry, began publishing estimates of how much water each dam was holding back, in near real time and free to anyone. Upstream denials suddenly had to contend with a public ledger. The lesson is not that satellites solve politics. It is that transparency, once produced and published by a coalition, stops being a favour the upstream power can grant or withdraw.

From protest to coalition

India should convene that coalition for its own rivers. Bangladesh and Bhutan share the basin and share the blindness; the Mekong states have already run the experiment. A standing platform of shared gauges, satellite altimetry, open hydrological models and common flood bulletins, published on a fixed schedule, would cost a fraction of one flagship dam. The output must be public by default and machine-readable, so that insurers, farmers and municipal engineers come to use it daily. A transparency programme whose data is embedded in ordinary decisions cannot be switched off by one capital's displeasure.

Credibility begins at home, and here the Union government must swallow hard. India classifies much of its own river data, and downstream neighbours have long complained that flood readings arrive late or not at all. India has also discovered the appeal of the tap: placing the Indus Waters Treaty in abeyance turned water information into a bargaining chip against Pakistan. Whatever the merits of that decision, a country cannot run a transparency coalition on one frontier and an opacity strategy on another. The norm India proposes must visibly bind India, or it will recruit nobody.

The mechanics are affordable. A dedicated budget line for basin observation, covering small satellites, altimetry subscriptions and a public modelling centre staffed jointly with partner states, would be a rounding error beside the flood damage it could pre-empt; losses in a bad year run to several thousand crore rupees in Assam alone. China should be invited into every layer of the platform, sincerely and publicly. It may decline. The point of a coalition standard is that it exists either way: it can be boycotted, but it cannot be vetoed.

Three steps suggest themselves before the monsoon after next. Table a hydrological transparency compact with Bangladesh and Bhutan, with publication schedules written into the text. Fund the observation platform in the next budget and place its data in the public domain from the first day. And commit India to releasing its own flood-season readings to all neighbours by default, so that the standard has a founder worth imitating. The dam on the great bend will rise regardless. Whether the river's information rises with it is still India's choice.