Ganges Water Treaty 2026: Measuring the Ganges at Farakka as the Renewal Talks Begin

1. The Treaty and the Barrage: What Farakka Is For

Farakka barrage, in Murshidabad district of West Bengal, was built between 1961 and 1975 with a single engineering purpose: to divert dry-season Ganges water down the 38 km feeder canal into the Bhagirathi–Hooghly, flushing silt out of the Kolkata port channel and pushing back salt-water intrusion. Model studies by India’s Central Water and Power Research Station (CWPRS) and the West Bengal river research institute settled the required diversion at 40,000 ft³/s (about 1,134 m³/s) — a figure that has defined every Ganges negotiation since, as documented in the FAO’s review of the diversion’s environmental effects.

The downstream state — East Pakistan, then Bangladesh — always had a competing claim on the same water. After the 1976 dry-season deadlock, a 1977 agreement allocated shares at 10-day intervals with a guarantee clause; interim memoranda followed in 1982 and 1985; and after a gap without agreement (1988–1996), the two governments signed the 30-year Ganges Water Sharing Treaty on 12 December 1996, to run from 1 January 1997. Its dry-season allocation bands at Farakka, per the treaty text as documented in the IUCN study of the Farakka measurement arrangements:

Two features of the 1996 treaty deserve attention from a hydrometry perspective. First, the 10-day period is the treaty’s atomic unit — allocations are set and settled every ten days, which means the measurement cycle must be faster than the allocation cycle. Second, the 1996 text dropped the guarantee clause that the 1977 agreement carried — the provision requiring a minimum release even in exceptionally low-flow years. That clause is now at the centre of the 2026 renewal debate.

2. Why Measurement Is the Treaty’s Foundation

A water-sharing treaty without a measurement protocol is a political document; with one, it is an engineering system. The 1996 treaty chose the latter path, and its protocol is unusually concrete. As documented in the IUCN study, the two sides established a Joint Committee whose teams observe flow daily during the January–May season at:

• Downstream of Farakka barrage — the total flow remaining in the Ganges main stem;
• The feeder canal — the diversion to the Bhagirathi–Hooghly;
• The navigation lock — a third regulated outlet;
• Hardinge Bridge, on the Bangladesh side — the cross-check station for what actually arrives downstream.

Observations run in 10-day cycles matching the allocation periods, with an annual joint report submitted to both governments. The design intent is unambiguous: each side must be able to verify the other’s numbers at the section where the water is divided.

The historical record explains why the negotiators were so specific. The 1975–76 dispute that led to the 1977 agreement was, at its core, a data dispute — conflicting readings of what the Ganges carried at Farakka in a dry year. And the post-1996 academic literature has continued to flag performance gaps: the IWA journal Water Policy review of the treaty’s implementation found actual Farakka flows below the indicative levels on which allocations were premised in years including 2010 and 2016 — exactly the situation a guarantee clause would have addressed. When the flow drops below the treaty’s lowest band, the treaty’s own arithmetic stops working; measurement is what tells both sides that moment has arrived.

3. 2026: The Renewal Talks Timeline

With expiry approaching, the two sides have moved through three documented technical contacts in 2026, each of which is, in substance, a measurement event:

• January 2026 — a Central Water Commission team hosted a four-member Bangladesh technical delegation at Farakka for an on-site assessment of the measurement arrangements (The Business Standard).
• May 2026 — a six-member Bangladesh delegation joined the 90th Technical Committee meeting of the Joint Rivers Commission in Kolkata, including joint flow measurement on the Ganges (TBS).
• July 2026 — experts from both countries carried out a joint field inspection at Farakka (TBS).

India’s Ministry of External Affairs has stated the talks continue at technical level under the JRC mechanism, which covers all 54 shared rivers (Economic Times). Bangladesh’s stated opening position, per its State Minister for Water Resources, is to seek a share higher than the 1977 agreement, the restoration of the guarantee clause, and stronger information-sharing provisions — with reference to international forums if talks stall (TBS; News18). The Indian side has signalled a preference for a comprehensive renegotiation rather than a simple extension, citing changed conditions — state-level dry-season needs, glacial retreat and groundwater recharge dynamics (The Statesman).

4. The Domestic Dimension: Bihar, West Bengal and the Data Question

The renewal debate is not purely bilateral. Two Indian states have entered the argument, and both arguments are ultimately about flow numbers:

• Bihar — excluded from the 1996 negotiations, the state’s political leadership has argued the treaty costs Bihar water during dry seasons. A JD(U) parliamentarian has claimed Buxar receives only about 400 cumecs while roughly 1,500 cumecs must be released downstream to Bangladesh — figures framed politically and not from CWC’s official record, but indicative of how the state is mobilizing around measurement claims. Chief Minister Nitish Kumar has gone further, publicly questioning the barrage itself (ETV Bharat).
• West Bengal — where Farakka physically sits — the state government’s position is that any future agreement requires the state’s concurrence (ETV Bharat).

For our purposes, the technical significance is this: claims about who gives and who gets water in the Ganges system are only resolvable with independently verifiable discharge series — at Farakka, at upstream stations like Buxar, and at the border. The more stakeholders enter the negotiation, the more the settlement depends on measurement everyone accepts.

5. What It Takes to Measure the Ganges at Farakka

The Ganges at Farakka is one of the harder measurement problems in hydrometry. During the January–May treaty season the flow runs between roughly 50,000 and 75,000 cusecs (≈1,400–2,100 m³/s) through a kilometre-scale, alluvial, actively migrating channel; in monsoon the river transforms into a multi-kilometre floodplain flow. A treaty-grade discharge record must handle all of it:

• Width and access: dry-season braiding means multiple channels and shifting sandbars; a fixed rating curve at a single vertical cannot survive the season, let alone the monsoon.
• Moving bed: the bed load moves beneath the boat — the classic error source that uncorrected discharge measurements inherit, and the reason DGPS bottom-tracking correction matters.
• Monsoon re-rating: after each flood, the stage–discharge relationship must be rebuilt; the treaty’s daily 10-day-cycle observations are what generate those data.
• Speed and auditability: joint teams need measurements both sides can witness, repeat and archive — a full-section profile in minutes, not a day-long cableway operation.

This is precisely the profile of the moving-boat ADCP method standardized in ISO 24578:2021: a 600 kHz-class profiler — the frequency class India’s CWC already specifies for wide sediment-laden rivers (see our analysis of India’s NHP ADCP modernization) — measures the complete cross-section in minutes, corrects for the moving bed via bottom tracking, operates at night, and produces a digital record that a joint committee can audit line by line. For the general moving-boat method behind these measurements, see our dam and reservoir monitoring guide (Section 3.1).

6. What’s at Stake: Performance Gaps and Changing Hydrology

The 2026 negotiation is happening against a backdrop that makes accurate measurement harder — and more consequential — than in 1996:

• Performance gaps on record: the IWA Water Policy review documents years (2010, 2016) in which actual Farakka flows ran below the indicative levels embedded in the treaty’s allocation schedule.
• Changing dry-season hydrology: glacial retreat in the Himalaya and upstream storage development are shifting the shape of the dry-season hydrograph — the treaty’s operating window — on both the Ganges and its transboundary tributaries; India’s negotiators have explicitly cited these changed conditions.
• More stakeholders, more scrutiny: with Bihar and West Bengal now formal voices, and Bangladesh seeking restored guarantees, every allocation claim will be tested against data.

None of these factors changes the treaty’s arithmetic; they change the confidence required in its inputs. A 10-day allocation cycle only functions if both sides trust the 10-day flow record.

7. What the Next Treaty Will Need — From a Measurement Standpoint

Whatever the diplomats settle, the enforcement architecture will have to deliver four things, each with a direct instrumentation counterpart:

The institutional buyers are already identified: the Central Water Commission on the Indian side (which has operated Farakka’s joint measurement and runs the national gauging network), the West Bengal water and port authorities, and on the Bangladesh side the BWDB and hydrology directorate — plus the international consultants who periodically audit transboundary measurement regimes. The treaty’s renewal, in other words, is also a refresh cycle for the measurement fleet at one of Asia’s most scrutinized river sections.

FAQ

When does the Ganges Water Sharing Treaty expire?

The 1996 treaty was signed on 12 December 1996 for a 30-year term and expires on 12 December 2026. Technical-level talks on what follows are ongoing under the Joint Rivers Commission (MEA via Economic Times).

What is the treaty’s minimum water guarantee at Farakka?

Each country is guaranteed 35,000 cusecs (≈991 m³/s) in alternating 10-day blocks during the leanest stretch. Across January–May the treaty applies flow bands: ≤70,000 cusecs shared 50:50; 70,000–75,000 cusecs — Bangladesh 35,000, India the balance; >75,000 cusecs — India 40,000, Bangladesh the balance (IUCN documentation of the treaty).

Where is the Ganges measured under the treaty?

Joint teams observe daily flows downstream of Farakka barrage, at the feeder canal and at the navigation lock, with a cross-check at Hardinge Bridge in Bangladesh. Observations run in 10-day cycles from 1 January to 31 May, with an annual joint report (IUCN).

What is the guarantee clause Bangladesh wants restored?

The 1977 agreement required a minimum release even in exceptionally low-flow years; the 1996 treaty dropped it. Bangladesh’s water minister has sought its restoration — alongside stronger information-sharing — in any renewed deal (News18).

How would ADCP measurement improve treaty compliance at Farakka?

Treaty compliance turns on daily discharge across a kilometre-scale, braided, moving-bed river. An ADCP profiles the full section in minutes, works at night and through monsoon sediment, and bottom-tracks the moving bed — producing the jointly verifiable, audit-grade records that the treaty’s 10-day cycles require, following the ISO 24578:2021 moving-boat method.