About Kolkata Water Intelligence

How “Days of Water Left” Works

We compute three scenarios based on current reservoir storage, daily consumption, and inflow patterns:

Home / dashboard

Kolkataimpounds nothing. There is no dam anywhere in its water system: supply is run-of-river abstraction from the Hooghly at Palta, about 22 km north in Barrackpore, plus roughly 110 MLD of deep tube wells. So the "days of water left" question every other city on this platform answers has no answer here - the numerator does not exist. That is not a data gap; it is the shape of the city.

The headline instead is drainage capacity. KMC's own sewerage document states the main network "was designed to discharge a rainfall of 6 mm. per hour", across 180 km of century-old brick sewer with most pumping stations built 50 to 100 years ago. We measure how often the sky beats that, from 26 years of hourly rainfall (232,896 values, 2000-2025): a mean of 31.8 hours a year, but 19.2 hours a year in 2000-2012 against 44.5 in 2013-2025. The threshold slider is the divisor - move it and the count recomputes, because most Indian storm-water codes use 12-25 mm/h and Kolkata's standard is a Victorian one.

Below the hero sits the sewage balance, because for a city whose emergency is drainage rather than scarcity, "where does it go" is the question straight after "how often do the drains fail".

The rainfall is ERA5-family reanalysis, which smooths short convective bursts, so every exceedance count is a lower bound rather than a rain-gauge reading. The 6 mm/h standard comes from a 2009 KMC document, and we hold it to the scope KMC gives it: the British-era brick trunk network, 180 km of it, not every drain in the city. KMC publishes no design standard for the areas added since, so the chart reads on the core city. The standard is configuration with a citation rather than a constant, so a newer published rating is a one-line change.

Rivers

Four channels: the Hooghly, which supplies essentially all the city's drinking water; the Adi Ganga, the original course of the Ganga running through south Kolkata past Kalighat, now largely the engineered channel also called Tolly's Nullah; the Bidyadhari, which drains the East Kolkata Wetlands eastward; and the Saraswati, carried for basin context.

The water quality is the deepest series on this platform - 41 stations and 3,209 samples from 2010 to 2026, roughly quarterly, from the West Bengal Pollution Control Board's EMIS portal. The longest single series is the Ganga at Dakshineswar: 281 samples spanning 28 January 2010 to 7 July 2026. For comparison, Chennai's river page runs on annual CPCB sampling.

Tidal station pairing is unique to Kolkata here. WBPCB samples each of six Adi Ganga points separately at high tide and at low tide, because the Hooghly is tidal this far inland and the same location is a different water body six hours apart. We keep the two apart rather than averaging them - and that is what surfaces the finding: at Bansdroni on 7 May 2026, dissolved oxygen was NIL and faecal coliform 4,900,000 MPN/100ml at high tide, and worseat low tide, BOD 14.53 against 10.75 and faecal coliform 8.4 million against 4.9 million. The board's own observers recorded the water as "Blackish" and "Pungent".

Flood risk

Kolkata's flood risk is drainage-capacity-driven, not release-driven. It has no dam and no barrage, so there is no upstream gate to watch: the trigger is rainfall intensity against the same 6 mm/hour standard the dashboard runs on. Most of the core city drains through a combinedsystem, carrying sewage and stormwater in one conduit, which is the single fact tying the city's flooding, its river pollution and its dependence on the wetlands together. A drainage failure and a pollution event here are the same event.

The page is anchored on KMC's own weekly waterlogging register - the chart its Mechanical Sewer Cleansing wing publishes listing every pocket it sent de-silting machines to, with a borough/ward attribution on every row. The counts on the flood page are read from the register itself and change as it does, because KMC republishes it weekly and overwrites the previous week in place. Flooding in Kolkata is not an annual event; it is a weekly operating condition.

Groundwater

This page corrects an assumption the research started with. Kolkata is not groundwater-poor: a full India-WRIS station census, paged to exhaustion over 2010-2026, finds 703 observation wells and 201,221 readingsacross the six districts of the metropolitan area - 23 in Kolkata district itself, live to 4 June 2026. That is denser than Delhi's 237-well network.

Two districts have gone quiet and are shown as such rather than interpolated over: Howrah has not reported since 30 April 2023, Hooghly since 30 November 2022. Liveness is itself a finding, so per-district recency is published alongside the readings.

Water bodies and lake restoration

5,526 water-body polygons from OpenStreetMap, plus 3,051 GPS-located bodies from the 1st Census of Water Bodies - ten times the census layer Chennai carries. That census matters more here than anywhere: KMC's own working inventory is a departmental tank list compiled in 1993, supplemented by a 2004 aerial map, so the national census is the only modern per-water-body register Kolkatahas and it is not the corporation's.

Nobody agrees how many ponds there are. For 2006 alone there are four published counts: KMC's own list said 3,873, the National Atlas organisation's map census said 8,731, a satellite count said 4,889, and KMC's earlier 1997 list had said 1,786. That last pair is a trap - KMC's count rose because the corporation searched harder, not because ponds appeared. The widely-cited 44% loss comes from comparing the map census against the satellite count in the same year.

The lost water bodieslayer is built differently here for an honest reason: no per-pond inventory of what was filled exists. So it enumerates toponymic evidence instead - localities and streets still carrying a pond's name with no water body mapped within 300 m today. Eleven of sixteen such localities and twelve of twenty-five such streets qualify, which corroborates the 44% figure by a completely different method. Each entry is evidence a named pond is gone, not a death certificate for a specific pond, and every one is flagged as toponymic.

Restoration is a court docket rather than a programme. Four flagship bodies (Rabindra Sarobar, Subhas Sarobar, Lal Dighi and the East Kolkata Wetlands), four projects, and the NGT's November 2017 order appointing KMDA custodian of Rabindra Sarobar - which the Tribunal then upheld against KMDA's own plea to relax it. Water quality for both Sarobars is pulled from our own WBPCB ingest, so this page and the rivers page read the same samples.

Sewage, allocations and commitments

The signature finding sits on the dashboard: by KMC's own statutory accounting, 910 of Kolkata's 1,400 MLD of sewageis treated in the sewage-fed fisheries of the East Kolkata Wetlands - roughly five times what all five of the city's treatment plants manage between them (179 MLD), with 311 MLD (22.21%) untreated or only partly treated. The wetlands are 12,500 hectares of working fisheries that nobody built, nobody pays for, and which lie outside the corporation's boundary, in North and South 24 Parganas. That fact is why this platform models Kolkata as a region rather than a city.

The allocation ledger is unusual because Kolkata has no entitlement document for its own water - run-of-river abstraction is not a quota with a receipt. What exists on paper is what KMC sells onward: 90 MLD to Bidhannagar and 22.7 MLD to Budge Budge, both entitled-but-unverifiable, since neither seller nor buyer publishes delivered volumes.

The commitments register has a real verification calendar for the first time. Alongside two 2021 promises with no follow-up, it now tracks the KMC-SHARP sewerage packages financed under ADB Loan 4584-IND - named contractors, contractual completion dates, and a percentage-progress figure refreshed every six months by a disclosure obligation rather than at the implementing agency's discretion. As of 31 December 2025, five packages had been awarded and commenced, and every one stood at 0.0% physical progress. That is a baseline the next report either moves or does not.

What {cityName} does not have, and why

Several Chennai surfaces are absent here, and most are differences rather than deficiencies. There are no reservoir cards or storage history because nothing is impounded; no shoreline page because Kolkata is a tidal river port some 130 km upstream of the Bay of Bengal, not a coastal city; no tank cascades because the Gangetic delta drains rather than cascades; and no dam-release flood trigger because there is no dam.

Ward-level surfaces - my-ward, ward profiles, the per-ward risk composite - are switched off for a concrete reason. The only public ward geometry carries 141 of KMC's 144 wards. Wards 142, 143 and 144 are missing, and they are not slivers: they account for about 18.93 km², roughly 9.2% of the city. A ward map that silently drops three wards is worse than none, because a resident of ward 143 would get "not found" rather than "not yet mapped".

A full graded comparison against Chennai - every feature scored, with the reason recorded wherever parity is not reachable - is published in the repository at docs/cities/kolkata/parity-scorecard.md, alongside the provenance file at docs/cities/kolkata/data-sources.md.

The long-form account of how this system came to work the way it does is on the Kolkata origins page.

How we classify river health

CPCB publishes two parallelriver-water-quality classification systems, and they don't always agree. Knowing which one we use - and why - matters for reading our river status badges honestly.

Our status badges ("dead", "severely degraded", "degraded", "stressed", "healthy") are computed from current readings via the Designated Best-Use thresholds- not from the PRS Priority list. We take the worst classification across a river's monitored stations and surface that as the river-level status.

Practical consequence: a river on CPCB's PRS Priority list (e.g. the Madurai-Manamadurai stretch of the Vaigai is Priority III) won't automatically render as "severely degraded" here. If the underlying NWMP readings show only Class C/D conditions, the badge reflects that. The PRS designation belongs in the river description as historical context, not as the live status.

Status thresholds follow CPCB's published Designated Best-Use criteria; readings are from CPCB NWMP annual River Water Quality reports.

Known Limitations

• Estimates are approximations. Actual water availability depends on factors not modeled (groundwater extraction, Krishna water transfer, distribution losses, industrial use).
• Utility-published feeds may occasionally be stale (weekends, holidays, portal outages). Each surface shows a freshness indicator, and the freshness registry names any feed that has stopped updating.
• Groundwater data from OpenCity may lag by months. The map always shows the most recent available period.
• Forecasts use ARIMAX (AutoARIMA with inflow/outflow as exogenous regressors) and work best with 2+ years of daily data.
• Risk scores are relative indicators for comparison between wards, not absolute measures of water safety.
• Satellite spread is a summary of surface water extent, not a direct measure of storage volume, water quality, or inflow source. A lake can look broad and still hold less usable water than expected.
• Reservoir catchment polygons are reviewed operational geometries for rainfall context, not official legal boundaries. This matters especially in Chennai's managed canal and transfer system.
• Current satellite context relies on optical Sentinel-2 observations. During persistently cloudy periods, some water bodies may temporarily lose this insight until a radar fallback is added.

Disclaimer

Open Source

Neer Vazhvu is fully open source. The code, data pipeline, and methodology are transparent and available on GitHub. Contributions, bug reports, and data corrections are welcome.

Support this project

Neer Vazhvu is free and open source. If you find it useful, consider supporting us on Patreon to help cover satellite data, hosting, and API costs.