About Surat Water Intelligence

How “Days of Water Left” Works

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

Home / dashboard

Every other city on this platform opens on a question about how much water is left. Surat cannot answer that question, because it impounds nothing of its own. It abstracts from a weir-cum-causeway pond on the Tapi, which is a river reach rather than a reservoir, and the water that fills that pond is released from Ukai dam about 100 km upstream, which the Gujarat Water Resources Department operates rather than the corporation. There is no volume to run down, so a days-of-water headline here would not be merely awkward. It would be undefined.

So the dashboard asks the opposite question: how much room is left before water arrives. Surat Municipal Corporation publishes, hourly and free, a reading and the operational threshold it is measured againstat every link of the chain - Ukai's full reservoir level, the causeway's overflow level, and a danger level for each of five urban khadis. No other publisher on this platform hands over both halves. Every threshold on that card is the corporation's own figure; headroom is the only number we compute, and it is a subtraction.

Two honesty notes carried on the card itself. A level past a danger mark is a trigger for the corporation to act, not a forecast that a given street will flood. And the causeway routinely sits aboveits overflow level during the monsoon and is closed; that negative headroom is rendered as negative rather than clamped to zero, because "submerged" is the true state.

There is no storage history and there never will be. The chart other cities carry is replaced by a note saying so, rather than a promise that it "fills in automatically". Level over a full-reservoir mark is not a quantity of water, and Surat's share of Ukai is not published anywhere.

One structural caveat on the archive: SMC's page shows a rolling window of about ten readings, with no archive, no dated URL and no API. Our record therefore starts the day we began scraping and cannot be backfilled from any source.

Facts

A static snapshot rather than a live pipeline, every entry carrying its own citation. The reuse figures come from the corporation's own dated presentation rather than from press coverage, which reports materially different numbers for the same programme.

One fact is an absence, and it is deliberate. Surat publishes no measured non-revenue water. The national open-data release contains a "losses including NRW" column, but it is exactly 20.0000% of total supply on all 48 monthly rows, and the accompanying "actual supplied" column equals total supply on every row, which contradicts it. It is an assumption sitting in a measurement column. The same release's per-sub-ward consumption column is exactly 0.75 x capacity on all 233 rows. We publish neither, and say why rather than quietly dropping them.

Rivers

The Tapi and the Mindhola, with CPCB's 2022 national monitoring at eight stations. The Tapi's seven Gujarat stations happen to form a clean upstream-to-sea profile, from Ukai down through Mandavi, Kakrapar, Kathore and Rander to the ONGC bridge at Hazira, and reading them in order produces a finding that inverts the usual assumption about an Indian urban river.

The Tapi is not organically polluted as it passes through Surat. BOD sits at or below detection limit at most Surat stations. What climbs is conductivity - 369 to 513 umhos/cm at Ukai, 363 to 7,656 at Kathore, and 1,537 to 49,720 at Hazira, which is seawater. Surat's river problem is salinity and the estuary, not sewage.

Two limits worth stating. Only one CPCB edition is ingested; the annual backfill is available and not yet run. And river geometry comes from OpenStreetMap, which names the Tapi and the Mindhola but none of the five khadisthe corporation monitors against danger levels - so the creeks most central to the flood story have no line on the map. SMC's own GIS holds a creek layer, but it serves rendered tiles only.

Groundwater

94 India-WRIS observation stations across Surat district, 6,563 readings, 1970 to 2026. A 56-year record: deep in time, thin in space.

That shape decides how the page renders. The stations are drawn as click-through points rather than interpolated into a per-zone depth surface, because 94 stations spread across a district cannot support that precision - the same call Madurai made at four stations. The ward-depth and ward-risk choropleths other cities carry are off here, and that is a decision about honesty rather than a missing file.

Two source caveats we correct for rather than inherit. The block column is empty for every Surat row in the WRIS export, so block-level assessment cannot be derived from it. And telemetry readings arrive with mixed sign; raw values are carried unaltered so the correction stays visible rather than baked in.

Water bodies

3,401 polygons from ISRO's Space Applications Centre wetland atlas, 1,434 of them inside city limits, carrying area, an inland-versus-coastal split, a man-made-versus-natural split and turbidity flags.

It is a strong geometric base and a weak semantic one: only 34 of the 3,401 carry a name, and the source's type-classification fields are empty for every Surat feature. There is no census of water bodies for Suratequivalent to Chennai's or Bengaluru's, no restoration register, and no lost-bodies study - so the ranking, census and lost-bodies layers are absent rather than thin.

There is also no catchment atlas, and that is a judgement about the city rather than a missing dataset. The cascade layer reconstructs chained tank systems - the Tamil kanmoi networks, the Bengaluru kere chains - where each tank's surplus was engineered to feed the next over centuries. Surat's water bodies are coastal wetlands, tidal creeks and urban talavs on a flat estuarine plain. The algorithm would find downhill neighbours here because it always does, and drawing them as a cascade would assert an inheritance the city does not have.

Flood risk

Flooding is not a sub-page for Surat; it is the front door, and the dashboard hero is already a flood surface. This page carries the spatial half: the corporation's own depth-classed footprint of the August 2006 inundation, alongside the live khadi network.

You will not find a single number about the 2006 flood anywhere on this site, and that is deliberate. It is the defining event in Surat's water history and the obvious thing to render today's dam release against. But every figure for it - the peak discharge, the share of the city inundated, the death toll, the losses - currently traces to encyclopaedia entries, news coverage or advocacy reports rather than to a primary record. Our rule is that external numbers are primary-source verified before publication, so these wait for the People's Committee on Gujarat Floods report and the Surat Citizens' Council Trust report. The footprint is the half we can source, so the footprint is the half we show.

Commitments

A short register, because Surat's public promises are concentrated rather than scattered. All three entries come from one document and all three are owned by the corporation itself: reuse 70% of treated wastewater by 2030, reach 100% and zero liquid discharge by 2035, and achieve comprehensive sewerage coverage by 2033.

The sewerage entry carries a caveat the corporation's own slide supplies: its coverage percentages are marked "before city limit extension in June 2020", and that extension took SMC from roughly 326 to 462.149 sq km. The denominator moved after the percentage was calculated.

Status changes only against a dated citation, and history is kept rather than overwritten.

What is missing, and whose gap it is

No ward surfaces. "Ward" in Surat means three incompatible things: 30 electoral wards, about 134 census and administrative wards in SMC's own 1961-2011 table, and a third scheme inside the corporation's GIS. None has downloadable geometry. SMC's GIS is the richest municipal system we have found on this platform - roughly 390 layers including the full water and sewerage networks - but it serves rendered images and attributes only; the setting that would release the boundaries themselves is switched off. So the analytical unit here is the zone, of which there are nine, and which is the only unit carrying live data, an official current population and the city's own supply breakdown.

A vintage conflict we did not smooth over.SMC's Zones page and its GIS both describe nine zones, with South split into A and B. The live rainfall feed still reports eight, with a single South Zone. Neither surface has been reconciled by the corporation, so the flood-chain data carries the feed's eight names exactly as published rather than remapping them onto the nine.

No allocation ledger.The ledger's whole subject is entitled-versus-received, and no published drinking-water entitlement from Ukai or the Tapi was found. The entitled half does not exist in public.

No restoration ranking. SMC restores lakes and says so - its reuse programme routes 2 MLD of treated water to lake rejuvenation - but publishes no project list, dates, budgets or per-body status.

No industrial effluent layer yet.The Gujarat Pollution Control Board's continuous-monitoring dashboard and the Gujarat Environment Management Institute's discharge-point monitoring for the Pandesara and Sachin common effluent treatment plants are both identified and neither is built. For a textile city this is the most significant outstanding layer.

Full provenance, including the traps in each source, is in the project's facts page and in the per-city data-sources document in the repository.

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.