Ref: EWS/PA/2026/001
Version 2.0

Policy Advisory: Real-Time Open API Access to Dam Water Level Data for Flood Early Warning

A formal recommendation for saving lives through data transparency

To: Hon'ble Minister, Ministry of Jal Shakti, Government of India; Chairman, Central Water Commission (CWC); Chairman, Bhakra Beas Management Board (BBMB); Chairman, Punjab State Power Corporation Ltd (PSPCL)
From: EWS-Dams-India — Agentic AI Flood Early Warning System (Proof-of-Concept Pilot)
Date:
CC: National Disaster Management Authority (NDMA), Punjab State Disaster Management Authority (PSDMA)
URGENT — PUBLIC SAFETY

1. Executive Summary

During India's 2025 monsoon, dam discharges with inadequate advance warning caused catastrophic flooding across Punjab — over 100 lives lost, 1,900+ villages submerged, and an estimated ₹5,000–10,000 Crore in economic damage. International evidence shows that 72-hour early warning can prevent 80–90% of flood fatalities.

The technology for prediction exists today. The only barrier is data accessibility. This document recommends that the Government mandate real-time, machine-readable publication of dam water level data via open API — enabling automated early warning systems that can save hundreds of lives every monsoon season.

2. The Problem: Inadequate Warning Caused Preventable Deaths

100+ lives lost in 2025 Punjab monsoon due to dam discharges with hardly any credible advance warning to downstream populations.

Impact Category2025 Monsoon EstimateSource
Human lives lost100+State disaster reports, media
Livestock drownedThousandsDistrict damage assessments
Economic damage₹5,000 – ₹10,000 CroreCrop + infrastructure + relief
Villages inundated1,900+Mongabay, Times of India, Indian Express
Advance warning givenHardly any credible warningMedia reports (Indian Express, Tribune)

Constitutional Imperative: Under Article 21 (Right to Life), citizens have a fundamental right to know when dam gates will open. This is life-or-death information that must reach downstream populations days in advance — not hours.

3. The Data Gap: Why Early Warning Systems Cannot Operate Reliably

The barrier to effective flood prediction is not technology — it is data accessibility. Critical dam operations data exists in SCADA/telemetry systems at every major dam, but remains locked behind formats that prevent automated analysis:

Data SourceMethodFrequencyMachine-Readable?API?
CWC India-WRISHTML web pageOnce dailyNoNo
BBMB Daily BulletinHTML / PDFOnce dailyNoNo
IMD Rainfall ForecastHTML / Image mapsTwice dailyPartialNo
Punjab State (PSPCL)Not published at allN/ANoNo
Gate Opening DecisionsPress release / verbalAd-hocNoNo

4. Ranjit Sagar Dam: Proposal for Hourly Data Exposure

Ranjit Sagar Dam (Thein Dam) on the River Ravi is the largest dam in Punjab, with a storage capacity of 3.28 billion cubic metres. It is managed by Punjab State Power Corporation Limited (PSPCL) and sits upstream of Pathankot, Gurdaspur, and densely populated Ravi floodplain villages.

Currently, Ranjit Sagar reservoir data is not published through any public channel. This represents an opportunity — by establishing even a basic hourly data exposure, Punjab State can bring Ranjit Sagar into the early warning network and protect downstream communities during monsoon season.

4.1 Proposed: Minimum Hourly Read Exposure

We propose that PSPCL publish the following data at minimum hourly intervals during monsoon season (June–September):

Data PointMinimum FrequencyFormat
Reservoir level (ft above MSL)Every 1 hourJSON endpoint or public bulletin
Inflow rate (cusecs)Every 1 hourJSON endpoint or public bulletin
Outflow / discharge (cusecs)Every 1 hourJSON endpoint or public bulletin
Gate status (open/closed, count)On change + hourlyJSON endpoint or push notification

Why hourly? Dam levels can rise 2–5 feet within hours during peak monsoon inflow. Daily data (as currently published by BBMB for Bhakra/Pong) is a starting point, but hourly readings enable predictive models to detect danger 48–72 hours in advance — the difference between a successful evacuation and a disaster.

4.2 Benefits for Punjab State

4.3 EWS-Dams-India: Ready to Integrate

Our early warning system is fully built and tested for Ranjit Sagar — the data pipeline, analysis engine, flood simulation, and multi-language alerting are all deployed. The moment Punjab State exposes Ranjit Sagar data through any public channel (API, web page, or even a daily PDF bulletin), our system will activate monitoring within hours. The technology is ready; we only need the data feed.

5. Proposed Solution: Real-Time Open Data API

We recommend the Government mandate a Real-Time Open Data API for all CWC and state-monitored reservoirs. The data already exists in SCADA/telemetry systems at each dam — only the publication layer needs to be built.

5.1 Proposed Data Points and Frequency

Data PointFrequencyFormatLatency
Current water level (ft above MSL)Every 1 hourJSON REST API< 5 minutes
Inflow rate (cusecs)Every 1 hourJSON REST API< 5 minutes
Outflow rate (cusecs)Every 1 hourJSON REST API< 5 minutes
Gate opening scheduleAs decided (push)JSON + SMS pushImmediate
72-hour inflow forecastEvery 6 hoursJSON REST API< 30 minutes
Historical readings (90-day)On-demand queryJSON REST API< 10 seconds

5.2 Implementation Phases

Key Point: No new sensor hardware is required. The data already exists in SCADA systems at every major dam. Only the publication layer (API endpoints) needs to be built — a straightforward engineering task achievable in weeks, not months.

5.3 Technical Requirements

6. Impact: Lives Saved With Early Warning

International evidence (UNDRR, Bangladesh FFWC, European EFAS) consistently shows that warning lead time is the single most critical factor in flood survival:

Warning Lead TimeEvacuation RateDeaths PreventedEconomic Savings
< 6 hours (current)20–35%~30%Minimal
12–24 hours60–75%~60%₹1,000–2,000 Cr
48–72 hours (proposed)90–95%80–90%₹1,500–3,000 Cr

Per major flood event (2025 Punjab scale):
• Lives saved: 60–90 (of ~100 fatalities)
• Livestock saved: 5,000–15,000 animals
• Economic savings: ₹1,500–3,000 Crore

Early warning saves lives, livestock, and movable property. Standing crops and fixed infrastructure cannot be moved — but people and animals can.

7. Recommendations

Immediate Actions

  1. Direct CWC and BBMB to publish dam data via JSON API — beginning with Bhakra, Pong, and Ranjit Sagar
  2. Direct PSPCL to publish Ranjit Sagar data — at minimum, daily reservoir level bulletin matching BBMB standard
  3. Mandate 48-hour advance public notification before planned gate openings via API push, SMS, and public broadcast
  4. Issue data-sharing MoU with verified Early Warning Systems for real-time API access during monsoon

Medium-Term Actions

  1. Establish National Dam Data API — unified platform aggregating all 150+ CWC-monitored reservoirs
  2. Integrate AI early warning outputs into NDMA infrastructure — connect prediction systems to existing disaster communication channels
  3. Define data quality SLA: 99.9% availability, ≤5-minute latency during monsoon season

Policy Framework

  1. Dam Safety Data Transparency Guidelines under Dam Safety Act 2021 — mandate real-time public data access for all major reservoirs
  2. National Flood Early Warning Platform — unified API with standardized alerting protocols
  3. Minimum notification standards — clear timelines, compliance review, and accountability for advance warning failures