From River Linking to River Intelligence: What a Digital Operating System for India’s Rivers Could Look Like
This essay follows the published Policy Circle article
On 28 August 2026, Policy Circle published my article, “River-linking plan needs a data backbone”, examining why India’s renewed river-interlinking discussion needs an information and governance layer alongside the engineering proposition.
The Policy Circle article focuses on the policy question: how can transferable water, competing forecasts, interstate consequences and operating decisions be made more transparent and evidence-led?
Read the Policy Circle article →
https://www.policycircle.org/opinion/river-linking-plan-data/
This longer essay takes the argument one step further.
It examines what such a River Intelligence architecture could actually look like through the Saptarishi Framework — connecting hydrology, land, infrastructure, environmental intelligence, economic exposure, institutional authority and disaster response within a federated national operating environment.
The distinction is deliberate:
Policy Circle examines why the data backbone is necessary.
This essay explores how the governance architecture could work.
Download the detailed 8-page policy note →
https://apurvapathak.short.gy/saptarishi-river-intelligence
India’s river interlinking debate is usually framed as an engineering question.
Where is water available? Where is it needed? What reservoirs, canals, pumping systems and transfer structures would be required to move it?
These are essential questions.
But there is another question that should come before them:
Before India connects its rivers physically, can it first connect the intelligence surrounding those rivers digitally?
India’s National Perspective Plan for Interlinking of Rivers identifies 30 proposed links — 14 under the Himalayan component and 16 under the Peninsular component. As of August 2026, the Ken–Betwa Link Project is the only priority link under the plan that has entered implementation.
A national river-linking system would therefore not be one simple canal between the Brahmaputra and Godavari. It would eventually involve a sequence of river basins, reservoirs, infrastructure systems, environmental conditions, state jurisdictions and operating decisions.
The more interconnected the physical system becomes, the greater the need for an equally interconnected decision system.
That is where the idea of River Intelligence becomes important.
Flood management in India is already a systems problem
Consider what happens during a major flood.
Meteorologists monitor rainfall.
Hydrologists monitor catchments and river levels.
Reservoir authorities examine storage and release capacity.
Satellite systems observe inundation.
State governments monitor roads, villages and utilities.
Municipal authorities operate drainage and pumps.
Police and transport departments manage closures.
Disaster-management agencies plan evacuation and emergency response.
Agricultural departments assess crop exposure.
Power and telecommunications operators watch critical networks.
The flood itself recognises none of these institutional boundaries.
Water moves through the entire physical system.
This is why better flood management in India cannot depend only on better flood forecasting.
Forecasting answers:
Where is the water likely to go?
Governance must then answer:
What is there?
Who and what will be affected?
Which infrastructure becomes unavailable first?
What intervention options exist?
Who has authority to act?
What happens downstream after that action?
And eventually:
Did the action work?
These questions require different forms of information to become visible within one decision environment.
India already has much of the digital foundation
A River Intelligence architecture does not require India to begin from zero.
The Government’s C-Flood system already provides two-day advance inundation forecasts at village level, using two-dimensional hydrodynamic modelling together with satellite and ground-based hydrological observations. Its initial coverage includes the Godavari, Mahanadi and Tapi river basins.
The Brahmaputra Board is separately advancing river-basin master planning, flood and erosion management, sediment management and coordination between basin states.
The National Water Development Agency has decades of work on inter-basin links.
The Central Water Commission has hydrological and flood-forecasting capability.
IMD brings rainfall and weather forecasting.
ISRO and NRSC bring earth observation and geospatial intelligence.
States possess enormous operational knowledge of reservoirs, irrigation systems, embankments and local river behaviour.
NDMA, NDRF and State Disaster Management Authorities hold disaster-response responsibilities.
The problem is therefore not simply that India needs another digital platform.
The larger challenge is:
How can these capabilities operate around the same event without requiring every institution to surrender its authority or data to one central system?
That is a digital public infrastructure question.
What the Saptarishi Framework adds
The Saptarishi Framework was originally developed as a seven-layer Digital Public Infrastructure architecture for India’s built environment.
Its central idea is federation.
Different institutions retain responsibility for their own authoritative information, while shared identifiers, standards and APIs allow those systems to interoperate.
Applied to river basin governance, the seven layers become particularly useful.
Atri — physical asset intelligence
Atri identifies and describes the physical assets within the water system:
dams, barrages, embankments, pumping stations, culverts, canals, bridges, drainage structures and control infrastructure.
The question is:
What physical infrastructure exists, what is its condition and what can it safely do?
Bharadvāja — land and settlement intelligence
Bharadvāja connects the hydrological event to land:
parcels, villages, settlements, floodplains, agricultural land, easements, ecological buffers and affected ownership.
The question becomes:
What lies in the path of the water?
Gautama — infrastructure and network intelligence
Gautama sees interconnected systems:
rivers, reservoirs, canals, roads, rail, utilities, ports and logistics corridors.
A flooded bridge is not merely a damaged structure. Its closure may disconnect a district, delay emergency services or interrupt a national transport corridor.
Jamadagni — environmental and geospatial intelligence
This is the core hydrological layer.
Rainfall, terrain, river levels, catchment saturation, groundwater, inundation, erosion, sediment, environmental flows and climate conditions belong here.
Jamadagni asks:
What is happening in the natural system?
Kaśyapa — capital and risk intelligence
A flood has economic consequences.
Kaśyapa can translate exposure into property loss, crop exposure, infrastructure damage, business interruption, insurance risk and avoided-loss scenarios.
This helps decision-makers compare interventions not simply by engineering feasibility but by consequence.
Vasiṣṭha — institutional governance
Information is not action.
Vasiṣṭha connects intelligence to legal and administrative authority.
Who may alter reservoir operations?
Who closes a road?
Who issues an evacuation warning?
Who coordinates across districts or states?
A national operating system must make authority clearer rather than hiding it behind algorithms.
Viśvāmitra — disaster response and resilience
Viśvāmitra converts the shared operating picture into emergency preparedness and response:
evacuation, NDRF/SDRF deployment, critical-infrastructure protection and situational awareness.
Taken together, the seven layers create something significantly larger than BIM.
Why this is not simply a BIM framework
BIM can model a dam.
GIS can model a floodplain.
A hydrological application can model water.
A satellite system can observe inundation.
A digital twin can simulate infrastructure operations.
These technologies are extremely valuable.
But none of them individually defines how different institutions should use their information together, how decisions are authorised, or how the outcome of those decisions becomes institutional learning.
That is the governance gap Saptarishi attempts to address.
The architecture is therefore not:
one enormous national database.
It is:
federated institutions + authoritative data + common standards + secure APIs + clear decision rights.
That is why the river case is such an important test of the Saptarishi Framework.
If Saptarishi works only when coordinating architectural models, it remains primarily a construction framework.
If it can coordinate hydrology, infrastructure, land, economics, state authority and emergency response around one evolving event, it begins to demonstrate itself as a governance architecture.
From flood forecasting to a closed control loop
A River Intelligence system should not end with a dashboard.
It should operate as a closed loop:
OBSERVE → PREDICT → SIMULATE → DECIDE → ACT → VERIFY → LEARN
OBSERVE
Collect rainfall, river-level, reservoir, satellite, soil-moisture and catchment information.
PREDICT
Estimate flows, flood peaks, arrival times, inundation depth and geographic extent.
SIMULATE
Ask what those conditions mean.
Which villages are exposed?
Which road becomes inaccessible?
Which bridge is at risk?
How does a reservoir release change downstream conditions?
What alternative action produces a better outcome?
DECIDE
Present those consequences to the legally competent authority.
The system informs the decision.
It does not replace sovereign institutional judgement.
ACT
Issue warnings.
Protect infrastructure.
Close roads.
Prepare evacuation.
Alter operations where legally authorised.
VERIFY
Compare what was predicted with what actually happened.
LEARN
Update models, operational thresholds, asset information, emergency plans and design standards.
This final step is essential.
A river-intelligence system should become better after every flood.
The harder question: how much water is actually transferable?
The same architecture becomes even more important when discussing India river interlinking.
Traditional water-transfer planning necessarily distinguishes between surplus and deficit basins.
But real rivers are dynamic.
A basin that appears to have abundant water in a long-term planning model may, at a particular moment, also face:
downstream demand;
limited reservoir capacity;
environmental-flow requirements;
sediment constraints;
groundwater stress;
uncertain future rainfall;
or ecological vulnerability.
A national River Intelligence system could therefore supplement static classifications with a Dynamic Transferable-Water Envelope.
Conceptually:
Potential transferable water
= forecast inflow
safely available storage/release capacity
− essential basin demand
− downstream commitments
− environmental flows
− flood-safety reserves
− sediment and morphological constraints
− drought-security reserves
− conveyance and energy constraints.
This is not intended as a final engineering formula.
It is a governance principle.
The quantity of water regarded as transferable should be time-specific, evidence-based, auditable and capable of changing as conditions change.
Why the Brahmaputra demands particular caution
The Brahmaputra demonstrates why simple surplus-deficit language can be inadequate.
Its behaviour involves not only water volume but also erosion, sediment transport, shifting channels, floodplain processes and complex basin-wide relationships.
The Brahmaputra Board’s own current work combines flood and erosion management with sediment management and river-basin master planning.
A River Intelligence system must therefore ask more than:
How much water can be removed?
It must also ask:
What happens to sediment?
What happens downstream?
What happens to wetlands and floodplains?
How does the river morphology respond?
And what happens if the forecast is wrong?
The same discipline must apply at the receiving end.
Moving water towards the Godavari or another recipient basin cannot be considered successful if it merely transfers risk from one geography to another.
Donor and recipient basins must be modelled as one decision.
Digital link before physical link
This leads to perhaps the most practical proposition.
Before a major inter-basin network is relied upon physically, India could operate parts of it digitally in shadow mode.
No new water rights are required for the test.
No reservoir gate needs to move because the model says so.
No physical link needs to be assumed correct.
Instead, a digital representation of selected basins could run alongside real operations.
First: replay history
Major past floods could be reconstructed using available rainfall, river, reservoir, inundation and infrastructure data.
The question would not be:
Could Saptarishi have prevented this flood?
That would be an unrealistic claim.
The useful questions are:
Could the exposed settlements have been identified earlier?
Could critical infrastructure have been protected sooner?
Could an evacuation decision have been made earlier?
Would different reservoir operations have changed the outcome?
Would a hypothetical inter-basin transfer have reduced the flood peak?
What would that transfer have done to the recipient basin?
Then: run live shadow monsoons
During future monsoons, the digital system could record what it would have recommended while existing authorities continue operating normally.
After each event:
forecast can be compared with reality;
predicted inundation with observed inundation;
recommended action with actual action;
and
hypothetical transfer with donor and recipient consequences.
Over several monsoon cycles, this creates evidence.
Some river links may look more compelling.
Some may require modification.
Some may show limited value.
All three outcomes are useful.
River basin governance matters as much as modelling
The challenge is not only technological.
River basin governance in India is inherently federal.
Water crosses administrative boundaries.
A future interconnected river network could increase the number of decisions that affect more than one state.
A trusted digital operating picture could make those discussions more transparent.
Every significant variable should have:
an authoritative source;
a responsible institution;
a time stamp;
a confidence range;
and an audit trail.
If two agencies disagree, the system should not conceal that disagreement behind one apparently precise number.
It should make the disagreement visible.
Technology cannot eliminate interstate negotiation.
But it can improve the evidence on which negotiation occurs.
The real objective is not river linking
This may be the most important distinction.
The purpose of River Intelligence should not be to justify physical river links.
Its purpose should be to improve national water decisions.
Even if no new link were constructed, the same architecture could improve:
flood management;
reservoir coordination;
drought preparedness;
urban inundation response;
critical-infrastructure resilience;
agricultural planning;
floodplain governance;
emergency management;
and long-term climate adaptation.
India therefore does not need to choose between “river linking” and “no river linking” before exploring this digital capability.
The intelligence layer has independent public value.
Connect the intelligence first
India has already demonstrated through Digital Public Infrastructure that national transformation does not necessarily require one giant institution controlling everything.
Aadhaar, UPI and other DPI systems demonstrate another possibility:
shared standards;
trusted identifiers;
interoperable interfaces;
federated institutions;
and population-scale digital coordination.
India’s rivers present a much more physically complex problem, but the governance principle may be similar.
C-Flood is already improving inundation intelligence.
CWC, IMD, ISRO/NRSC, NWDA, the Brahmaputra Board, states and disaster-response agencies already hold significant parts of the national water picture.
The question is whether those pieces can increasingly become one trusted decision environment.
That is the River Intelligence proposition.
Before India connects its rivers physically, connect the intelligence surrounding those rivers digitally.
If that intelligence later demonstrates that a physical link is hydrologically sound, environmentally responsible, economically justified and institutionally manageable, the decision will rest on stronger evidence.
If it demonstrates that a link should be modified, deferred or reconsidered, that too is national value.
The objective of a national digital operating system should never be to manufacture a predetermined answer.
Its value lies in improving India’s ability to arrive at the right one.
Continue the discussion
This essay forms part of a wider exploration of River Intelligence as an applied Saptarishi governance model.
Read the published Policy Circle article:
River-linking plan needs a data backbone
https://www.policycircle.org/opinion/river-linking-plan-data/
Download the detailed 8-page policy note:
From River Linking to River Intelligence
https://apurvapathak.short.gy/saptarishi-river-intelligence
The three pieces serve different purposes:
Policy Circle sets out the national policy argument.
This essay explains the underlying Saptarishi River Intelligence architecture.
The policy note sets out the proposed control loop, seven-layer structure, dynamic transferable-water concept and digital shadow-mode Prayoga in greater detail.
The common proposition remains simple:
Before India connects its rivers physically, connect the intelligence surrounding those rivers digitally.
Apurva Pathak, Architect
www.linkedin.com/in/apurvapathak-designgovernance
Author, The Saptarishi Framework
