The 2023 Yamuna flood carried less water than the 1978 flood - and still rose higher in Delhi. That isn't the river behaving worse. It's the river having nowhere left to go.

The River Didn't Get More Dangerous

• It Got Narrower A flood that rises higher on less water isn't a stronger flood. It's a smaller channel. That distinction is the entire finding of this study and it reframes what "Delhi flooded again" actually means.

The Data Point That Should Change How We Read Every Future Flood Headline

The single most important number here isn't the 68% channel narrowing - it's the comparison between 1978 and 2023. The 1978 flood carried more water than 2023, yet 2023 produced a higher water level in Delhi. If you only tracked rainfall or discharge, that result looks impossible.

It only makes sense once you know that a third of the river's floodplains - the lateral space it used to spread into during high flow - have been disconnected by embankments over the past century. The water in 2023 had less room to go sideways, so it went up instead.

That's not a freak weather event; it's a structural consequence that was decades in the making before a single drop of that specific monsoon fell.

Why "Natural Disaster" Is the Wrong Frame Entirely Here's the

angle a standard flood report misses: barrages like Tajewala, Wazirabad and Okhla were built for irrigation and water supply, not flood management and each one incrementally reduced what the study calls "formative discharge"

• the long-term flow a river needs to actively maintain its own channel width and flush sediment. Without that maintenance flow, sediment piles up (deposit area fell from 20 sq km to just 4 sq km since 1985) and the channel itself physically shrinks. Embankments built to protect expanding urban settlements then finished the job by permanently walling off the floodplain the river would otherwise have used as overflow space. Neither decision was made with flood risk as the primary variable - irrigation diversion and urban protection were the stated goals - but flood vulnerability was the accumulated byproduct. This is why treating each new flood as a discrete weather event misdiagnoses the problem. Rainfall varies year to year, but the river's structural capacity to absorb a rainfall spike has been declining in one direction for two centuries. Every fresh embankment or reclaimed floodplain parcel doesn't just add development - it permanently removes buffer capacity that the next major rainfall event will need and won't have. For the exam, the transferable insight is this: urban flood vulnerability in India is frequently classified as a climate or disaster-management issue when the underlying cause is often decades of unrelated infrastructure decisions - irrigation barrages, protective embankments, floodplain reclamation - that quietly convert a river's natural buffer into permanent risk. Reading floods as infrastructure history rather than isolated weather events is exactly the kind of causal reframing that distinguishes a strong GS3 answer from a descriptive one.