Not the Rain: Kerala's Floods Are a Land-Use Failure

The Lift Line

Kerala is not getting more rain. It is getting less, in harder bursts, on ground that has lost the ability to hold it. Only the bursts are beyond the state’s control.

Why This Editorial Matters for Your Exam

Disaster questions are usually answered with the NDMA structure, the Sendai Framework and a gesture at preparedness. This supplies a physical mechanism, the difference between rainfall volume and rainfall intensity mediated by infiltration rate, that converts a generic answer into a technical one, and it transfers directly to urban flooding in Chennai, Mumbai and Bengaluru, where the same argument holds with concrete substituted for paddy.

GS Paper 1: Important geophysical phenomena; changes in critical geographical features including water bodies; climatology; distribution of key natural resources.

GS Paper 3: Disaster and disaster management; conservation and environmental degradation; land reforms and land use.

Data and Institutions Vault

Prelims-grade facts:

• Kerala: about 580 km of coastline; 44 rivers of at least 15 km, of which 41 flow west to the Arabian Sea
• The east-flowing Kabini, Bhavani and Pambar are tributaries of the Cauvery
• Periyar is Kerala’s longest river (244 km); the Bharathappuzha has the largest basin (6,186 sq km)
• Kerala Conservation of Paddy Land and Wetland Act, 2008; its 2018 amendment created “unnotified land” and Section 27A conversions
• Gadgil report (2011): three graded zones across the Western Ghats; Kasturirangan report (2013): a smaller ESA of natural landscape
• The Western Ghats ESA has gone through seven draft notifications without a final one; Kerala’s draft extent is 9,993.7 sq km

Watch the trap: the 2018 Kerala floods led to a Post Disaster Needs Assessment and the Rebuild Kerala Development Programme, not a judicial commission of inquiry.

Background and Context

The author is K. A. Shaji, writing in Down To Earth on 5 August 2026.

The empirical starting point is stronger than a hedge allows. Kerala’s monsoon rainfall has not risen at all. On the Ministry of Earth Sciences’ Assessment of Climate Change over the Indian Region (2020), summer monsoon precipitation declined by about 6 per cent between 1951 and 2015, with notable decreases over the Western Ghats. What has changed is distribution: longer dry spells punctuated by shorter, more intense wet spells, with widespread extreme rain events over central India and parts of the Ghats up threefold since 1950, driven by surges of moisture from a warming Arabian Sea.

So the position is not that more rain is falling on a damaged landscape. It is that less rain, arriving in harder bursts, is producing worse floods, which makes the absorption question unavoidable rather than merely arguable.

Kerala flooded again in late July and early August 2026. Between 31 July and 2 August the state recorded 13.9 cm of rain against a normal of 4 cm, more than three times normal, with Vadasserikkara in Pathanamthitta recording 361 mm in 24 hours. As of 5 August the toll stood at 28 dead and four missing.

The 2018 floods remain the reference point: the most severe the state had recorded in nearly a century, with over 400 deaths. They did not produce a judicial commission of inquiry, and describing the aftermath that way is a common error. What they produced was a Post Disaster Needs Assessment in October 2018, the Rebuild Kerala Development Programme that followed from it, and an amicus curiae report to the Kerala High Court in 2019 by senior advocate Jacob P. Alex, which found that none of Kerala’s 79 dams had been operated for flood moderation, that reservoir storage was held high into the first week of August, that no integrated operation on scientifically derived rule curves existed, and that flood warning systems were absent. It recommended a fuller inquiry headed by a former Supreme Court judge.

The State Government moved the High Court to reject that report. The charge in this editorial is that the findings were absorbed rhetorically and never administratively, and the state’s own response to the amicus report is the clearest evidence for it.

The Core Argument / Issue

Why intensity matters more than volume

This is the mechanism, and getting it right is the difference between a good answer and a general one.

The intuitive account is that rain arrives faster than soil can absorb it and the excess runs off the surface. That mechanism, infiltration-excess or Hortonian overland flow, is real, and it is what produces urban flooding on concrete. It is not what happens in the Western Ghats, and getting this right matters, because the correct mechanism is far more sensitive to land use than the intuitive one.

These are steep, deeply weathered, forested catchments with high infiltration capacities. Measurement across sites in the Sahyadri ranges has found true infiltration-excess overland flow confined to small degraded or grassy patches, with Horton’s model failing to explain streamflow behaviour in this high-rainfall terrain. Most water goes into the ground and travels downslope as subsurface stormflow, frequently through natural soil pipes, a tunnel-erosion phenomenon documented specifically in Kerala’s lateritic highlands and implicated in slope failure and subsidence.

The governing process is therefore saturation, not exclusion. Weeks of antecedent monsoon rain fill the soil mantle; pore-water pressure builds at the soil-bedrock interface; shear strength collapses; the slope lets go.

This inverts one intuition entirely. Dry soil has higher initial infiltration capacity, because strong matric suction draws water in, so a dry profile initially generates less runoff, not more. Antecedent wetness is not incidental to Kerala’s floods and landslides; it is the precondition for them. Analysis of 225 rainfall-triggered landslides in Idukki between 2010 and 2018 concluded that antecedent rainfall, not any single day’s total, explains their occurrence, and cumulative rainfall over 10 to 40 days predicts failure better than daily intensity. World Weather Attribution’s own account of the Wayanad disaster records that the rain fell on soils already saturated by two months of monsoon.

Why this strengthens the argument rather than weakening it: infiltration-excess flow depends mainly on rainfall intensity, which policy cannot touch. Saturation and pore-pressure mechanisms depend on the depth, structure and drainage of the soil mantle and on what sits above it, all of which forest cover, slope cutting and quarrying directly change. Removing forest and cutting slopes shortens the drainage path, concentrates subsurface flow and lowers the threshold at which the mantle fails.

The redistribution of the same annual total into fewer and heavier events therefore produces substantially more damage with no increase in rainfall at all, because the soil mantle reaches saturation sooner and stays there longer. That disposes of the objection that rainfall has not increased, which is the objection most commentary treats as decisive.

What Kerala lost

Kerala’s flood-absorption infrastructure was overwhelmingly natural and incidental, which is why its removal attracted no infrastructure debate.

The recurring feature is that each of these performed a flood-management function without having been built as flood management, which meant no agency owned it, no budget line protected it, and its loss appeared in no infrastructure account.

Why topography compresses everything

Kerala sits between the Western Ghats and the Arabian Sea, a strip between 35 and 120 km wide along a coastline of about 580 km, covering some 38,863 sq km. It has 44 major rivers, meaning those at least 15 km long, of which 41 flow west into the Arabian Sea; the three that flow east, the Kabini, Bhavani and Pambar, are all tributaries of the Cauvery.

The dimensions are the point. The longest river, the Periyar, runs 244 km; the Bharathappuzha, at 209 km, is shorter but has the largest basin at 6,186 sq km, which is why it is often described incorrectly as the longest. More telling than either is the distribution: Kerala’s rivers average just 64 km, and 33 of the 44 run under 100 km from the Ghats to the sea.

The consequence is a very short catchment lag time. Where a large basin gives days of warning between upstream rainfall and downstream peak, Kerala gives hours. A 64 km river descending from the Ghats has almost no floodplain in which to spread and almost no residence time in which to attenuate a peak. That has two implications a strong answer should draw out.

Warning systems have limited headroom. Even a perfect forecast converts into a very short evacuation window when the lag is measured in hours.

Upstream retention is worth more than downstream response. Because there is so little time to react, capacity that holds water in the catchment is disproportionately valuable relative to capacity that responds to it on the plain. That is an argument for restoring absorption rather than for buying more rescue equipment, and it runs against the political economy of disaster spending, which favours the visible response over the invisible prevention.

Where the counter-argument is strong

The climate signal is real, and the attribution evidence is genuinely mixed. This is worth stating precisely, because both sides of the public argument overstate it.

As background, World Weather Attribution found the rainfall that triggered the Wayanad landslides of 30 July 2024, which killed more than 230 people, about 10 per cent more intense because of warming, with a further 4 per cent expected at 2 degrees. The study flagged high uncertainty, given the complex rainfall dynamics of a small mountainous region. Notably, the same study identified a 62 per cent reduction in forest cover in the affected region, and quarrying for building materials, as contributors to slope susceptibility, which makes it a citation for the land-use thesis rather than against it.

But a 2023 study in Environmental Research Letters found that anthropogenic climate change made the 2018 Kerala flood less likely, not more, attributing this to weakening South Asian monsoon low-pressure systems. Two major Kerala events, attributing in opposite directions. Anyone asserting that climate change simply causes Kerala’s floods is going beyond the evidence, and so is anyone denying the signal.

Dam operation is a genuine variable. Reservoirs serve irrigation and power, which argue for keeping them full, and flood moderation, which argues for keeping cushion. A reservoir filled early in the season has little room later, and rule curves that resolve this trade-off are a real management question rather than a distraction. Sequential releases from multiple reservoirs into the same short river system compound the problem further.

The statute was weakened rather than merely unenforced. This distinction matters, because “the law exists but is ignored” and “the law was amended to permit what it prohibited” call for different remedies. The 2018 amendment to the 2008 Act, Act 29 of 2018, created a new category of “unnotified land”, meaning land recorded as paddy or wetland in the Basic Tax Register but absent from the data bank that local committees were supposed to compile. It then inserted Section 27A, allowing the owner of such land to apply to the Revenue Divisional Officer for conversion to residential or commercial use on payment of a fee, bypassing the District Level Authorised Committee entirely. The data bank had been compiled incompletely and with errors; the amendment converted that administrative gap into a legal entitlement. The Supreme Court held in The Tahsildar v. Renjith George (May 2025) that these amendments cannot operate retrospectively, but the prospective effect stands.

Restriction is distributive, not technical. This is why the expert reports failed, and it deserves stating plainly. Kerala has very high population density, very high land prices and sustained remittance-funded housing demand. A restriction on hill-slope construction is a restriction on the principal asset-building route available to a large number of households. The Gadgil report (2011) proposed extensive ecologically sensitive area coverage in three graded zones and was politically rejected. The Kasturirangan report (2013) proposed a much smaller area confined to natural landscape, and became, as background, the basis for seven successive draft notifications: March 2014, 2015, 2017, 2018, 2022, 31 July 2024, and most recently 27 July 2026, nine days before this edition, covering 56,825.7 sq km across six states.

Not one has ever been finalised. Twelve years, seven drafts, no notification. Successive National Green Tribunal deadlines, March 2020 and December 2020, lapsed, and the expert committee examining state objections now has tenure running to July 2027.

Two details make the deadlock sharper than a simple Centre-versus-states story. Kerala’s ESA extent in the current draft, 9,993.7 sq km, is Kerala’s own figure, produced by the state’s Oommen V. Oommen committee, which cut the Kasturirangan proposal for Kerala by about a quarter; the Centre accepted it. And Kerala is now pressing for the exclusion of all human habitation and agricultural land from the ESA altogether.

The most telling fact in this background is the timing of the sixth draft: it issued on 31 July 2024, the day after the Wayanad landslides. Even a disaster that killed more than 230 people did not break the deadlock, and the seventh draft two years later is substantively unchanged in area. The stronger report was unacceptable and the weaker one has proved unimplementable, which suggests the obstacle is not the content of either.

How to Think About This (Analytical Frame)

When a service is performed incidentally by something that was never designed to perform it, its loss will be invisible in every account until the service fails. Paddy fields storing floodwater, mangroves absorbing storm surge, urban lakes buffering runoff, hedgerows fragmenting fuel loads, informal drainage channels conveying monsoon flow: each is doing work no agency has been assigned and no budget records. Because the service is unowned and unpriced, converting the land to a use that is owned and priced looks like an unambiguous gain, right up to the point of failure. The diagnostic question for any land-use change is not what the land will now produce but what it was quietly producing before.

The Diagram in Words

Picture a steep green slope with a sponge lying across it, and below the sponge a shallow bowl, and below the bowl a narrow channel running quickly to the sea. Rain falls; the sponge takes most of it; what escapes collects in the bowl and drains away slowly; the channel carries a manageable flow. Now, over thirty years, cut pieces out of the sponge for a quarry and a road and some houses. Fill the bowl and build on it, because it is flat land and flat land in Kerala is valuable. Narrow the channel where it passes a town. The rain has not changed much, but it now arrives in shorter, harder bursts. What is left of the sponge saturates in an hour, there is no bowl at all, and the same water that used to take three days to reach the sea arrives in one afternoon, through a channel half its old width, in a town built on the place the water used to sit.

Way Forward

1. Treat absorptive capacity as infrastructure with a budget line, so that paddy land, wetland and drainage restoration compete for funds on the same basis as embankments and rescue equipment.
2. Enforce the Kerala Conservation of Paddy Land and Wetland Act, 2008 through restoration, requiring reversal of unlawful reclamation rather than levying a penalty and permitting the fill to remain.
3. Zone by hazard, not by category, using slope, soil and flood-history mapping to restrict construction on steep gradients and in floodplains, with the mapping published so that it constrains permission at the point of grant.
4. Restore encroached drainage and canal networks, which is the cheapest single intervention available and the one most often deferred because it produces no new asset.
5. Revisit reservoir rule curves to preserve flood cushion through the monsoon, with coordinated operation across reservoirs feeding a common river, and publish the operating rules so that releases are predictable.
6. Replace the failed uniform ecologically-sensitive-area designation with graded, village-level regulation negotiated through local bodies, on the reasoning that a rule which is locally agreed and partially strict will outperform a rule that is scientifically ideal and never enforced.
7. Apply the framework to Indian urban flooding generally, since Chennai, Mumbai and Bengaluru present the same structure with concrete substituted for paddy and stormwater drains for backwaters.

PYQ Linkage and Practice

UPSC has tested Kerala’s floods, Western Ghats ecology, the Gadgil and Kasturirangan reports, wetland conservation and urban flooding across GS1 and GS3. The infiltration-rate mechanism and the incidental-service frame are what allow those to be written as one argument rather than as separate recollections.

Practice question: “The same annual rainfall delivered in fewer and heavier events, onto a landscape that has lost its absorptive capacity, produces flooding without any increase in rainfall at all.” Examine this claim with reference to Kerala, and assess why successive expert recommendations on Western Ghats land use have not been implemented. (250 words, 15 marks)

Interview angle: The Gadgil report was scientifically stronger and politically impossible; the Kasturirangan report was politically survivable and ecologically weaker, and neither has been implemented. If you were asked to produce a third report, what would you do differently to avoid the fate of both?

Sources: Down To Earth, Ministry of Environment, Forest and Climate Change, India Meteorological Department

Cite this page: Not the Rain: Kerala's Floods Are a Land-Use Failure — Ujiyari.com | Free UPSC & State PCS Editorial Analysis