CBAM May Hurt India’s Steel Exports Less Than Feared. Here’s What Mills Can Do
India’s steelmakers face growing pressure from the EU’s CBAM, but cleaner production could help reduce costs and protect exports. A new analysis outlines how mills can adapt by directing lower-emission steel to Europe, improving emissions reporting and investing in cleaner capacity.
Steel bars stacked at a manufacturing facility as Indian steelmakers adapt to the EU’s Carbon Border Adjustment Mechanism (CBAM). Representational image. Image credit: Willians Huerta/Pexels
For India’s steelmakers, the European Union’s Carbon Border Adjustment Mechanism (CBAM) is no longer a distant policy concern. The system entered its definitive phase in January 2026, putting the carbon intensity of exported steel firmly on the industry’s balance sheet. The focus is now shifting from the potential impact of CBAM to how Indian steelmakers can adapt to it.
A new analysis by climate think tank Sandbag argues that the impact could be substantially smaller than feared if Indian producers make strategic changes to where and how they produce steel. Its modelling suggests that, under an “expected” scenario, total CBAM costs for India’s steel exports could fall from 762 million Euro to around 79 million Euro as exporters optimise low-emission production. Under a more ambitious scenario involving new low-emission capacity, the modelling puts the sector at a potential 44 million Euro net benefit.
The Impact Is Already Showing Up in Trade
There are early signs of pressure on India’s steel exports to Europe.
India’s iron and steel exports to the EU fell 13% in the four months through April 2026 following the rollout of CBAM, according to Department of Commerce data cited by ICRIER. A June 2026 ICRIER study estimated that India’s steel exports to the EU could eventually fall by 24% because of CBAM.
However, the decline cannot be attributed to CBAM alone. Steel exports are also influenced by global prices, demand, freight costs and trade measures. The significance for Indian mills is therefore not just the possibility of selling fewer tonnes. Higher carbon-related costs could also squeeze margins or make Indian steel less competitive against lower-emission producers.
This is where the Sandbag analysis offers a different perspective. Its modelling suggests that Indian steelmakers can reduce their exposure by changing what they produce, where they sell it and how quickly they lower emissions.
First: Stop Treating Indian Steel as One Block
India’s steel industry is not uniformly high-emission. Different plants and production routes have different carbon intensities. That matters because CBAM does not simply impose the same burden on every tonne of steel. The emissions embedded in production determine the exposure.
Steel sheet being processed in an industrial manufacturing facility, illustrating the production processes behind India’s steel industry. Representational image. Image credit: jarmoluk/Pixabay
Sandbag identifies Indian steel plants whose emission intensity is already low enough to potentially benefit from CBAM. It specifically points to ArcelorMittal’s NS Hazira plant as an example of existing relatively low-carbon capacity that could serve European demand. For the industry, the first step is therefore to know where it stands.
Rather than approaching CBAM as a sector-wide penalty, steelmakers need to identify which plants, production lines and products have the lowest emissions and where those tonnes can earn the greatest value.
Send The Cleaner Tonnes to Europe
Immediate recommendation is strategic rather than technological: allocate lower-emission steel to the European market.
Report finds that Indian companies could reduce their CBAM liability by strategically reallocating low-emission output towards Europe. The same producers could also benefit from higher European steel prices, potentially offsetting the cost of CBAM certificates. That changes the way exporters need to think about their production portfolio.
If a company has several plants with different emissions profiles, sending the same type of steel to every market may no longer make commercial sense. Europe could increasingly become the destination for the cleaner end of the portfolio, while more carbon-intensive production faces greater pressure elsewhere. In effect, carbon intensity becomes another variable in export planning, alongside price, freight and demand.
Use Existing Low-Carbon Capacity Before Building from Scratch
India does not necessarily have to wait for a completely new generation of green-steel plants. Some relatively low-emission capacity already exists. Sandbag argues that this infrastructure can be leveraged to maintain Indian steel’s competitiveness in Europe, provided companies make the necessary operational adjustments.
That means the transition does not begin and end with new technology. Steelmakers can first examine how existing plants can be operated more efficiently, which products can be routed towards Europe and whether current capacity is ready for the reporting and verification requirements associated with CBAM.
For an industry under pressure to keep capital expenditure and production costs under control, that could be a more immediate route to reducing exposure.
Investment in New Low-Emission Capacity
The longer-term strategy is harder — and more expensive. Sandbag’s “ambitious” scenario assumes the deployment of new low-emission production capacity. Its modelling suggests that this could change the economics of CBAM enough to turn a net cost into a projected 44 million Euro net benefit.
This is where CBAM intersects with India’s larger steel decarbonisation challenge.
The industry is expected to expand significantly to meet domestic demand, but much of India’s steelmaking remains tied to coal-intensive production. Moving towards lower-emission technologies therefore requires substantial investment as well as changes in energy sources, raw materials and production processes.
For exporters targeting Europe, however, those investments could increasingly be viewed not simply as compliance costs but as a way of protecting market access.
Get The Carbon Numbers Right
There is another piece of groundwork that cannot be ignored: emissions measurement and reporting. CBAM effectively puts a carbon price on the emissions embedded in imported products. That means steelmakers need to know, and be able to demonstrate, the emissions associated with their production.
The Sandbag report itself flags “readiness for CBAM reporting” among the factors that will determine which existing low-emission facilities are best positioned to benefit. For Indian mills, reducing emissions without being able to credibly quantify them could limit the commercial advantage.
The carbon footprint of steel is therefore becoming part of the product information that exporters need to manage.
The Industry Cannot Rely on Cleaner Steel Alone
There is a catch. The Sandbag modelling does not suggest that CBAM will simply disappear as a problem for Indian steel. Its projected outcomes depend on producers responding appropriately — by optimising production, reallocating cleaner output and investing in low-emission capacity.
That makes the transition uneven. Companies with access to relatively low-emission plants, capital for technological upgrades and the ability to accurately measure emissions could be better placed than smaller or more carbon-intensive producers.
This could create a new fault line within India’s steel industry: not simply between large and small producers, but between those that can demonstrate lower-carbon production and those that cannot.
Europe May Become the Proving Ground
India has strongly opposed CBAM, calling it discriminatory and protectionist and questioning its compatibility with international trade principles. But whatever the broader policy dispute, Indian steelmakers now have to deal with the market reality. Europe is putting a price on carbon-intensive imports. At the same time, India is looking to strengthen its position in international steel markets.
That makes the European market a useful proving ground for India’s transition to cleaner steel. The Sandbag analysis suggests that Indian steelmakers have several levers to pull: identify their lowest-emission capacity, direct those tonnes towards Europe, improve operational efficiency, strengthen emissions reporting and invest in new low-emission production where commercially viable.
The opportunity is not guaranteed. Nor will every steelmaker benefit equally. But the takeaway is that CBAM does not have to be treated simply as a bill Indian-steelmakers must pay. For companies that can clean up production and put the right steel in the right market, it could become a competitive test they are capable of passing.
EP Staff is the editorial team at EdPublica, an independent media organisation focused on science, education, environment and public policy. The team produces evidence-based news, features, explainers and analysis on issues that shape society and everyday life.
India’s electricity is shifting fast to non-fossil sources — but oil, gas and LPG imports are set to rise, not fall, under the IEA’s projections to 2035. Illustration: EdPublica
IEA’sWorld Energy Outlook 2026 has India’s electricity shifting fast towards non-fossil sources. For oil, gas and cooking fuel, the same projections show more imports, not fewer.
India has said since 2021 that it wants to be energy independent by 2047. The International Energy Agency’s World Energy Outlook 2026, published on 19 November, does not say India is on course for that. Its projections stop at 2035, and for oil and gas they point the other way.
India already meets about 40 per cent of its energy needs through imports, including 90 per cent of its crude oil and 50 per cent of its natural gas, according to the Outlook. Over the next decade its energy demand is projected to grow by 2.5 per cent a year in the agency’s Stated Policies Scenario (STEPS), roughly three times the global rate. STEPS counts existing policies plus measures that governments have formally proposed. “In all our scenarios, India accounts for more growth in energy demand over the next decade than any other country or region in the world,” the IEA said.
The picture is uneven. Electricity is moving quickly away from imported fuel. Oil and LPG are not. And the one scenario in the Outlook where India pushes harder on imports shows gains that are real but partial.
Is India Really Heading Toward Energy Independence?
India’s electricity generation rises from 2,114 terawatt-hours (TWh) in 2025 to 3,585 TWh in 2035 under STEPS, an increase of about 70 per cent. The non-fossil share, meaning renewables and nuclear, goes from 28 per cent to 51 per cent. “As non-fossil capacity continues to expand, India generates well over half of its electricity from non-fossil sources by 2035 in the STEPS,” the report says. Solar PV accounted for half of India’s capacity additions over the past decade.
Illustration: EdPublica
A quick calculation with those rounded shares adds a wrinkle. Non-fossil output roughly triples, from about 590 TWh to about 1,830 TWh. But fossil generation, 72 per cent of the 2025 total and 49 per cent of the 2035 total, still rises from about 1,520 TWh to about 1,760 TWh, an increase of around 15 per cent. The share of coal and gas falls while the amount of electricity made from them goes up. (These are our own calculations from rounded IEA shares, not figures the IEA publishes.)
Globally, Ben Backwell, chair of the Global Renewables Alliance, pointed to the IEA’s expectation that “Renewables are now expected to grow from 33% of electricity generation today to around 50%, and potentially even higher, by 2035.” India’s non-fossil share, which includes nuclear, goes from 28 to 51 per cent over the same years. The measures differ, but the direction and rough scale are similar.
Dave Jones, chief analyst at Ember, reads the Hormuz crisis as an accelerant. “Hormuz will speed up electrification and renewables, likely faster than many people think,” he said. The IEA’s own summary lists a push for faster electrification among the common responses it sees from countries hit by the crisis.
Oil and LPG: the other direction
Domestic crude production is projected to fall by nearly 20 per cent by 2035 under STEPS. Over the same period refining capacity grows by about 1 million barrels a day from nearly 6 million in 2025, with another 840,000 barrels a day already under construction, 660,000 of it expansions at existing plants and 180,000 greenfield. India ranks fourth in the world by refining capacity and third by refinery runs, and it remains a net exporter of petroleum products. But a larger refining industry fed by a shrinking home supply means more crude has to come from abroad.
The Outlook expects crude imports from the Middle East to stay broadly flat while total imports rise, with most of the extra supply coming from outside the Gulf. West Asia supplied roughly half of India’s refinery intake in 2025.
Cooking gas is the sharper problem. India produced about 450,000 barrels a day of LPG in 2025, around 40 per cent of what it uses. Under STEPS, domestic supply grows by less than 100,000 barrels a day by 2035 while demand grows by nearly 400,000, which could take LPG imports to nearly 1 million barrels a day.
This year showed what that exposure costs. The IEA estimates that the LPG used in half of India’s residential cooking cylinders passes through the Strait of Hormuz. LPG imports fell by more than 50 per cent in the first two months of the conflict, and the government maximised domestic production and put household supplies ahead of other users. “The Strait of Hormuz remains critical to India’s energy security,” the report said.
IEA Executive Director Fatih Birol put the wider lesson plainly. “I don’t see any credible pathway for the energy world to return to the way it was before,” he said. He also warned that “In pursuit of greater energy security, countries must take care not to replace one dependency with another.”
What a harder push would buy
The Outlook includes a scenario in which India goes further to cut its reliance on imported fuel: faster adoption of electric vehicles, more industrial electrification, better efficiency, more renewables and more domestic coal. By 2035 electric car adoption is 30 per cent faster than under STEPS. India’s oil requirement is 580,000 barrels a day lower and its natural gas use is 8 per cent lower than in STEPS.
For scale, Jan Rosenow, Professor of Energy and Climate Policy at the University of Oxford, noted that “China’s electric vehicle fleet, for example, is avoiding around 1.5 million barrels of oil a day.” The baselines differ, since that is China’s fleet today and India’s figure is an extra saving in 2035, but it gives a sense of the size of the prize. Rosenow’s broader point was that “countries which had electrified more of their energy use were better protected from this year’s supply shock.”
Coal does much of the work on the import side. Greater domestic coal production could bring steam coal imports down from nearly 150 million tonnes of coal equivalent in 2025 to around 30 million by 2035, and cut overall coal import dependence from about 30 per cent to 15 per cent. But rising steel output pushes coking coal demand up by 55 per cent, and India would still depend on imports for much of that extra requirement.
That sits awkwardly with climate goals, and campaigners have said so. David Tong of Oil Change International said the IEA “has yet again reiterated that there is no room for new oil, gas, and coal beyond existing fields to limit global warming to 1.5ºC.” Maria Pastukhova of E3G made a different point about the global picture: “over three quarters of the additional reduction in oil and gas imports that could result from policy responses to the Hormuz crisis comes from energy efficiency, electrification and renewables.” That is a worldwide figure, not an Indian one.
The cooling load
Over the next decade India is expected to add more than 12,000 cars to its roads every day, see a 33 per cent increase in built-up space and install another 200 million air conditioners. Globally, the IEA finds that energy needed to keep buildings cool is now overtaking the energy needed to heat them for the first time.
Pastukhova called this one of the report’s most alarming and least noticed findings: “by 2035, cooling will account for 30% of global peak electricity demand.” She added that “In emerging and developing economies outside China, 85% of households are exposed to high temperatures, yet fewer than one in five of these households have air conditioning.” India is in that group. Her warning on the fix was blunt: “Falling clean technology costs alone will not prevent it”.
Air conditioning is what drives peak demand, and the Outlook says coal will keep an important role in meeting India’s electricity demand. Anyone judging how independent India’s power system can become by 2047 has to count that load.
Reading 2047 against the data
The 2047 goal was announced by the Prime Minister in his 2021 Independence Day speech, with the stated aim of bolstering energy security by reducing imports. The IEA’s own policies database lists it as announced. The Outlook does not test it directly, because its projections end in 2035.
The case that it can be done comes from elsewhere. A 2023 modelling study from Lawrence Berkeley National Laboratory set out a pathway to near-complete energy independence by 2047 across power, transport and industry, mainly by cutting imported oil used in road transport and imported coal used in industry and power.
Set against the IEA’s numbers, the power half of that pathway is visibly under way. The oil half is not yet there. Under STEPS, crude output falls, refining grows and LPG imports climb. The stronger scenario narrows the gap but does not close it, and it leans on coal to do so.
Birol said that “energy security and energy transitions have never been so closely aligned.” For India, the next decade will show whether that alignment reaches the fuels it buys from abroad, or stays mostly in the power grid.
Challenge AI on Climate Science, and It May Abandon the Facts
A UNESCO–UCL study finds that AI models can abandon established climate-science evidence when users challenge their answers. The research also identifies weaknesses in climate-policy reasoning and significant differences in performance across languages, raising questions about the reliability of AI-generated climate information.
AI models' accuracy fell by up to 35.1 percentage points when users persistently challenged their answers, regardless of the professional identity they claimed, according to the UNESCO–UCL study.Image credit: UNESCO
Artificial intelligence models can explain established climate science with impressive accuracy. But when users challenge their answers, present new evidence or ask them to assess complex climate policies, their reliability declines sharply. A new UNESCO–University College London (UCL) study warns that the fluency of AI-generated explanations can conceal serious weaknesses in how these systems evaluate climate information.
The report was presented at COP31 in Antalya on November 16. Researchers tested up to 33 leading models against nine climate and environmental datasets, across 13 languages. They also assessed 656 climate-policy questions against En-ROADS, a climate–energy simulation tool.
The findings reveal a distinction between recalling established scientific knowledge and reliably assessing evidence. On questions about established climate science, the best-performing models scored between 88% and 96%. But when asked whether new evidence supported or contradicted accepted scientific claims, even the strongest systems achieved only 71–77% accuracy.
When AI agrees instead of correcting misinformation
One of the study’s most concerning findings concerns sycophancy — the tendency of AI systems to agree with users even when they are wrong.
Researchers tested 15 large language models using climate claims, including the assertion that carbon dioxide is not the main driver of climate change. Although the models initially answered neutral questions accurately, roughly eight in ten abandoned a correct answer at the first sign of user disagreement.
The problem intensified when users claimed professional expertise or introduced fabricated evidence. A false assertion from someone identifying themselves as a climate scientist reduced average accuracy by 12.9 percentage points. When users sustained their challenges over several exchanges, accuracy fell by between 30.9 and 35.1 percentage points. All 15 models were affected.
The researchers attribute this behaviour partly to training incentives that reward agreeable and validating responses. Leona Verdadero, UNESCO’s lead author, said AI could make complex climate science more accessible, but warned that its tendency to agree with users undermines its reliability as a source of established facts.
Policy analysis exposes another weakness
The report found that strong performance on individual climate questions does not translate into reliable policy analysis.
Using En-ROADS, researchers examined whether models could identify the effects of individual policy measures and understand what happens when two measures are implemented together. The best-performing model achieved 94.5% accuracy when identifying the direction of an individual policy’s effect on projected warming.
But when asked whether the combined effect of two policies would be greater than, smaller than or approximately equal to the sum of their individual effects, average accuracy fell to 33.8%. With three possible answers, random guessing would produce 33.3%.
This matters because climate policies rarely operate independently. Measures can overlap, reinforce one another or reduce each other’s effects through changes in energy demand, fuel prices and emissions. Increasing model size did not meaningfully resolve this weakness.
UCL senior author Dr Maria Perez-Ortiz cautioned that AI systems’ ability to explain scientific knowledge could create a false sense of their wider capabilities, particularly when users move from asking what is known about climate change to asking what should be done.
The problem extends to how people interpret AI responses
Other research suggests that the issue is not limited to the accuracy of a model’s answers. A September 2026 study published in PLOS Climate, involving 790 US adults, examined how people used ChatGPT to evaluate four false climate claims.
The researchers found that positive language and semantic richness in ChatGPT’s responses were associated with higher perceived credibility of the responses. Participants’ existing acceptance of climate science, attention to climate change and familiarity with ChatGPT also influenced their credibility judgments.
The findings do not establish that persuasive language makes people believe misinformation. They do, however, highlight how the presentation of an AI response can shape users’ perceptions of its credibility — an important consideration when chatbots are used to assess misleading climate claims.
A separate study, published in Public Understanding of Science identified another form of bias. Comparing GPT-family model responses with expert assessments based on the IPCC’s 2023 Synthesis Report, the researchers found that the models systematically overestimated climate change impacts. The tendency was more pronounced when models were prompted to respond in the role of climate scientists.
Language gaps and possible safeguards
UNESCO’s report also found that climate information generated by AI is not equally reliable across languages. Average accuracy was 88.6% in English, compared with 83.6% in Hindi and Bengali, 81.4% in Urdu and 78.3% in Swahili. Across five models, fewer than half of the questions in two climate datasets received the same answer in all 13 tested languages.
A single false assertion from a “climate scientist” lowered accuracy significantly in 13 of the 15 models; sustained pushback did so in all 15 under every identity.Image credit:UNESCO
The researchers link these differences partly to the quality and quantity of digital text available for model training in different languages. The findings raise concerns for communities that depend on languages in which AI systems have been less extensively trained.
Research also points to ways of improving AI-assisted climate fact-checking. A 2025 study in npj Climate Action introduced Climinator, a system that combines large language models with authoritative sources, including IPCC and World Meteorological Organization material, to assess climate claims. At one level of detailed claim classification, it achieved an F1 score of 72.7%, compared with 58.5% for GPT-4o. Another study, published in Climatic Change in September 2026, found that structured prompts and retrieval of authoritative external sources improved some aspects of AI-generated climate debunking.
These approaches do not eliminate the need for verification, but suggest that models can be made more useful when their responses are grounded in scientific evidence rather than generated from their internal knowledge alone.
UNESCO recommends stronger disclosure of model performance across languages and under user pressure, alongside independent evaluation before deployment in public-interest settings. It also calls on educators, journalists and civil society groups to verify AI-generated claims against reliable sources.
The central warning is not that AI has no place in climate communication. It is that the ability to explain climate science convincingly should not be mistaken for the ability to evaluate evidence, resist misinformation or independently guide climate policy.
India Records 436 Extreme Weather Events in 30 Years, Loss of 170 billion USD
India recorded 436 extreme climate events between 1996 and 2025, averaging one event every 25 days. These disasters claimed over 78,000 lives, affected 1.2 billion people and caused more than 170 billion dollars in economic losses.
India has faced a relentless succession of cyclones, floods and deadly heatwaves over the past three decades. Between 1996 and 2025, the country recorded 436 extreme weather events, which together claimed more than 78,000 lives, affected over 1.2 billion people and caused economic losses exceeding 170 billion dollars.Averaged across the 30-year period, that amounts to one extreme weather event every 25 days. The scale of this recurring toll places India 10th among the world’s most climate-affected countries, according to the Climate Risk Index 2027, published by Germanwatch, a Germany-based environmental and development organisation.
The figures capture the cumulative toll of recurring disasters across the country, from destructive cyclones and monsoon floods to prolonged heatwaves. Unlike countries whose long-term rankings were driven largely by one exceptionally devastating catastrophe, India’s position reflects repeated exposure to extreme weather events over several decades. The Philippines and Nicaragua are also included in this group.
Countries such as Dominica, Myanmar and Honduras, meanwhile, rank highly because of exceptionally destructive disasters that accounted for a substantial share of their recorded losses. For India, the pattern is different. Its 436 recorded events represent a succession of climate-related emergencies across different regions and seasons.
India ranks 10th among the countries most affected by extreme weather events between 1996 and 2025.The Germanwatch Climate Risk Index 2027 breaks down the ranking across six indicators covering fatalities, people affected and economic losses.
Three Decades of Cyclones, Floods and Extreme Heat
India’s 10th-place position in the long-term ranking is largely the result of repeated disasters over three decades. Germanwatch describes India as a country facing “continuous threats”, where one extreme weather event follows another and the losses accumulate over time. The events range from the 1998 Gujarat and 1999 Odisha cyclones to Cyclone Hudhud in 2014, Cyclone Amphan in 2020, the Uttarakhand floods of 2013 and severe floods across 13 states in 2019.
The financial toll of individual disasters shows how sharply these losses can rise. Cyclone Amphan alone caused more than 16 billion dollars in damage in India, adjusted for inflation. It was the most destructive cyclone recorded in the country’s 30-year assessment period.
The report also identifies four years — 1998, 2002, 2003 and 2015 — when recurring and unusually intense heatwaves, with temperatures around 50°C, claimed many lives. These episodes sit alongside the deaths and displacement caused by floods and cyclones in the country’s long-term disaster record.
India’s geographical diversity exposes it to several hazards. Its long coastline faces tropical cyclones and storm surges, the Himalayan region experiences intense rainfall, floods and landslides, while large parts of the country depend on the monsoon for agriculture and water availability. Extreme heat affects both rural and urban populations.
The Germanwatch report attributes countries’ high climate risk rankings to heightened exposure to hazards or increased vulnerability. In India’s case, recurring disasters mean that the effects of one event can overlap with the economic and social consequences of earlier ones. A flood can destroy homes, roads, crops and public infrastructure. When another extreme weather event strikes before affected communities have recovered, households and local economies face additional financial pressure. Farmers may lose successive harvests, while damage to roads, electricity networks and water systems can disrupt essential services.
Urban development presents another challenge. Rapid construction and the expansion of built-up areas can reduce the land available to absorb rainwater, increasing the risk of urban flooding. The concentration of buildings, roads and other infrastructure can also intensify heat exposure.
A 2025 World Bank report, Towards Resilient and Prosperous Cities in India, examined climate risks in 24 Indian cities, with particular attention to Chennai, Indore, New Delhi, Lucknow, Surat and Thiruvananthapuram. It found that urban heat island effects can raise temperatures in city centres by 3–4°C compared with surrounding areas. The report also warned that expanding built-up areas are reducing cities’ capacity to absorb stormwater.
The scale of future urban development makes these risks particularly significant. India’s urban population is projected to increase from 480 million in 2020 to 951 million by 2050. More than half of the infrastructure required for that future urban population is yet to be built. Decisions about housing, transport, drainage and public spaces will therefore influence how well cities withstand future climate shocks.
Economic vulnerability compounds these geographical pressures. The World Bank has noted that more than 80% of India’s population lives in districts at risk of climate-induced disasters. Communities dependent on agriculture, fishing and outdoor work can face significant livelihood disruptions when extreme weather affects their primary sources of income.
Climate Resilience Must Become Part of Development Planning
The scale of India’s recorded losses makes climate resilience an important component of economic and social planning. It involves preparing communities and institutions for extreme weather, reducing their exposure to hazards and ensuring that essential services can continue during and after disasters.
Early warning systems are an important part of this effort. Forecasts need to reach people in time and communicate the likely impact of an event, allowing local authorities and communities to act before a disaster occurs. Evacuation arrangements, accessible shelters, trained emergency personnel and reliable communication networks can reduce casualties and disruption.
India’s experience with heat preparedness in Ahmedabad offers an example. Following a deadly heatwave in 2010 that caused 1,334 additional deaths in May alone, the city introduced South Asia’s first comprehensive Heat Action Plan in 2013. The programme combines temperature-based warnings with public communication, healthcare preparedness and measures to protect people exposed to extreme heat. A World Bank case study credits the plan with helping avert around 1,100 deaths annually in the years immediately following its introduction.
Flood resilience requires a different set of interventions. Cities need drainage systems designed around local rainfall patterns, better stormwater management and restrictions on construction in vulnerable floodplains. Wetlands, lakes and open spaces can also help retain excess rainwater. The World Bank’s 2025 urban resilience report recommends combining engineered infrastructure, such as drains and flood embankments, with nature-based measures that improve water absorption and storage.
Infrastructure outside cities also needs attention. Roads, bridges, electricity networks, hospitals and water supply systems must be designed to withstand the hazards prevalent in their locations. In agricultural regions, better access to weather information, water conservation, resilient farming practices and financial protection against crop losses can help households recover from extreme weather.
Community preparedness is equally important. Local authorities need to identify vulnerable households, understand the risks faced by different communities and ensure that disaster warnings translate into practical action. Outdoor workers, older people, children, people with disabilities and low-income households may require specific protection during heatwaves, floods and cyclones.
Climate resilience also requires sustained financial investment. The World Bank’s 2025 assessment estimates that India will need more than $2.4 trillion by 2050 to meet its requirements for resilient, low-carbon urban infrastructure and services. It estimates that timely adaptation measures could avert annual urban flood-related losses of up to $5 billion by 2030 and $30 billion by 2070. Investments in adaptation could also prevent more than 130,000 deaths from extreme heat by 2050.
These estimates underline the economic value of preparing for climate risks before they cause widespread damage. Spending on resilient infrastructure, early warning systems and local adaptation can reduce the costs of repeated reconstruction and emergency relief.
India’s 436 recorded extreme weather events over three decades provide a measure of the pressure that recurring disasters place on its people and economy. Building resilience requires planning for the hazards that different regions face, protecting vulnerable communities and ensuring that development does not create new sources of risk. The decisions taken on infrastructure, land use, public health and livelihoods today will influence how much damage future extreme weather events cause.