extreme weather in myanmar essay

• Geopolitics
• Environment

Myanmar at risk from extreme climate

According to the Global Climate Risk Index 2020, Myanmar has had the highest weather-related losses in the past two decades, alongside Puerto Rico and Haiti. It is said that Myanmar is also one of the most vulnerable countries at risk of climate crisis. The consequences of climate change  can be seen around the world, with natural disasters and rising sea levels headlining global news. In Myanmar, severe flooding in recent years and 2008’s disastrous Nargis cyclonic storm have affected the lives of millions of locals and caused over 100,000 deaths. The deadly tropical cyclone was deemed as the worst natural disaster recorded in Myanmar’s history. 

The dry zone of Myanmar lies in the central portion of the country, home to nearly a third of Myanmar’s total population. According to media reports, temperatures there are projected to rise by up to three degrees Celsius (3° C) after 2040. In the Irrawaddy delta, in the south – the mid-level projection for sea-level rise is up to 40 centimetres (cm) by 2050. 

Shashank Mishra from Myanmar Climate Change Alliance, which is a body that straddles the United Nations (UN), the government, and civil society – told the media that “the total monsoon period has already decreased from 144 days per year in 1998 to 125 days.” He also added that the number of extremely hot days is projected to increase from one day a month to between four and 17 by 2041. This will cause serious health problems to the locals, damage ecosystems, crops and infrastructure.

Historian and author of The Hidden History of Burma, Thant Myint-U pointed out that the impact of climate change on Myanmar could be catastrophic in terms of rising sea levels, extreme heat and extreme weather events. 

Myanmar’s unstable weather has resulted in loss of production and rising indebtedness for local farmers.

Agriculture under threat

Based on a 2017 report titled, ‘Assessing Climate Risk in Myanmar’ by the World Wide Fund for Nature (WWF), agriculture is the main economic activity in Myanmar and the largest employer of the labour force. The rise in temperature has severely affected the agriculture sector in Myanmar.  The WWF report stated that crop productivity could decline, as some crops are especially vulnerable to temperature increases. Drought incidence would likely increase as well, affecting agriculture, livestock, wildlife and communities alike that struggle with declining water availability resulting from increased evaporation. 

The Irrawaddy delta is known for its fertile area for rice growing and was once called the “rice bowl” of the British Empire. It was reported that production was so good that Myanmar could feed its people a high amount of rice with enough left over to be sold. Unfortunately, the cyclone in 2008 flooded paddy fields with sea water, damaged irrigation systems and destroyed seed supplies. The UN Food and Agriculture Organization (FAO) estimated that cyclone Nargis impacted 65 percent of the country's paddy fields. This is one of the many examples of how climate crisis could affect food security and the livelihoods of local farmers  and the public in general.

Other issues faced by farmers as a result of climate change include an increase in weeds, diseases and insect pests that may find the new climate more hospitable. 

Stopping climate change

Based on a paper published in the Munich Personal RePEc Archive in 2019 titled, ‘Economic Assessment of Climate Adaptation Options in Myanmar Rice-Based Farming System’, the central dry zone of Myanmar suffers a significant climate variability due to droughts and floods resulting in crop destruction and increased vulnerability of farm households. 

From 2015 to 2019, a project called ‘Addressing Climate Change Risks on Water Resources and Food Security in the Dry Zone of Myanmar’ initiated by the United Nations Development Programme (UNDP) was carried out to provide farmers with the resources, knowledge and tools needed to support good harvests, despite changing weather patterns. The project aimed to reduce the increasing impacts of climate change on agricultural and livestock production cycles in the dry zone of Myanmar.

In 2019, the government of Myanmar launched two new policies that will guide Myanmar’s environmental management and climate change strategy. The National Environmental Policy and the Myanmar Climate Change Policy were introduced as there’s a growing awareness that extreme weather and other climate change events could threaten and impact the country’s economic and social development. 

A number of initiatives have been proposed by the government and related organisations to help fight climate crisis in Myanmar. The Myanmar Climate Change Strategy 2018-2030 was also drafted by the country’s Ministry of Natural Resources and Environmental Conservation (MONREC) with assistance from other organisations. The long-term goal of this strategy is to achieve climate-resilience and a low-carbon growth pathway to support inclusive and sustainable development by 2030. 

Related articles:

Global strike for climate change

Protecting our food by protecting our farmers 

Tangshan And Xuzhou: China's Treatment Of Women

Myanmar's suu kyi: prisoner of generals, philippines ends china talks for scs exploration, ukraine war an ‘alarm for humanity’: china’s xi, china to tout its governance model at brics summit, wrist-worn trackers detect covid before symptoms.

• High contrast
• Children in Myanmar
• Where we work
• Our Representative
• Our partners
• Work with us
• Press Centre

Search UNICEF

Cyclone mocha: latest example of dire climate threat facing myanmar’s children, 3.4 million people live in the areas highly impacted by the cyclone.

• Available in:

More than 90 per cent of children in Myanmar face three or more overlapping climate and environment shocks, hazards or stresses, according to a new UNICEF regional report , ‘Over the Tipping Point’.

The report reveals that children in the East Asia and Pacific region face a greater exposure to multiple climate disasters than in any other region of the world. It calls for urgent investment in climate-smart social services and policies to protect children.

Ten days after Cyclone Mocha battered areas in Myanmar already hard-hit by years of conflict and deprivation, the impact of climate change on children and families is clear for all to see, UNICEF said. It is estimated that 3.4 million people live in the areas highly impacted by Cyclone Mocha.

With temperatures and sea levels rising and extreme weather such as typhoons, severe floods, landslides and droughts increasing globally, millions of children are at risk. Many children and their families face displacement and struggle to survive, with limited or no access to healthcare, education, and water and sanitation services.

In Myanmar, in addition to the impact of COVID-19 and conflict, many families have been forced to pull their children out of school to help with agricultural work, or because they cannot afford the cost of education due to the economic impact of climate change on their livelihoods.

According to the latest analysis, which is based on the Children’s Climate Risk Index (CCRI), in the East Asia and Pacific region over 210 million children are highly exposed to cyclones; 140 million children are highly exposed to water scarcity; 120 million children are highly exposed to coastal flooding; and 460 million children to air pollution. Ultimately, these effects exacerbate inequalities that children already face, pushing the poorest further into poverty. 

As families grapple with the aftermath of Cyclone Mocha - amid an ongoing conflict - it is important to remember girls are more likely to be affected by disasters than boys or men, UNICEF said. Also, children with disabilities are at greater risk of adverse outcomes following a natural disaster.

In addition to its ongoing humanitarian support to children affected by the ongoing conflict across Myanmar, UNICEF is working to provide lifesaving assistance to children and families affected by Cyclone Mocha in Rakhine, Chin and Kachin States and Sagaing and Magway Regions.

Related topics

More to explore, how you can support your child during bombing.

Tips for parents on physical safety and how to reduce stress

Delivering child protection support in central Myanmar

Danger in the Dry Zone

Healing children’s trauma with a simple box of crayons

Giving children a chance to be a child again

UNICEF relief supplies reach desperate families

Humanitarian assistance

• Get involved

Why Climate Change Matters for Myanmar’s Development, and What We Are Doing About It

September 22, 2019.

Op-ed by:  U Ohn Win, Union Minister, Ministry of Natural Resources and Environmental Conservation, and Mr. Peter Batchelor, UNDP Myanmar Resident Representative

Mangrove forest in Tanintharyi region Photo Credit: Singay Dorji

Myanmar faces many development challenges, but climate change presents the greatest challenge of all. And while the impacts of climate change are felt in many ways, it is the threat to the country’s future development that makes it so significant.

Myanmar’s geographic location and incredible physical diversity means climate change takes many forms – in the dry zone, temperatures are increasing and droughts becoming more prevalent, while the coastal zone remains at constant risk from intensifying cyclones. And extreme flooding in the current wet season has already seen over 190,000 people seek emergency shelter, with the damage to homes, schools and farms compounding the impact of last year’s floods, and those from the year before.

The possibility of more intense and more frequent climatic events is already impacting Myanmar. The country is already one of the most vulnerable in the world to such extreme weather events. With the memories of 2008’s catastrophic Cyclone Nargis still vivid, the development gains that have been made in recent times remain highly susceptible to such risks. Even without a single Nargis-scale event, the loss and damage caused by floods, landslides and droughts in recent years runs into the billions of US dollars, not to mention the deaths and turmoil for people’s lives. The need to prepare for, respond to, and recover from, these natural disasters costs time and resources that could otherwise be spent on more pressing development priorities.

There is no question that Myanmar must work with the international community to slow down and reverse global warming, while also building its resilience by adapting to the reality of a changing climate.

The Government of the Republic of the Union of Myanmar recognizes that a clean environment, with healthy and functioning ecosystems, is the foundation upon which the country’s social, cultural and economic development must be sustained. It has therefore committed to a national development framework that incorporates the notion of environmental sustainability for future generations by systematically embedding environmental and climate considerations into all future policies and projects. The Myanmar Sustainable Development Plan (2018-2030) has committed Myanmar to a climate-sensitive development pathway and is complemented by the new National Environmental Policy and Myanmar Climate Change Policy, which were both recently launched by the President. Both policies have benefitted from technical support from international development partners like UNDP and extensive public consultations across Myanmar.

Together, these new policies set a vision for Myanmar as a climate-resilient, low-carbon society that is sustainable, prosperous and inclusive, for the well-being of present and future generations. They are also the basis for Myanmar’s implementation of the Paris Agreement to help keep global temperature increases to 1.5 degrees Celsius above pre-industrial levels.

Myanmar’s ambitious approach to reducing greenhouse gas emissions includes reversing the decline of the country’s forests. For example, the Government has committed 500 million US dollars over 10 years for the Myanmar Rehabilitation and Reforestation Programme. Nature-based solutions, such as protecting coastal mangrove forests, can help mitigate climate change by storing huge amounts of carbon dioxide while also building natural barriers to reduce the impact of cyclones and storm surges on coastal communities.

The energy sector contributes two-thirds of greenhouse gas emissions in the Asia-Pacific region. Therefore, energy development in Myanmar must be climate smart and cannot ignore the sector’s changing economic outlook. The Government is committed to increasing the use of renewable energy while helping provide electricity to the millions of people who still lack reliable access. New forms of renewable energy – including solar and biomass – will contribute 9% of the country’s energy mix by 2030. The distribution of fuel-efficient cookstoves is being rapidly expanded to 5 million households – this will improve people’s health, while also helping avoid deforestation from people gathering firewood.

Myanmar’s private sector has a vital role to play in responding to climate change, but this also presents great opportunities as the economy expands. Disruptive green technology and innovation will help us accelerate towards a low carbon economy. The government will continue promoting green businesses and aligning incentives for the private sector to introduce low carbon technologies.

To empower younger generations with the knowledge, skills and attitudes to prosper in 21 st century Myanmar, climate change must inform the strengthening of the education system – in the curriculum, as well as by developing climate-resilient schools.

Such approaches also show that what is good for the climate is also good for our country’s development and the lives of our people.

Myanmar is ready to be part of the climate change solution at this month’s United Nations Climate Action Summit, to be held in New York on 23 September, 2019. The Government is continuously intensifying its efforts, but does need international support from partners like UNDP. Technological know-how is needed to support actions on the ground, such as in the form of affordable renewable energy technologies. Finance is needed to support investments in human resources and innovative solutions. And training and technical assistance is needed to strengthen the capacities of all players – governments, communities and the private sector.

The Climate Action Summit in New York will be a critical forum for global leaders to come together and present strong new actions to reverse climate change. It is also an opportunity to recognize the valuable efforts of developing countries like Myanmar, and to redouble support for countries pursuing sustainable, low-carbon development pathways.

The world is now in a desperate race against climate change. Strong and urgent action, both internationally and locally, is critical to protecting Myanmar’s current and future development. We must all act together if we are to win this race and ensure a sustainable future for the world and for Myanmar and its people.

Related Content

Rakhine State

Publications

Myanmar at a crossroads: past trends of human well-being and a future outlook.

Myanmar’s current crisis is derailing the progress it had been making in terms of economic growth, reducing poverty, creating jobs and reaching Sustainable Develo...

Helping communities weather the socio-economic downturn: Building urban resilience

Following the twin crises of COVID-19 and the military takeover in Myanmar, Yangon’s inhabitants face unprecedented new pressures. The poverty rate in Yangon was ...

Household Vulnerability Survey 2020

The COVID 19 pandemic is having significant and, probably, lasting consequences on the Myanmar economy. The effects of the pandemic have resonated well beyond the...

The Critical Value of PPE in Myanmar’s Mon State

To cope with the situation, Mon State has been working with the UN Development Programme, to try to control the spread of the virus through an Awareness Raising a...

Climate danger grows in ‘vulnerable’ Myanmar after military coup

Fears that military could step up exploitation of resources to shore up finances, putting one of the world’s most climate-vulnerable nations at increased risk.

Yangon/Taipei – There is increasing concern that Myanmar is at risk of a serious environmental crisis, as the generals who seized power in a coup on February 1 focus on cementing their control and shoring up their position by stepping up lucrative but devastating policies of exploiting the country’s vast natural wealth.

The Global Climate Risk Index puts Myanmar among the countries most at risk from the climate crisis, frequently experiencing devastating floods and landslides as well as drought, exacerbated by decades of uncontrolled deforestation and mining of minerals and gems.

Keep reading

China-asean summit begins without a myanmar representative, myanmar students boycott classes following school reopening, infographic: what has your country pledged at cop26, what’s in the glasgow climate pact.

Over the past 20 years, the Southeast Asian country has experienced the highest weather-related losses alongside Puerto Rico and Haiti.

But tentative efforts to pursue more renewable energy projects and develop climate resilience under Aung San Suu Kyi’s government have been derailed since the military overthrew her National League for Democracy’s elected administration on February 1, suspending aid programmes and leading to the departure of private investors.

Developers who were awarded a solar power tender last year — totalling more than 1GW or one-third of Myanmar’s current dry season available capacity of 3.1GW — were unable to deliver, partly because of the coup.

The military in May launched its own solar power tender but was forced to extend the bidding deadline three times due to a lack of bidders. The latest deadline passed in mid-October but no official results have been announced to date.

Difficulties facing solar power companies mirror the broader risk of Myanmar missing out on climate finance opportunities post-coup.

“There are good investable projects in Myanmar which would build climate resilience such as natural reforestation and renewable energy projects,” said Vicky Bowman, director of Yangon-based Myanmar Centre for Responsible Business and former British ambassador to Myanmar. “But development partners seem frozen since the coup, and private sector investors instinctively now view Myanmar as high risk and look to alternatives in Southeast Asia, even though climate investments there may have as many problems in practice as Myanmar.”

Investors should see that there are still opportunities to work with local communities and companies to invest in natural capital and climate resilience, Bowman told Al Jazeera. “Otherwise the Myanmar people are hit with a double whammy of military rule and international neglect.”

Myanmar’s absence from the world’s top climate negotiations at COP26 in Glasgow last month reflected the country’s coup-induced international isolation, and the ongoing battle for recognition between the coup leaders and the National Unity Government (NUG), the parallel administration including officials from the elected government that was overthrown.

COP26 hosts, the UK, left Armed Forces Chief Min Aung Hlaing off the summit guestlist, while the event organisers, the UN Framework Convention on Climate Change (UNFCCC), disinvited Myanmar military government representatives, according to two sources involved in the matter.

Chit Win, the military-appointed chief diplomat in London who had evicted the removed government’s ambassador from the embassy after the February 1 coup, did manage to register temporarily on the event page with three associates. But they were denied entry and were subsequently taken off the system following a backlash from people in Myanmar.

Al Jazeera has seen copies of both nationally determined contributions (NDCs) — climate action plans and policy commitments — submitted by the NUG and the State Administration Council, as the coup leaders have dubbed their ruling body.

Both NDCs estimate the business-as-usual (BAU) scenario for coal to be about 30 percent of the country’s total power generation, which was what the NLD deputy energy minister Tun Naing reaffirmed in 2019.

The NUG claimed that they plan to decrease the share of coal from 33 percent (about 7940MW) to between 20 percent (3620MW) and 11 percent (2120MW) by 2030. The SAC gave the same figures.

But coal’s share of power generation is currently less than 1 percent, 30 times less than the higher-end estimates provided by the NUG and SAC. Sources attributed the discrepancy to efforts by some producers to encourage Myanmar to use more coal.

“The NLD government’s deputy minister [Tun Naing] at the time was being egged on by Japanese, Chinese and Indian coal interests, which no longer would be interested both for policy reasons and because it’s Myanmar post-coup,” an industry source in Yangon told Al Jazeera.

The NUG said it stuck with the overthrown NLD administration’s NDC for COP26 because they felt it had legitimacy from being drawn up by the government elected by the people, according to two senior officials at the NUG’s Ministry of Natural Resources and Environmental Conservation who requested anonymity due to security reasons.

“Considering the legitimacy provided by the Myanmar people to the ousted administration, we [NUG] submitted the NLD government’s NDC to COP26,” said a senior official at the NUG’s Ministry of Natural Resources and Environmental Conservation.

Another senior NUG official said that there had not been enough time for them to redraft the NDC.

“We have received some comments from Indigenous groups and will take their input into account as we revise the NDC,” the official said when asked about the role of ethnic communities in protecting forests.

NUG Deputy Electricity and Energy Minister Maw Htun Aung acknowledged the criticisms and said that the coal policy would be “reconsidered” and the energy master plan reviewed, although it is currently the SAC rather than the NUG which is in the capital, Naypyidaw.

“It does not make sense to focus on coal power. Even China is phasing out coal financing. We do not plan to scale up coal projects, and will work with ethnic communities to draft an energy policy on a federal level,” Maw Htun Aung told Al Jazeera.

According to government estimates last year, electricity in Myanmar comes from 20 gas-fired power stations, 62 hydropower facilities and a single coal-fired plant.

Resource exploitation

In addition to the slowdown in climate-related action and investments, environmental activists and analysts fear that the military will scale up logging, the teak trade, palm oil plantations and the exploitation of natural resources, such as jade, which supported the long-term survival of previous military regimes even under international sanctions.

The generals have also long profited from gem sales, and local media report a gem fair is due to take place in Naypyidaw this month.

Military-appointed agriculture minister Tin Htut Oo in November spoke about expanding palm oil plantations, according to the state-run Global New Light of Myanmar. The official paper said “implementations are underway” to make Tanintharyi Region, a major region in southern Myanmar bordering the Andaman Sea and Thailand, “a big oil pot based on palm oil”.

Mary Callahan, a Myanmar expert at the University of Washington in the United States, says the proposal is “disastrous for fragile ecosystems and endangered species”. Promoting palm oil plantations could lead to a new wave of land confiscation and more deforestation, she told Al Jazeera.

Weeks after seizing power and detaining Aung San Suu Kyi and her allies, Min Aung Hlaing also talked about developing hydropower dams.

This has sparked fears that the military might decide to restart the controversial China-backed Myitsone Dam in northern Myanmar, a pet project of former strongman Than Shwe that was halted by then-president Thein Sein in 2011 in the face of significant public protests . The generals have not mentioned Myitsone directly.

“We are very concerned that the military will fall back on old policies like large-scale hydropower, which could spell disaster for the country’s two major rivers – the Ayeyarwady and Thanlwin – the last two remaining large free-flowing rivers in tropical Asia,” said a senior staff member at an environmental NGO working on Myanmar, who declined to be named for security reasons.

Ethnic communities along the borders, coasts and hilly regions are also concerned about the climate risks.

“Of course we are worried about climate change. We are working on forest management and climate issues,” said a senior official of an ethnic armed group in northern Myanmar, who declined to be identified due to the sensitivity of the matter. Even though most of the territories controlled by his group are mountainous and protected from flooding, other climate-induced disasters such as cyclones, drought and landslides remain a threat to the local population. Since the coup, his group, which has long sought autonomy, has renewed fighting against the armed forces.

“Because of the coup and political crisis, it has become more difficult to address environmental challenges. For one, more and more international investors and partners have withdrawn from Burma,” he said. A key reason, he added, is that “the Burmese military leader will rely on natural resources to resolve their finances problem. Not only this junta but also successive regimes in the previous State Peace and Development Council [SPDC] era.”

The SPDC was the official name for the military government that seized power in 1988.

“Forests in the border areas controlled by ethnic groups are more secure than those in government-held regions,” the staff member from the environmental NGO said. “To help protect these forests, we need neighbouring countries and economic blocs like ASEAN and the EU to be on high alert for illegally-traded timber. Tackling demand is key.”

Still, in his written remarks submitted to COP26, the military-appointed Natural Resources and Environmental Conservation Minister Khin Maung Yi pledged to achieve 50 percent net emissions reductions by 2030 “if adequate international assistance is received”.

“Similarly, by 2030, the share of new renewable energy targets (solar, wind) will be increased from 2000MW to 3070MW,” Khin Maung Yi wrote.

But as long as the political crisis continues its downward spiral, neither the foreign assistance nor the energy investments on which the military is banking — with the possible exception of China — is likely to be forthcoming, according to diplomats and investors in Yangon.

Experts say environmental exploitation risks pushing more into poverty and increasing food insecurity, but as the generals focus on crushing any resistance to their rule, few have any confidence they will have the will to address Myanmar’s impending climate nightmare.

• Election 2024
• Entertainment
• Newsletters
• Photography
• Press Releases
• Israel-Hamas War
• Russia-Ukraine War
• Global elections
• Asia Pacific
• Latin America
• Middle East
• AP Top 25 College Football Poll
• Movie reviews
• Book reviews
• Financial Markets
• Business Highlights
• Financial wellness
• Artificial Intelligence
• Social Media

Heavy flooding in southern Myanmar displaces more than 14,000 people

Local residents wade through a flooded road near Shwe Maw Taw pagoda in Bago, about 80 kilometers (50 miles) northeast of Yangon, Myanmar, Monday, Oct. 9, 2023. Flooding triggered by heavy monsoon rains in Myanmar’s southern areas has displaced more than 10,000 people and disrupted traffic on the rail lines that connect the country’s biggest cities, officials and state-run media said Monday. (AP Photo/Thein Zaw)

Local residents wade through a flooded road in Bago, about 80 kilometers (50 miles) northeast of Yangon, Myanmar, Monday, Oct. 9, 2023. Flooding triggered by heavy monsoon rains in Myanmar’s southern areas has displaced more than 10,000 people and disrupted traffic on the rail lines that connect the country’s biggest cities, officials and state-run media said Monday. (AP Photo/Thein Zaw)

A local resident drives motorbike on a flooded road in Bago, about 80 kilometers (50 miles) northeast of Yangon, Myanmar, Monday, Oct. 9, 2023. Flooding triggered by heavy monsoon rains in Myanmar’s southern areas has displaced more than 10,000 people and disrupted traffic on the rail lines that connect the country’s biggest cities, officials and state-run media said Monday. (AP Photo/Thein Zaw)

Volunteers use a boat on a flooded road in Bago, about 80 kilometers (50 miles) northeast of Yangon, Myanmar, Monday, Oct. 9, 2023. Flooding triggered by heavy monsoon rains in Myanmar’s southern areas has displaced more than 10,000 people and disrupted traffic on the rail lines that connect the country’s biggest cities, officials and state-run media said Monday. (AP Photo/Thein Zaw)

Half-submerged houses are seen on a flooded road in Bago, about 80 kilometers (50 miles) northeast of Yangon, Myanmar, Monday, Oct. 9, 2023. Flooding triggered by heavy monsoon rains in Myanmar’s southern areas has displaced more than 10,000 people and disrupted traffic on the rail lines that connect the country’s biggest cities, officials and state-run media said Monday. (AP Photo/Thein Zaw)

Vehicles drive along a flooded road in Bago, about 80 kilometers (50 miles) northeast of Yangon, Myanmar, Monday, Oct. 9, 2023. Flooding triggered by heavy monsoon rains in Myanmar’s southern areas has displaced more than 10,000 people and disrupted traffic on the rail lines that connect the country’s biggest cities, officials and state-run media said Monday. (AP Photo/Thein Zaw)

Local residents wade through a flooded road in Bago, about 80 kilometers (50 miles) northeast of Yangon, Myanmar, Monday, Oct. 9, 2023. Flooding triggered by heavy monsoon rains in Myanmar’s southern areas has displaced more than 10,000 people and disrupted traffic on the rail lines that connect the country’s biggest cities, officials and state-run media said Monday.(AP Photo/Thein Zaw)

People use their mobile inside their electronic shop on a flooded road in Bago, about 80 kilometers (50 miles) northeast of Yangon, Myanmar, Monday, Oct. 9, 2023. Flooding triggered by heavy monsoon rains in Myanmar’s southern areas has displaced more than 10,000 people and disrupted traffic on the rail lines that connect the country’s biggest cities, officials and state-run media said Monday.(AP Photo/Thein Zaw)

Local residents use a raft to pass a flooded road in Bago, about 80 kilometers (50 miles) northeast of Yangon, Myanmar, Monday, Oct. 9, 2023. Flooding triggered by heavy monsoon rains in Myanmar’s southern areas has displaced more than 10,000 people and disrupted traffic on the rail lines that connect the country’s biggest cities, officials and state-run media said Monday. (AP Photo/Thein Zaw)

Local residents wade through a flooded road in Bago, about 80 kilometers (50 miles) northeast of Yangon, Myanmar, Monday, Oct.9, 2023. Flooding triggered by heavy monsoon rains in Myanmar’s southern areas has displaced more than 10,000 people and disrupted traffic on the rail lines that connect the country’s biggest cities, officials and state-run media said Monday. (AP Photo/Thein Zaw)

Two girls have foods at a restaurant on a flooded road in Bago, about 80 kilometers (50 miles) northeast of Yangon, Myanmar, Monday, Oct.9, 2023. Flooding triggered by heavy monsoon rains in Myanmar’s southern areas has displaced more than 10,000 people and disrupted traffic on the rail lines that connect the country’s biggest cities, officials and state-run media said Monday. (AP Photo/Thein Zaw)

• Copy Link copied

BANGKOK (AP) — Flooding triggered by heavy monsoon rains in Myanmar’s southern areas has displaced more than 14,000 people and disrupted traffic on the rail lines that connect the country’s biggest cities, officials and state-run media said Monday.

State television MRTV reported Monday evening that the number of displaced people in Bago township, about 68 kilometers (42 miles) northeast of Yangon, the country’s biggest city, had climbed to that figure, and they were taking shelter in 36 relief camps. It said almost 1,000 more people in Mon state’s township, just east of Bago, were sheltering in three relief camps, and there some evacuations in a northern part of Yangon.

A senior official at the Ministry of Social Welfare, Relief and Resettlement, Lay Shwe Zin Oo, said that constant rainfall in the Bago region that began last week caused the flooding in the low-lying areas of its capital, Bago township. She said there were no casualties reported so far.

Bago township recorded 7.87 inches (200 millimeters) of rainfall, its highest level in 59 years, Myanmar’s Meteorological Department said Sunday. Rain or thundershowers was forecast for across the country until noon on Tuesday.

One of the leaders of an emergency rescue team in Bago told The Associated Press that the flooding was up to about eight feet (2.3 meters) deep in low-lying areas and four feet (1.2 meters) downtown.

“Almost the whole area of the town was flooded,” Thant Zin Maung, chairman of the Mizzima Thukha Charity Foundation said by phone on Monday. “It is the third flood in the town this year and the worst in many years. All the monasteries in the town have opened relief camps. Charity organizations are evacuating people from low-lying areas as much as they can.”

A 55-year-old resident of Bago’s Pan Hlaing ward interviewed by phone said the floodwaters were as much as six feet (1.8 meters) deep in her neighborhood, and her family members were living on the second and third floors of their house.

The woman, who spoke on condition of anonymity because Myanmar’s military government prefers to tightly control the release of information, said the water was still rising steadily in her neighborhood, which had never flooded badly before.

Social Welfare Ministry official Lay Shwe Zin Oo said Bago evacuees were sheltering in relief camps, schools and Buddhist monasteries, while the authorities were providing food, drinking water and other essential assistance.

Reports in the state-run Myanmar Alinn newspaper on Monday said that trains that departed from Mandalay, the country’s second-largest city in central Myanmar, and from southern Mawlamyine township were halted en route. Scheduled departures from Yangon, the biggest city in the country, were canceled after rail lines were flooded by the rapid flow of water from mountain torrents and the spillage from dams in the Bago region.

MRTV said almost 1,000 people in Mon state’s Kyaikto township, just east of Bago, were sheltering in three relief camps, and there some evacuations as well in a northern part of Yangon.

Myanmar experiences extreme weather virtually every year during the monsoon season. In 2008, Cyclone Nargis killed more than 138,000 people. In July and August this year, floods in Mon, Kayin and Rakhine states and the regions of Bago and Magway killed five people and displaced about 60,000.

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 29 September 2023

The global costs of extreme weather that are attributable to climate change

• Rebecca Newman 1 &
• Ilan Noy   ORCID: orcid.org/0000-0003-3214-6568 2  

Nature Communications volume  14 , Article number:  6103 ( 2023 ) Cite this article

30k Accesses

11 Citations

1335 Altmetric

Metrics details

• Natural hazards

Extreme weather events lead to significant adverse societal costs. Extreme Event Attribution (EEA), a methodology that examines how anthropogenic greenhouse gas emissions had changed the occurrence of specific extreme weather events, allows us to quantify the climate change-induced component of these costs. We collect data from all available EEA studies, combine these with data on the socio-economic costs of these events and extrapolate for missing data to arrive at an estimate of the global costs of extreme weather attributable to climate change in the last twenty years. We find that US \(\$\) 143 billion per year of the costs of extreme events is attributable to climatic change. The majority (63%), of this is due to human loss of life. Our results suggest that the frequently cited estimates of the economic costs of climate change arrived at by using Integrated Assessment Models may be substantially underestimated.

Similar content being viewed by others

Quantifying uncertainty in aggregated climate change risk assessments

Luke J. Harrington, Carl-Friedrich Schleussner & Friederike E. L. Otto

Towards reliable extreme weather and climate event attribution

Omar Bellprat, Virginie Guemas, … Markus G. Donat

The burden of heat-related mortality attributable to recent human-induced climate change

A. M. Vicedo-Cabrera, N. Scovronick, … A. Gasparrini

Introduction

Extreme weather events have significant adverse costs for individuals, firms, communities, and regional economies. Based on the available data from the International Disaster Database (EM-DAT), the World Meteorological Organization 1 reports that there has been a sevenfold increase in the reported disaster losses from extreme weather since the 1970s.

While a part of this increase is due to increased reporting of disaster damage (especially in lower-income countries/regions or countries/regions that were previously more isolated), and because of increased exposure brought about by population growth and internal migrations to more exposed urban and coastal areas, a part of it is attributable to climate change. The most recent Intergovernmental Panel on Climate Change Report 2 notes it is virtually certain that there is a climate change component in the increase in reported disaster damage (at least of some types, with weaker evidence for others). The detection of anthropogenic changes in the frequency, severity, spatial location, and extent of extreme weather events is consequently important.

Extreme Event Attribution (EEA) is a methodological approach that examines the degree to which anthropogenic greenhouse gas emissions had changed the occurrence of specific extreme weather events that have indeed occurred. Using climate modeling tools, EEA quantifies the causal link between anthropogenic climate change and the probability and/or the intensity of specific extreme weather events by focusing on their specific circumstances and characteristics. EEA was first conceptualized by Allen 3 , who, together with some co-authors, developed a method to analyze the contribution of climate change to the risk of an individual weather event that could be clearly defined and quantified. This approach was first implemented for the 2003 continental European heatwave 4 —an event that led to high mortality, especially in France.

The EEA methodology compares the probability of an event that occurred with the probability or intensity of the same event occurring in a counterfactual world without anthropogenic emissions. From a probabilistic perspective, a Fraction of Attributable Risk (FAR) metric is calculated to describe what portion of the risk of an extreme weather event occurring is the result of climate change. Methodologically, these probabilistic methods have been approached from both a frequentist or a Bayesian perspective 5 , with possibly important consequences for the results thus obtained. We do not distinguish between these in our work here, given the relative paucity of Bayesian attribution work. The attribution approach based on FAR is known as the risk-based approach 6 . The alternative intensity approach calculates what share of a specific aspect of the risk (e.g., rainfall) was due to climate change. For instance, the 2017 Hurricane Harvey’s climate change-induced economic costs were analyzed by both risk-based 7 and intensity-based 8 approaches.

The economic costs associated with extreme weather events can be measured in two ways: First, these include direct economic damage, which occurs during or immediately after the event. Using flooding as an example, where the hazard is heavy precipitation, direct economic damage may include destroyed housing and roads, or lost crops. However, an extreme weather event can also cause indirect economic losses. These are declines in economic value-added because of the direct economic damage. Examples of these indirect losses are wide-ranging. For the flood example, they could include microeconomic impacts such as revenue loss for businesses when access routes are inundated by floodwater, meso-economic impacts such as temporary unemployment in the affected area, or even wider-ranging macroscale supply-chain disruptions. These indirect economic losses can often spill out beyond the affected area, and indeed even beyond the affected country/region’s borders. Indirect losses may also have long time lags, making them difficult to quantify. Generally, events that cause more damage will also lead to more losses, ceteris paribus. However, this relationship between direct damage and indirect loss is nonlinear, with high-damage events causing disproportionately many more losses as well. Because of these difficulties in quantifying indirect (flow) losses over a large variety of extreme weather phenomena in a large diversity of countries/regions and economies (thereafter referred to as countries for ease of exposition) and affected regions, this paper only focuses on the more easily quantified stock of direct damages.

By combining the data on direct economic damages, with the attributable share of the risk, we can quantify the climate change-attributable cost of these events. This attribution-based method for calculating the costs of climate change (from extreme weather events) differs fundamentally from other approaches to climate change cost estimation. Those other approaches use macroeconomic modeling embedded within climate models in various types of Integrated Assessment Models (IAM).

Given some of the data deficiencies in terms of temporal and spatial coverage, described in the following chapters, the purpose of this paper is not to produce a definitive quantification. At the current rate of progress in attribution research in meteorological science, we are still years away from obtaining a thorough and reliable global coverage of most socio-economically damaging extreme weather events. Our ability to measure the damage associated with these events is also far from being sufficiently comprehensive or accurate. Therefore, our aim is to demonstrate the use-value of the methodology, rather than reach an unimpeachable set of estimates. As better EEA studies and more thorough and exhaustive economic costs estimates for extreme events become available over time, and the method is refined, the precision of this approach’s estimates will increase in tandem.

Here, we use the frequency approach to aggregate the global economic damage from extreme weather events attributable to anthropogenic climate change. For that, we collect data from all available attribution studies with a frequentist analysis and extract their FAR estimate. We then combine these FAR estimates with data on the socio-economic costs of these events. While our research is not directly comparable to the IAMs, it provides an additional form of evidence that suggests that most IAMs are substantially under-estimating the current economic costs of climate change.

By examining the attribution information in conjunction with the cost information, we can calculate the climate change-attributed economic costs of extreme weather events. We first present these costs for the events in the master dataset, and then the results we obtain by extrapolating our findings to create a global estimate of these costs.

Attributed costs for events in the dataset

From the 185 events in the dataset—a net of 60,951 deaths are attributable to climate change—75,139 deaths that occurred due to climate change in events that became more likely and 14,187 deaths in events that have become less likely due to climate change. The net statistical value of life cost attributed to climate change across the 185 events in the master database is United States (US) $431.8 billion.

Anthropogenic climate change is responsible for a net $260.8 billion of economic damages across the 185 matched events (without the extrapolation described in the next section). This is equivalent to 53% of the total damages recorded. More than 64% of the climate change-attributed damages are connected to storms, which is expected given the high damages from events such as Hurricane Harvey. Furthermore, 16% of the attributed damages resulted from heatwaves, while floods and droughts are each responsible for 10%, and wildfires account for 2% of the net attributed damages. Lastly, cold events, calculated as a fall in climate change-attributed damages, are responsible for only −2% of net attributed damages.

Extrapolated global climate change-related economic costs of extreme weather

The results from extrapolating the attribution data across all global economic costs from extreme weather events are described below for the two extrapolation methods. Furthermore, we analyze the heterogeneity of globally attributed costs across time and event type.

The total climate change-attributed impacts, dictated by the respective extrapolation methods, have varying degrees of similarity. For heatwaves, the extrapolated estimates for deaths and damages are very closely aligned—less than one percentage point between the results from the two methods. For other event types, the disparities are wider. Notably, storm damages contribute substantially to attributed economic costs, making up over 60% of the total damages recorded in the EM-DAT extreme weather event dataset. There are two data comparisons where the estimates differ widely (greater than ten percentage points) between a global and continental approach: flood deaths (45%) and storm deaths (132%). These discrepancies in flood cost calculations occur because the FAR data points vary widely across attribution studies. These flood results are significantly impacted by a regional average FAR for floods in Africa of a decrease of 0.49, meaning that an estimated 49% of the decrease of risk of flooding in Africa can be attributed to anthropogenic climate change. Comparatively, the regional average FAR for floods in all other regions is positive, indicating an increase in risk resulting from climate change. This has a relatively large impact on the regional extrapolation results as floods cause a relatively high number of deaths in Africa and a comparatively low level of damages. This is a common pattern for disaster mortality and damage in low-income countries 9 , 10 . Moreover, the discrepancy between climate change-attributed deaths from storms is primarily driven by a regional average FAR in Asia (0.81) at least 20 percentage points higher than the FAR in all other regions. This has a notable impact on the results given the high number of storm-related deaths in Asia. However, it is important to recognize that the two noted regional average FARs that impact these results are calculated from a few data points—3 for floods in Africa and 1 for storms in Asia. Due to the lack of data relating to important event-type and continental combinations, the global average extrapolation method is used in all the tabulations described below, to minimize over-reliance on a small number of attribution studies.

The economic value of life lost to climate change-attributed extreme weather is obviously very dependent on the assumed value of statistical life. When using the US-UK mean Value of Statistical Life (VSL; as described in the data section), the climate change-attributed cost associated with mortality is a net US$ 1.79 trillion from the global extrapolation method.

The estimated global cost of climate change over the 2000–2019 period is summed up in Fig.  1 . These results are calculated using the global average FAR extrapolation method, which is less sensitive to singular studies than the regional average FAR approach. In aggregate, the climate change-attributed costs of extreme weather over 2000–2019 are estimated to be US \(\$\) 2.86 trillion, or an average of US \(\$\) 143 billion per year.

These are the globally aggregated data for climate change-attributed impacts of disasters that were associated with extreme weather, using data collected from the Emergency Management Database - EM–DAT. Total costs represent the full estimate of the economic damages associated with an event, while the climate-attributed costs represent only the portion for which climate change is responsible. The combined bar represents the full cost, with the transparent portion representing the (statistical) lives lost and the solid portion are economic damages.

In an alternative calculation, in which we used the median FAR for each type of event, instead of the average (mean) FAR, the results are only larger: US \(\$\) 167 billion. This larger result is somewhat surprising, as our intuition was that larger events are more likely to be investigated in EEA projects, so that the median FAR should, in general, be smaller than the average FAR. This is not actually the case, suggesting this possible bias may be overstated.

This aggregate of US \(\$\) 143 billion, annually, is split across attributed human costs (statistical loss of life) of nearly \(\$\) 90 billion and economic damages of \(\$\) 53 billion per year. The distribution of costs is highly variable across years. The year with the lowest costs attributed to climate change is in 2001 at \(\$\) 23.9 billion, while the year with the highest climate-attributed costs is 2008 with \(\$\) 620 billion. The years in which costs reach high peaks—notably 2003, 2008, and 2010—are predominantly because of high mortality events. The events that drive these peaks are the 2003 heatwave across continental Europe; Tropical Cyclone Nargis in Myanmar in 2008; and the 2010 heatwave in Russia and drought in Somalia.

The aggregate result presented is subject to uncertainty given the limited number of data observations and the exploratory nature of this methodology. When considering a global FAR, for each type of weather event, one standard deviation below and above the mean the respective attributed cost per year is US \(\$\) 58 billion and US \(\$\) 228 billion, respectively. Storms drive the largest difference, given their contribution to absolute cost, however, the largest standard deviation is in flood events.

The peaks in climate change-attributed costs differ when we look solely at damages and exclude the statistical loss of life. The greatest peaks in monetary damages occur in 2017 and 2005. Storm events in the United States drive these—in 2005, Hurricanes Katrina, Rita, and Wilma together caused \(\$\) 123 billion in attributed damages, and in 2017, Hurricanes Harvey, Irma, and Maria were responsible for \(\$\) 139 billion in climate change-attributed damages.

Figure  2 shows how total and climate change-attributed costs are distributed across high (gross national income GNI per capita>USD 12,535), upper-middle (GNI per capita between USD 4046 and US \(\$\) 12,535), lower-middle (GNI per capita between US \(\$\) 1036 and US \(\$\) 4046), and low-income (GNI per capita <US \(\$\) 1036) countries. This provides context for how different countries, especially the vulnerable ones, are being impacted by climate change-induced extreme weather. As per the available data, high-income countries have the highest climate change-induced economic costs at around 47% of the total. A few elements drive this, the primary being the United States having high asset exposure to storms.

The aggregated mortality and economic damage costs for each country/region income group, using the 2020 World Bank’s income classification.

However, the distribution of economic costs from extreme weather events across low to high-income countries is also likely a product of data availability and measurement. High-income countries have more resources and expertise to gather economic data when an extreme weather event occurs, while lower-income countries do not have this same level of resource availability.

These extrapolated estimates for the climate change-induced cost of extreme weather can be calculated as a proportion of gross domestic product (GDP), as shown in Fig.  3 . Using the global average extrapolation method, the total economic cost, inclusive of damages and statistical loss of life, can be presented as a proportion of annual global GDP. This is not a direct comparison because GDP is a measure of economic flow, i.e., measured over a defined period, whilst damages and loss of life are a stock variable, i.e., measured at one point in time. It is, however, still a measure of the relative importance of these shocks on the affected economies. Climate change-attributed economic costs from extreme weather events vary between 0.05% to 0.82% of global GDP annually over the study period.

These are the globally aggregated data for climate change-attributed impacts of disasters that were associated with extreme weather, using data collected from the Emergency Management Database - EM–DAT. These represent only the portion which are attributable to climate change.

The limitations of comparing stock and flow withstanding, we can compare the annual average attributed costs for levels of GDP across countries at varying levels of development. This shows that low-income nations, as a cohort, experience the relative economic costs of climate-attributed extreme events to a greater degree—at near an average of 1% of GDP per annum, compared to 0.2% for high-income countries, as shown in Fig.  4 . This differential is almost entirely driven by high levels of loss of life in lower-income countries, which may be the result of fewer early warning systems and safety procedures in place in these areas. In this context, we note our decision to use a uniform Value of Statistical Life across countries. That, of course, means that in lower-income countries, where mortality is highest, the relative importance of the loss-of-life measure is higher (as the value of assets is lower). Moreover, a smaller difference in economic damages may be the result of offsetting factors of higher-value assets in high-income countries, although buildings and infrastructure are likely to be more resilient to weather events. This finding uncovers disparity in the costs of climate change, and a potential for inequality to become further entrenched due to greater extreme events.

This figures uses the 2020 World Bank’s income classifications and are based on the average cost per annum over the 2000–2019 sample period.

Comparing the cost estimates with integrated assessment models

There are several different approaches used to estimate the economic impact of climate change, with the attribution-based method of this research presenting an alternative option. The attribution-based method is an event aggregation approach; it therefore differs significantly from the macroeconomic methodology used in IAMs. Commonly, IAMs characterize damages as a polynomial function of the deviation of average annual temperature from pre-industrial times, as done, for example, in the Dynamic Integrated Climate Economy (DICE) model 11 , 12 . DICE approximates the damages from climate change, as a proportion of the global economy, according to the damage function shown in Eq. ( 1 ).

Where \(T\) is the change in global mean surface temperature above the pre-industrial threshold, currently estimated to be around 1.2 °C in 2020 13 . To allow us to compare the results from attribution to those of DICE, we used the parameters from the DICE 2016R model: φ 1 = 0; φ 2 = 0.00236, and the same temperature deviation data. This approach from DICE is not unique for IAMs. The Policy Analysis of the Greenhouse Effect (PAGE) model, which was used in the Stern Report 14 , also calculates economic and non-economic damages from climate change using a polynomial function. However, PAGE uses regional temperature deviations rather than the global one 15 .

From this basic calculation, as per the DICE model, the assessed global damages from climate change over 2000–2019 is estimated to be US \(\$\) 4.04 trillion. Based on an aggregated event attribution approach, the approximation in this research is \(\$\) 2.86 trillion, meaning the DICE estimate is ~40% larger. The comparative calculations of climate change costs from DICE and the attribution-based approach, by year, are shown in Fig.  5 . However, these two metrics are not attempting to measure the same quantity, with two key differences:

The total cost calculated by the extreme event attribution method is shown relative to that calculated by the DICE damage function, as total in gray, and only the DICE estimates for extreme weather events (EWE) and other environmental damages in yellow.

First, the IAMs produce a measure of decline in economic flow (proportional to global GDP) while attribution-based estimates measure loss in economic stock. This is the same distinction between damage and loss we described earlier.

Second, the attribution-based estimates solely measure the net economic cost of extreme weather events caused by anthropogenic activity, while IAM models attempt to estimate the overall annual loss caused by climate change. This should include extreme weather costs as well as many other types of costs and benefits from changing crop yields, ocean acidification effects, sea-level rise and its attendant impacts, environmental degradations and ecosystem disruptions, spending on adaptation, and many other types of impacts.

These factors limit the comparability of the IAMs measures and the attribution results. However, it is notable that extreme weather events are only one category of the damages that are, in theory, included in the DICE measure. The key limitation of IAMs, which is highlighted through comparison with the attribution-based approach, is that they account only for changes in average temperature rather than the change in temperature distribution, and specifically in the tail end of the distribution of weather-attributed. By focusing on the deviation in the average temperature, the IAMs fail to capture changes in extremes, plausibly the most important current impact of climate change.

Nordhaus acknowledges that DICE, and other IAMs, generally omit the impacts of extreme weather (as well as biodiversity, ocean acidification, catastrophic climate risks, and more). The solution he used to account for this limitation is to add 25% of the monetized damages in the DICE model 16 . This is a very subjective adjustment, which would assume that extreme weather accounted for a maximum of \(\$\) 0.8 trillion ( \(\$\) 0.55 trillion would mean that extreme events account for the full value of the 25% adjustment to the DICE estimate) across 2000–2019, relative to the climate attribution-based figure of \(\$\) 2.86 trillion. This suggests a large underestimate that exhibits how DICE fails to accurately assess the economic impacts of climate change from extreme weather.

In addition, we can compare the attribution-based results to the Framework for Uncertainty, Negotiation, and Distribution (FUND) IAM, which is notably more complex than DICE. The FUND model differs from DICE as it calculates damages at a sectoral level, with nine sectoral damage functions operating across 16 regions of the world 17 . The key sector of interest in FUND, for this research, is the storm sector which is the only sector that is reflective of how climate change impacts the economic cost of extreme events. The FUND model calculates estimated damages (capital loss) and mortality for tropical and extra-tropical storms. This is a more sophisticated inclusion of extreme weather event costs compared to the DICE approach. As an example, the total damages and mortality from tropical storms in FUND are calculated for each region using Eqs. ( 2 ) and ( 3 ).

In the FUND model, the key inputs in the damage function are temperature change over pre-industrial levels (T), per capita income ( \(y\) ), current damage as a fraction of GDP (α), current mortality as a fraction of the population (β), and income elasticities of storm damage ( ϵ , η ).

The MimiFUND web page, an accessible source for viewing the FUND model and results, estimates current damages from tropical cyclones as higher than the damages from extreme weather events calculated in the attributed results 18 . FUND calculates the current damage from tropical cyclones as, on average globally, 0.08% of GDP. Comparatively, the climate change-attributed damages from storms calculated in this research are 0.06% of GDP on average per annum. Further, climate change-attributed damages from all extreme weather events in the research equate to an average of 0.07% of GDP per annum. The difference in the FUND tropical cyclone estimation and the climate change-attributed costs of storms is an interesting comparison. It may be a discrepancy that can, to some degree, be explained by underestimated economic data recorded in EM-DAT the attribution estimates use. Furthermore, FUND estimates the current mortality from tropical cyclones to be on average 0.00015% of the population, while attribution-based results estimate that storms on average have a climate change-attributed mortality rate of 0.00009% per annum. These inconsistencies are illustrative of how, especially when data is lacking, it is beneficial to analyze multiple approaches to quantitative research—with the macroeconomic IAMs and event attribution techniques providing valuable contrasts.

Limitations of the attribution-based approach

This research explores the potential of an attribution-based method for estimating the human-induced climate change costs of extreme weather globally. Although event attribution has been used to measure the climate change-related economic impact of individual extreme weather events before, this methodology has not yet been extended to a global approximation 7 , 8 , 19 , 20 . As such, this study does not provide a silver-bullet approximation of the cost of extreme weather events. There are important limitations of the attribution-based approach, primarily due to restrictions on the quantity and quality of data. These limitations are explored in detail below to highlight the progress required so as to improve these estimations.

When examining methodological limitations, we note that extreme event attribution is a young but rapidly expanding sub-field of climate science. The literature is limited, methodologies are continuously being refined, and the field’s development faces some methodological and epistemological challenges. Notable limitations are the uneven geographical coverage of attribution studies and the lack of attribution studies conducted on several important classes of extreme weather events. These lacunae are significant, given the relatively small number of attribution studies conducted overall.

Extreme event attribution studies are more commonly conducted in high-income countries, with lower-income regions barely represented in the literature. In our database, only 8% of the attribution studies are conducted on extreme events in Africa, while over half of the events studied are in either North America (23%) or Europe (25%). In recent years, there has been a greater attempt to balance the geographical distribution, particularly by the World Weather Attribution (WWA) network 21 . The WWA use the following human-based threshold to determine which events to consider for study: the event resulted in greater than 100 deaths, 100,000 people affected, or more than half of the total national population affected 22 . In contrast with an economic loss threshold, a human-based threshold leads to less bias against low-income countries where physical assets are of lesser value 23 .

Still, extrapolation based on the total average FAR per event type leans over-proportionately on event probabilities from high-income regions (and China). The data gaps in Africa, South America, and Oceania, in particular, result in over-reliance on few data points in the calculation of a regional average FAR, or the use of an imperfect substitute (e.g., the global average FAR). This is a notable limitation because different regions of the world are subject to different climatic systems and environmental conditions. Consequently, the FAR for specific extreme weather events will differ by region and even more locally within countries. Improved geographical coverage of event attribution studies would improve the robustness of the methodology presented, especially if this allowed for greater granularity in the extrapolation method.

The second issue with event attribution data is the uneven spread of research across different event types. About a third of all attribution studies analyze the role of climate change in inducing heatwaves, the best-represented event category. Comparatively, storms, which are most important when considering the economic cost of extreme weather, make up only 8% of the studies in this dataset. One reason behind this discrepancy is the degree of difficulty associated with attributing different event types. Heatwaves, and similarly extreme cold events, generally result in the most reliable event attribution estimates as the direct thermodynamic effects for these events are comparatively straightforward 24 . In contrast, events such as agricultural droughts are caused by several compounding factors—such as precipitation, temperature, and soil moisture—making the attribution process significantly more complex. Cyclones are also complicated to model, which means that large-ensemble attribution studies of these storms have only become feasible in recent years, though a high computational cost for each simulation still persists 24 . As an example, Tropical Cyclone Idai that hit Malawi and Mozambique in 2019, and caused additional damage in Madagascar and Zimbabwe, was the costliest cyclone to have hit Africa with record-setting intense winds and rainfall, but even this event has not yet been analyzed in an attribution study.

Beyond the spatial and event-type coverage deficiencies, the framing of an event attribution study can induce large differences in how the role of anthropogenic emissions is quantified. Different framings would be appropriate for answering different questions 25 . One such example, which gained significant attention, was the 2010 Russian Heatwave. Two seemingly contradictory event attribution studies were conducted—one finding a negligible role of human-induced climate change, and the other identifying a fivefold increase in likelihood 26 , 27 . However, the framing of this event was central to this difference. The first paper analyzed the change in intensity, whilst the second analyzed the change in frequency. Moreover, subtle framing differences—such as whether attribution is conditioned on the background atmospheric conditions (e.g., El Niño-Southern Oscillation), or sea surface temperature conditions, or whether the counterfactual removes a single factor (greenhouse gas emissions) or all anthropogenic factors—can have a notable impact on the attribution quantification 6 , 28 . More reassuringly, recent examination of the variability of results due to different methodologies used in the EEA studies themselves suggest these results do not vary that much 29 .

An attribution study must also define the spatial and temporal boundaries of the event being analyzed. These decisions ultimately impact the final FAR that is calculated 30 , 31 . As long as these definitions of the event in the attribution study align well with the extent of the economic estimates produced by EM-DAT, this issue may not be as important. However, given the paucity of attribution studies, and the lack of detail about the geographical span of the EM-DAT data we could use, this was not always verifiably the case.

Commonly, event definition should reflect the main determinants of the event’s impacts, as the authors seek to answer what role anthropogenic climate change played in creating the economic and societal impacts of an event 6 . For example, calculating a FAR using a single-day rainfall measurement (rather than, say a 7-day aggregate measure) may be preferable when a flood has caused devastation because of the short burst of intense rainfall that caused water to accumulate. For this study, attribution studies that define events based on the determinants of the most important human and economic impacts are beneficial. There of course can be multiple impacts, and these can be related to different event definitions. It is therefore not always clear which impact should be used when defining the event parameters. This is particularly salient for extreme events that are not meteorological in nature, such as flooding (hydrological) and wildfire (ecological), as these are also related to multiple climate parameters.

A closer geographical and temporal match between the FAR and economic impact data recorded in the dataset makes the calculation of attributed costs more reliable. However, events are not always defined in this way, as there may be barriers that prevent climate researchers from using such impact-based definitions. For example, it is often found that meteorological observational datasets are not extensive enough—across time or space—to allow an attribution study based at a specific locality or on a specific factor. Therefore, the event definition sometimes must deviate from the boundaries of the actual impacts to ensure the adequacy of data records 6 .

Finally, it could be argued that attribution studies using the intensity approach, together with well-calibrated damage functions (that define the functional relationship between damage and the intensity of an event) might be a more appropriate input into our analysis. Two reasons led us to prefer relying on FAR quantifications. First, these are much more common in the attribution literature, allowing us to expand our sample of events. Second, we do not have well-calibrated damage functions, as these can be spatially and temporally specific, and are unique to each type of event (even different types of storms, for example, will necessitate different damage functions) and the social and built infrastructure exposures and vulnerabilities that differ substantially across locations 32 .

One possibility to account for the uncertainty associated with the results is to generate a range of estimates based on the identified range of FARs. For example, we could make a similar calculation to the one we present, but based on the lowest (or highest) FAR identified for each category of hazard/region combination, rather than the average. While this will create a range of estimates, we are not able to conclude anything about the distribution function underlying this range. We therefore see such a range as potentially misleading, and prefer not to present these kinds of exploratory sensitivity analyses. However, since all of our data are posted publicly, an interested reader can explore this further, of course.

This research looked at events that became more or less likely to occur due to anthropogenic climate change. However, there may still be an embedded underrepresentation of events that have become less likely because of human-induced climate change; maybe because of publication bias, or because other factors that are associated with event selection. This is because attribution studies are typically conducted on major events, one that attracted the researchers’ attention, and are not conducted at all on events that became mild because of climate change or have not occurred at all. Since there have been no recent occurrences for these events, it is impossible to quantify reliably their economic costs. There is no way to overcome this bias, but the available evidence seems to suggest that even before the main impacts of climate change have started to be felt, the importance of these type of decreasing frequency or intensity events had been relatively less prominent than that of increasingly likely ones.

The economic data used to quantify the global cost of climate change-attributed extreme weather events in this study are subject to an additional set of limitations. They reflect the current best-available estimates, but there are possible limitations regarding the data’s quality, coverage, and granularity.

The economic cost data used in this research underestimates the true costs of climate change over the study period. Most importantly, our estimates include only direct loss (damage) and not indirect loss ones. These later losses are difficult to measure. This is the case, among other examples, for productivity losses in a heatwave 33 . For example, the Australian Climate Council attempted a thorough approximation of the total economic impact of Australia’s southwestern heatwave in 2009 34 . They estimated that the heatwave was responsible for up to AU \(\$\) 800 million in indirect financial losses— predominantly caused by power outages and transport system disruptions. This same event, as recorded in EM-DAT, detailed no asset damages at all. An inventory of events with the economic impacts differentiated into direct and indirect economic losses, at a bare minimum, would give decision-makers a better understanding of the wider economic impact of anthropogenic climate change 19 .

The number of people affected by disaster events is recorded in EM-DAT. With the global average extrapolation approach, we found that climate change affected 1.4 billion people through extreme weather events between 2000 and 2019. Affected, in line with the EM-DAT definition, means requiring immediate assistance following the event. This could range from an acute need for life-saving medical attention and potentially sustaining life-long injuries, to the long-term provision of basic survival resources, or just supply of very short-term (hours or days) of emergency provisions. Clearly, there are significant economic costs associated with these affected people, including healthcare costs, costs of provision of other basic services such as emergency shelters, and potentially other longer-term welfare costs. However, given the extensive but imprecise range of costs that could be associated with someone being classed as affected, using a single monetary value for this group may be misleading. Therefore, these costs are not included in our calculations, but form an additional source of underestimation that is embedded in our results.

In addition, people can be adversely affected by an extreme weather event in ways that do not include requiring immediate medical assistance or basic survival needs. For example, people may suffer from mental health impacts (e.g., post-trauma), the loss of access to education, or the loss of their job if their place of employment is harmed. These will not be counted as having been affected, under the EM-DAT definition, yet suffer high economic loss. These costs are not captured in any available dataset.

While the limitations of this approach are significant, this research demonstrates how a more global approximation of the human-induced extreme weather event economic costs could be constructed. Each of the limiting factors described above has the potential to be reduced with more data collection and more research.

This research relies on two elements—the level of anthropogenic emissions and their consequential effect on climatic extremes (captured by the FAR), and the economic costs from extreme weather events. To minimize the climate change-attributed costs from extreme weather in the coming decades, there would need to be increased mitigation that will reduce the FARs, or an increased adaptation that will reduce the economic costs associated with extreme events, or preferably both.

Adaptation can make a considerable difference to the climate change-attributed economic impact of extreme weather events right now. Adaptation policies could include infrastructure development such as building flood protection or improving early warning signal systems for extreme weather events. A pertinent example of this, in our context, has been implemented in continental Europe, where the 2003 heatwave claimed upwards of 70,000 deaths, 55,400 of which were attributed to climate change. The extremely high mortality of this event shocked European countries into creating effective heatwave adaptation strategies to prevent a repeated high volume of deaths in the future. France, as an example, introduced a heat warning system that is triggered after three days of persistently high temperatures 35 . This system can enact the closing down of schools and public areas, the operation of a public heatwave helpline, and the opening of cool rooms in public buildings. This made a marked impact on the fatality of subsequent heatwaves. The heatwave in 2019 was hotter than that of 2003 in many locations, yet, in France, there were less than 1500 deaths, compared to over 19,000 in 2003. This clearly demonstrates how a well-designed and implemented adaptation policy can help reduce the climate change-attributed costs of extreme weather significantly. The results of this research, we hope, can provide an impetus to increase spending on climate change adaptation policies as it clarifies some of their benefits, in terms of avoided harm. It can also allow for better targeting of adaptation spending. This should ultimately help reduce climate change-attributed economic costs from extreme weather in the future.

For now, at the very least, more event attribution studies are needed, and the geographical and event-type representation of studies improved to align better with human impacts. This, in addition to better economic data, will allow the approximation of the global climate change-attributed economic cost of extreme weather to be improved, and thus form the basis for quantification of allocations through the Loss & Damage Fund. As such, this attribution-based method can also increasingly provide an alternative tool for decision-makers as they consider key adaptations to minimize the adverse impact of climate-related extreme weather events. This type of evidence can also fill, potentially, an evidentiary gap in climate change litigations that are attempting to force both governments and large emitting corporations to change their policies 36 , 37 .

Allen 3 suggested EEA as a method of comparing probabilities to quantify the contribution of climate change to the probability of an individual weather event occurrence. From this type of estimation, a FAR metric is calculated to describe what portion of the risk of an extreme weather event occurring is the result of climate change 6 . For this methodological approach, the weather is simulated under the current climate, and similarly, simulated under a counterfactual climate that is free from human greenhouse gas (GHG) emissions. This provides information on the degree to which climate change has altered the risk of event occurrence.

Economic costs of extreme weather disasters

An extreme weather phenomenon by itself is not a disaster, but when a weather-driven hazard intersects with an exposed and vulnerable population (i.e., populations with characteristics that make them susceptible to adverse hazard impacts 2 ), the extreme weather event becomes a disaster 2 . These events, when they occur, can cause a range of economic impacts. The Intergovernmental Expert Working Group on Indicators and Terminology Relating to Disaster Risk Reduction provides a set of relevant definitions. Firstly, a disaster can cause damages which occur during and immediately after the disaster. This is a stock amount that is measured in physical units and describes the total or partial destruction of physical assets, the disruption of basic services, and damages to sources of livelihood in the affected area. Relatedly, direct economic loss is the monetary value of these disaster damages, for example, the monetary value of totally or partially destroyed physical assets.

Secondly, disasters can cause indirect economic losses, defined as a decline in economic value-added because of direct economic loss (damages) and/or other disruptions caused by the disaster. These indirect losses can occur outside the disaster area and with a time lag and are measured as a flow variable (per unit of time). Indirect losses are more challenging to measure since they rely on developing a counterfactual (a without-a-disaster scenario). Finally, impact is the total effect of a disaster, including both negative effects (e.g., direct losses) and positive ones (e.g., indirect economic gains). Impact includes economic, human, and environmental impacts, including death, injuries, disease, and other adverse effects on human physical, mental, and social well-being. Some of these are intangibles that are rarely measured systematically after disaster events. This research will attempt to understand disaster impacts in aggregate and present them in terms of monetary valuation, referred to as the total economic cost. This is predominantly comprised of direct losses and the statistical value of life lost, given the limitations of the data collected in EM-DAT.

This approach, then, does not measure indirect losses. These may be significant. For example, the 2023 wildfires in Canada have imposed significant economic losses not only on Canadian cities impacted by the air pollution the fires generated, but they adversely impacted vast swathes of the densely populated North-East region of the United States (including New York City). None of the approaches discussed therein can account for these indirect losses, even though these could conceivably be orders of magnitude larger than the original damage wrought by these events (and were likely much larger in this specific case).

A challenge in aggregating damage data across international borders is the question whether damages have equal value in different countries. This problem is clearest for mortality. Typically, governments explicitly or implicitly attach a value of statistical life (VSL) to risk-of-mortality calculations, and these VSLs can be dramatically different across different countries. In low- and lower-middle-income countries (countries with GNI per capita <USD 1026 and USD 3995, respectively), a human life can be saved for a relatively much lower cost compared to upper-middle or high-income countries, so fiscally constrained governments in such countries typically use a much lower VSL in their policy decisions. However, we choose to use an identical value for a life anywhere. In this case, the flipside of this problem is to ask whether the monetary value of asset damage can be similarly aggregated internationally. Clearly, a \(\$\) 1 of value in a very wealthy country/region is a lot less consequential than a \(\$\) 1 in a very poor one. So as to clarify this further, we therefore present most of our results also separately for low, low-middle, high-middle, and high-income country groupings. Noy 9 and Wilson and Noy 38 provide more discussion of this issue and propose an alternative approach, which relies on a measure similar to Disability-Adjusted Life Year (DALY), instead of monetary values. However, the main argument for using a monetary unit of account herein is that our results are then comparable to others (for example, in the IAMs) and can also form a basis for Loss and Damage calculations.

Besides not measuring indirect losses, we also emphasize that the monetary measure we use (aggregating loss of life and direct damages) disregards any concerns about distributional consequences, even though these may very well have a significant impact on well-being. Our measure is purely utilitarian, but it could be enhanced with explicit assumptions about the specifications of the individual utility functions and the aggregate social welfare function. These kinds of approaches, however, will require a possibly controversial set of ethical and modeling choices.

Using event attribution to estimate the economic costs of climate change

Allen 3 proposed that EEA enables differentiating economic losses from extreme weather between those that are caused by natural variability and those caused by past anthropogenic activity. Frame et al. 7 suggested how this approach can attribute climate change-induced economic costs when both a fraction of attributable risk and economic cost inputs are available for a set of individual events. The approach they used is straightforward—multiply the fraction of attributable risk by the estimated economic costs. With some assumptions about aggregation and generalizability of the calculated FARs, this same process can be replicated across different types of economic impacts—including deaths, and even indirect losses—to provide individualized assessments of the climate change-attributed value of each of these impacts of extreme weather events. Frame et al. 20 estimated climate change-attributable insured costs of major flooding events in Aotearoa New Zealand based on the aggregation of attributed costs from 12 major flooding events. Some recent papers have looked at counting mortality and morbidity from heatwaves and attributing these to climate change 39 , 40 , 41 .

Here, we aggregate all the relevant EEA studies (see details below), and their corresponding economic impact assessments and then extrapolate from these to obtain an overall estimate of the climate change-attributed impact of all recent extreme weather events globally, for which economic impact estimates are available. We then compare these estimates to some of the existing assessments of the current costs of climate change from the IAMs.

Other methods for estimating the global economic impact of climate change

Most attempts to quantify the global impact of climate change use IAMs. Well-known, well-regarded, and equally well-criticized examples include DICE 12 and FUND 42 . The IAMs, typically, link the economic system with the climate system by using damage functions that express the economic impact of climate change as a function of a global or regional mean of annual mean temperature 43 . This, of course, captures the change in the mean, but not in the tail ends of the distributions of extreme weather 44 . Therefore, these models tend to include the costs of extreme weather using ad-hoc additional modification to the damage function, or they are omitted entirely 12 , 45 , 46 , 47 .

Of course, comparing the IAMs to our approach using FARs is problematic, since the two are aiming to measure different quantities. IAMs model the economy and measure the decline in the flow of economic activity over time because of climate change—a very different approach to ours. Our argument is not a criticism of the IAM approach, per se, what we suggest is that the adjustments IAMs typically make to account for the impact of extreme weather events are significantly understated.

Given the limited availability of FAR studies, our approach cannot be applied across every extreme weather event. Consequently, the global application we pursue here relies on the extrapolation of known FAR values to other events for which there are no EEA studies, and a reliance on patchy economic data, to assess impacts (we discuss these data limitations further in the following sections). Van Oldenborgh et al. 22 argue that, with the current stock of EEA studies, we should consider the possible selection biases in the availability of EEA studies. Generally, events with higher human and economic impacts will be more likely to be analyzed, events in high-income regions and more densely populated areas are more likely to receive attention, and event types that become less likely because of climate change may be underrepresented in the analyses as well 22 , 48 .

However, given the fundamental importance of empirical evidence to drive an informed climate change policy response, we use aggregation and extrapolation based on available knowledge, while acknowledging the limitations and inherent biases that might detract from the accuracy of such an exercise. Implicitly, we assume that the significant disaster events for which attribution studies are available are representative of the other damaging disasters of the same type, occurring in the same geographical region. Given the lack of a superior alternative, we see this is as an acceptable approach. We argue that all current approaches to estimate the costs of climate change are limited by their methodological straightjackets.

Indeed, we argue below that the conventional IAM assessments is even less robust, and underestimates many of the most important impacts associated with extreme weather. This makes the exploration of an alternative and complementary cost estimation method fundamentally important, even if this method has its own flaws.

Dataset collection and terminology

The fraction of attributable risk (FAR) is a metric that describes the portion of the risk of the extreme weather event for which anthropogenic climate change is responsible. When the risk of an event has increased due to anthropogenic GHG, it is calculated as shown in Eq. ( 4 ). This can be referred to as the fraction of attribution risk (FAR).

P 0 = Probability of a climatic event without anthropogenic GHG present.

P 1 = Probability of the event occurring within the current climate system (with anthropogenic GHG).

A FAR value of 1 means that the event would not have been possible in the absence of anthropogenic climate change. While a FAR of 0 indicates that climate change had no influence on the probability of the event occurring 49 . More information on the data collection procedures we used is available in the  Supplementary Information File.

To assess the economic cost of mortality, we utilize Value of Statistical Life (VSL) calculations; this is the standard approach in many policy decisions (for example, about road improvements for safety). The VSL describes a marginal rate of substitution between money and mortality risk in a defined period 50 and the VSL estimates differ very dramatically across countries 51 . The VSL we use here is an average of two VSL estimates used by the governments of the United States and the United Kingdom. The first is the United States Department of Transportation estimate for 2020, which sets the VSL at US \(\$\) 11.6 million, which itself is an average of VSL estimates from across the academic literature 52 . The second estimate is from the UK Treasury, which assesses the VSL to be £2 million, estimated from average values from survey data looking at representative samples of the population 53 . According to Viscusi 54 , the non-US median VSL is \(\$\) 7.36 million (adjusted to 2020 USD). For this study, the benchmark result of US \(\$\) 7.08 million per life lost is used, which incidentally is not very far from the non-US median reported by Viscusi 55 . For simplicity, and more importantly on equity grounds, we use this same VSL for deaths in every country/region, and every year, implying that death has an equivalent economic value regardless of the time and place in which it occurred.

Data for individual extreme weather events were matched, where both a FAR and economic data had been collected. These events were collated to form the dataset that provides the basis for our empirical analysis. The available data were refined to ensure the master dataset contained the best-available estimates for each included event.

When events with multiple attribution studies were available, the Scimago Journal Rank (SJR), in the year of publication, was used as a proxy for the research quality. The SJR impact factor was sourced from https://www.scimagojr.com/ ; it represents the rank of a journal’s scientific influence and is calculated from a weighted measure of the citations a journal receives. The weighting is determined by the prestige of the publishing journal from which a citation originates 54 . We acknowledge, of course, that this procedure is not full proof, and papers that are sometimes considered better are published in lower-ranked journals. However, we wanted to use an algorithm that does not require any subjective judgment.

A FAR measurement for a specific event is considered preferable if it comes from a higher SJR publication. For rapid studies conducted by the World Weather Attribution network, there was no recorded SJR as they are not refereed but are done by a large group of specialized climate scientists. Therefore, the average of the SJR impact factor scores for all other studies in the database was used as a rank for WWA studies when comparing them to others. When there are multiple attribution studies for the same event, with the same SJR, the preferred FAR was that with the closet spatial and temporal match to available economic data (as FARs can differ based on temporal and spatial event definition). When the scale is matched closely to economic data, the attribution of the cost will be more accurate.

The final dataset includes 185 events spanning 2000–2019. These events are gathered from 118 event attribution studies, as many attribution studies cover more than one event. Figure  6 depicts the hierarchical criteria applied in choosing the sample of attributed extreme weather events used to determine the FAR for analysis in this study, which is applied to the 2nd level of extreme events in the figure, that is 4864 events with human and/or economic costs recorded in EM-DAT.

Of the whole ‘universe’ of extreme weather events (EWE), 6135 EWE are recorded by the Emergency Management Database (EM-DAT). If these 4864 recorded quantities for damages/deaths, but only 185 were matched with at least one of the 357 papers on attribution included in the CarbonBrief database.

Methodological approach

Allen 3 states that “If [climate change] has trebled the risk over its ‘pre-industrial’ level, then there is a sense in which [climate change] is ‘to blame’ for two-thirds of the current risk….” (p. 891). This framing suggests that if anthropogenic climate change has made an extreme weather event three times more likely, then climate change is responsible for two-thirds of the economic cost caused by the set of similar events. Put differently, two out of each three events of the same class, and with the same calculated FAR, were caused by climate change, while the third would have happened even in a pre-industrial climate. Consequently, for each event ( i ) in the master database, we use Eq. ( 5 ) to estimate that individual event’s climate change-attributed economic cost.

Applying this approach to all events in the master dataset provides an estimation of the climate change-induced costs associated only with this specific list of events. To generate an estimate of the global cost of climate change from extreme weather events, we used the FARs from attribution studies in the dataset we collected and all the economic cost of extreme weather events across 2000–2019 recorded in EM-DAT. The events are limited to heatwaves, floods, droughts, wildfires, and storms, implicitly employing the EM-DAT definition of an extreme event. For inclusion in EM-DAT, an extreme event is one for which at least one of the following three criteria must be fulfilled: (1) 10 or more deaths; (2) 100 or more people affected/injured/homeless; or (3) declaration by the country of a state of emergency and/or an appeal for international assistance.

Two extrapolation methods were used—a global average extrapolation method and a regional average method. The global average extrapolation method relied on obtaining an average FAR for each specific type of event occurring anywhere from the FAR results recorded in the dataset. This event-type average FAR was then multiplied by the economic costs and mortality of all the relevant events in EM-DAT over the 2000–2019 period. The average FARs are calculated from individual attribution studies in the dataset (118 observations) rather than the FARs from the 185 individual events. This is because some studies cover a large number of events. Calculating an average FAR with each event as an individual data point would lead to much greater weight being placed on a smaller number of multiple-event studies.

The regional average extrapolation method was conducted by calculating an average FAR per event type and per continent. This was, similarly, calculated from individual attribution studies rather than events. This regional average FAR was then multiplied by the relevant event-type and region-specific events in the EM-DAT database and subsequently aggregated. This (partial) accounting for differences in how climate systems influence extreme weather across different regions is clearly an advantage of the regional approach. However, there are no, or very few, FAR studies for some event-type and continental combinations. For example, only one study examined a heatwave in Africa, so a regional extrapolation result relies solely on this one study, creating potentially an over-reliance on one modeling approach.

Furthermore, where there are no available attribution studies, for example, on storms in Europe, the global average for that event type is used as a substitute to fill in this data gap. There are significant number of event-type-region combinations for which this global compromise was necessary. The difference between the two methods, therefore, is not as large as it probably should be. In the future, with a more extensive set of attribution study results, it would clearly be preferable to use an approach that distinguishes between types of events, their location (even within continental-size regions), and potentially even their magnitude.

Extreme event attribution data

Of the 185 extreme weather events, the risk of 154 of these events increased because of anthropogenic climate change, while another 24 events were associated with decreased risk, and the risk of the remaining 7 events was unchanged (FAR = 0). These events cover the period from 2000 to 2019. Notably, 77% of these events occurred after 2013, because EEA studies have been conducted increasingly frequently only in recent years. Given the rapid evolution of the EEA methodology, the dominance of more recent studies in our dataset means the FAR records used for the results reflect mostly the higher quality, recent EEA research practices. A significant number of events are recorded for 2015 because of the study by Zhang et al. 56 , which covers a large spatial and temporal scale. With a larger dataset, we would have been able to place greater weight on more recent FARs, to reflect their higher reliability. However, our approach relies on a significant time span (in this case, 20 years) to account for the stochastic component in these event occurrences. If we were to place greater weight on, say, the last 5 years in the dataset, then our results will depend on what actually (stochastically) happened in those 5 years.

The geographical coverage of the matched attribution results included in the dataset is also important to note, as there are significant deficiencies in some regions. The studies cover Africa (10% of the studies), Asia (28%), the Americas (24%), Europe (20%), and Oceania (18%). North and South America are collected here as one grouping because there are very few FARs calculated for South America. The events in South America make up only 5% of the total matched events in the dataset (7 events in the aggregate across all event types). The matched events span 52 different countries. Events in China, the United States, New Zealand, the Philippines, Japan, the United Kingdom, and Australia combined make up over half (54%) of the total dataset. Similarly to the time-series coverage, this is impacted by attribution studies covering multiple events in a defined region, including Zhang et al. 56 that covered 26 cyclones across the Western North Pacific.

By construction, the dataset contains six types of extreme weather events. Notably, 52% of the attribution results in the master database are associated with high-temperature phenomena—heatwaves, droughts, or wildfires (31%, 16% and 4%, respectively). The remainder are either hydrological events–specifically floods (37%) and storms (5%)—or cold waves (6%).

In the initial search, we identified 112 weather events with at least one associated FAR, but which did not have matching economic data. A majority of these (51%) were heatwaves, since the science of attribution is well-established for heat events but measuring the economic impact of heatwaves is challenging and is rarely undertaken. After all, the main impact of heatwaves, aside from mortality, is their indirect losses in the flow of economic activity which are substantially harder to identify and measure than damages (stocks). These under-measured heatwave losses include economic disruptions due to disturbed hydroelectricity distribution, transport failures, ongoing harm to agricultural crop yields and health, and harm to the natural environment 57 , 58 . Moreover, a further 25% of events without economic data are droughts—with the majority occurring in Africa—which is reflective of the geographically uneven distribution of disaster cost records between lower and higher-income regions.

All the events included in the dataset have at least one FAR associated with them. Of the 185 events, 47 have multiple relevant attribution studies. For each, the best FAR was selected based on the spatial and temporal match between the FAR study and the available economic data (see more details in the data collection section). The distribution of FAR attribution results is shown in Fig.  7 . The peak at 0.3–0.4 is predominantly due to flood events—which make up 80% of the attribution results in this range. While 90% of the events with a FAR of between 0.7 and 1 are related to high temperatures, namely heatwaves, droughts, and wildfires. Interestingly, half of the attribution results with a zero or negative FAR are floods, while 32% are, less surprisingly, cold events. The remainder are three drought events with a zero FAR.

This figure shows the number of attribution events for each different FAR range, by the type of weather event.

To allow a global extrapolation of climate change-attributed costs to be made, a global average FAR for each event type has been calculated. On average, 77% of the risk of heatwaves occurring over the study period is due to anthropogenic climate change. Floods show the greatest distribution range and are the only event type where attribution results span both increasing and decreasing risk due to climate change. The global average FAR for floods, however, is 19%. Similarly, as for droughts (44%), wildfires and storms each have a FAR of 60%; however, these are calculated from a small number of data points. Lastly, on average, cold events are calculated as having a decreasing risk (−79%) because of climate change.

An average FAR per-continent per-event type has also been calculated to reflect the lack of uniformity in the global climate system (Fig.  8 ). There were very few, or any, matched attribution results to form the basis of a regional average FARs in many continent-type combinations.

This figure depicts the mean and the range of FARs matched, per weather type and by region. The number on the top of each vertical line represents the number of events for that region-type combination.

Economic data

The following section describes the features of the economic cost data collected regarding the events in the master dataset that is based on EM-DAT—the most comprehensive global database of disaster impacts. EM-DAT defines a disaster event as one that surpasses a clearly defined threshold of damage caused (10 people dead, 100 people affected, an official emergency declaration, or a request for international assistance). As such, what is missing is a large number of small disasters, whose frequency may have risen because of anthropogenic climate change. An alternative database that does aim to capture these high-frequency low-impact events, Desinventar, is not available globally (Desinventar is collected and managed by the United Nations Disaster Risk Reduction Office).

While for disasters more broadly, EM-DAT records deaths, dislocations, people affected, and monetary damages, here we use only the mortality metrics and monetary damage (if these are recorded). The other components include such a diversity of outcomes (from dislocation lasting just a few days to permanent and significant physical and mental injuries), are often inconsistently collected, and are missing for many disaster events, that we have decided to ignore them in our analysis. As such, our aggregation only includes death and damage (as defined earlier).

In the dataset, 114 of the 185 events have mortality estimates. Thirty-nine of these events are responsible for at least 100 deaths each, nine events with more than 1000 lives lost, and four are responsible for the deaths of over 10,000 people—a heatwave in Russia (> 55,000 deaths), a drought in Somalia (20,000 deaths), a heatwave in France (> 19,000 deaths), and a cold event in the United Kingdom (27,500 deaths). The total number of deaths recorded from the events in this dataset is 151,083, equivalent to a statistical value of life lost of US \(\$\) 1.07 trillion (using a VSL of $7.084 million).

Of the 185 events included in the dataset, 115 events have estimates for the economic damages caused. Across these 115 events, the total disaster damages stand at US$ 492.2 billion. The event with the highest damage recorded in this list of events is Hurricane Harvey in the United States in 2017, at US \(\$\) 100.3 billion. Harvey provides a useful example of some of the murkiness in definitions. EM-DAT classifies it as a tropical storm, and indeed the name refers to the hurricane. However, by far most of the damage was caused by the flood which the rainfall that came with the hurricane generated. The attribution papers that analyzed the event focused on the changing likelihood or intensity of the rainfall event, and not on the storm (measured and classified by windspeed). We follow the EM-DAT classifications, since these are available for all events, but note that these distinctions are not always immediately apparent.

Eighty-four of the events have estimated damages greater than US \(\$\) 100 million, and 8 of those are over US \(\$\) 10 billion. A small number of events in the dataset have insured loss estimates associated with them (48 out of 179). These data are heavily skewed to small number of countries, notably the United States, New Zealand, Australia, and Japan, as well as China. The restricted quality and quantity of data collection in low-income countries is one underlying reason for this, but it is also symptomatic of higher insurance penetration rates in high-income countries. Insurance costs from Hurricane Harvey in the United States and Hurricane Maria in Puerto Rico have the highest insurance payouts at US \(\$\) 31.7 billion each ( \(\$\) 30 billion in 2017 US dollars).

For all of these events, indeed for all the events we analyze here, the causes of the damage are complex, and are not just due to the hazard itself. The 2010 heatwave in Russia is a good example. In Moscow, mortality mostly arose from air pollution from peat fires that occurred in the surrounding area. The fires occurred due to a combination of drought and heat, but also the legacy of a Soviet policy of draining bogs, widespread ignition by humans, and confusion and inaction following a new Russian policy shift that had just transferred wildfire control from the national to the regional governments. Our analysis, however, assumes a ceteris paribus world in which all other pre-conditions still exist, but the amount of GHG in the atmosphere is pre-industrial. Clearly, one should view this counterfactual as a thought experiment, rather than a realistic scenario, since without the industrial revolution of the last 150 years, nothing in our world would have been the same.

Reporting summary

Further information on research design is available in the  Nature Portfolio Reporting Summary linked to this article.

Data availability

All the data generated in this study have been deposited in the Harvard Dataverse database [ https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/N3ED1N ]. The data are available under unrestricted access. More description of the data processing in this study are provided in the  Supplementary Information File.

World Meteorological Society (WMO). WMO Atlas of mortality and economic losses from weather, climate, and water extremes (1970–2019) . Retrieved from https://library.wmo.int/doc_num.php?explnum_id=10769 (2021B).

IPCC. Summary for policymakers. In Climate Change 2022: Impacts, Adaptation, and Vulnerability. Contribution of Working Group II to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (eds Pörtner, H.-O. et al.) 1–2272 (Cambridge University Press, 2022) https://doi.org/10.1017/9781009325844 .

Allen, M. Liability for climate change. Nature 421 , 891–892 (2003).

Article   CAS   PubMed   ADS   Google Scholar  

Stott, P., Stone, D. & Allen, M. Human contribution to the European heatwave of 2003. Nature 432 , 610–614 (2004).

Risser, M. D., Paciorek, C. J. & Stone, D. A. Spatially dependent multiple testing under model misspecification, with application to detection of anthropogenic influence on extreme climate events. J. Am. Stat. Assoc. 114 , 61–78 (2019).

Article   MathSciNet   CAS   MATH   Google Scholar  

Otto, F. Attribution of weather and climate events. Annu. Rev. Environ. Resour. 42 , 627–646 (2017).

Article   Google Scholar  

Frame, D., Wehner, M., Noy, I. & Rosier, S. The economic costs of Hurricane Harvey attributable to climate change. Clim. Change 160 , 271–281 (2020).

Article   ADS   Google Scholar  

Smiley, K. et al. Social inequalities in climate change-attributed impacts of Hurricane Harvey. Nat. Commun. 13 , 3418 (2022).

Article   CAS   PubMed   PubMed Central   ADS   Google Scholar  

Noy, I. A global comprehensive measure of the impact of natural hazards and disasters. Glob. Policy 7 , 56–65 (2016).

Serajuddin, U. & Hamadeh, N. New World Bank country classifications by income level: 2020–2021 . Retrieved from https://blogs.worldbank.org/opendata/new-world-bank-country-classifications-income-level-2020-2021 (2020).

Nordhaus W. & Boyer J. Roll the DICE Again: Economic Models of Global Warming (MIT Press, 1999).

Nordhaus, W. Revisiting the social cost of carbon. Proc. Natl. Acad. Sci. USA 114 , 1518–1523 (2017).

World Meteorological Society (WMO). State of the Global Climate 2020 . Retrieved from https://library.wmo.int/doc_num.php?explnum_id=10618 (2021A).

Stern, N. The Economics of Climate Change (UK Treasury, 2007).

Hope, C. The social cost of Co2 from the Page09 Model. Economics Discussion Paper No. 2011-39.

Nordhaus, W. & Sztorc, P. DICE 2013R: Introduction and User’s Manual . Retrieved from: http://www.econ.yale.edu/~nordhaus/homepage/homepage/documents/DICE_Manual_100413r1.pdf (2013).

Waldhoff, S., Anthoff, D., Rose, S. & Tol, R. The marginal damage costs of different greenhouse gases: an application of FUND. Economics 8 , 20140031 (2014).

MimiFUND. MimiFUND.ji . Retrieved from http://www.fund-model.org/MimiFUND.jl/latest/ (2023).

Clarke, B., E, L., Otto, F. & Jones, R. Inventories of extreme weather events and impacts: Implications for loss and damage from and adaptation to climate extremes. Clim. Risk Manag. 32 , 100285 (2021).

Frame, D. et al. Climate change attribution and the economic costs of extreme weather events: a study on damages from extreme rainfall and drought. Clim. Change 162 , 781–797 (2020).

World Weather Attribution (WWA). World Weather Attribution Initiative . Retrieved from https://www.worldweatherattribution.org/about/ (2023).

van Oldenborgh, G. et al. Pathways and pitfalls in extreme event attribution. Clim. Change 166 , 13 (2021).

Stott, P. et al. Attribution of extreme weather and climate‐related events. Wires Clim. Change 7 , 23–41 (2015).

National Academies of Sciences. Attribution of Extreme Weather Events in the Context of Climate Change (The National Academies Press, 2016).

Stone, D. A., Rosier, S. M. & Frame, D. J. The question of life, the universe and event attribution. Nat. Clim. Change 11 , 276–278 (2021).

Dole, R. et al. Was there a basis for anticipating the 2010 Russian heat wave? Geophys. Res. Lett. 38 , L06702 (2011).

Rahmstorf, S. & Coumou, D. Increase of extreme events in a warming world. Proc. Natl. Acad. Sci. USA 108 , 17905–17909 (2011).

Hannart, A., Pearl, J., Otto, F., Naveau, P. & Ghil, M. Causal counterfactual theory for the attribution of weather and climate-related events. Bull. Am. Meteorol. Soc. 97 , 99–110 (2016).

Stone, D. A. et al. The effect of experiment conditioning on estimates of human influence on extreme weather. Weather Clim. Extremes 36 , 100427 (2022).

Angélil, O. et al. On the nonlinearity of spatial scales in extreme weather attribution statements. Clim. Dyn. 50 , 2739–2752 (2018).

Uhe, P. et al. Comparison of methods: attributing the 2014 record European temperatures to human influences. Geophys. Res. Lett. 43 , 8685–8693 (2016).

Perkins-Kirkpatrick, S. et al. On the attribution of the impacts of extreme weather events to anthropogenic climate change. Environ. Res. Lett. 17 , 024009 (2022).

Orlov, A. et al. Global economic responses to heat stress impacts on worker productivity in crop production. Econ. Disasters Clim. Change 5 , 367–390 (2021).

Steffen, W., Hughes, L. & Perkins, S. Heatwaves: hotter, longer, more often. Climate Council Australia . Retrieved from https://www.climatecouncil.org.au/uploads/9901f6614a2cac7b2b888f55b4dff9cc.pdf (2014).

Pascal, M. et al. Evolving heat waves characteristics challenge heat warning systems and prevention plans. Int. J. Biometeorol. 65 , 1683–1694 (2021).

Article   PubMed   PubMed Central   ADS   Google Scholar  

Stuart-Smith, R. F. et al. Filling the evidentiary gap in climate litigation. Nat. Clim. Change 11 , 651–655 (2021).

Lott, F. C. et al. Quantifying the contribution of an individual to making extreme weather events more likely. Environ. Res. Lett. 16 , 104040 (2021).

Wilson, T. & Noy, I. Fifty Years of Peril: A Comprehensive Comparison of the Impact of Terrorism and Disasters Linked to Natural Hazards (Global Policy, 2023).

Mitchell, D. et al. Attributing human mortality during extreme heat waves to anthropogenic climate change. Environ. Res. Lett. 11 , 074006 (2016).

Vicedo-Cabrera, A. M. et al. The burden of heat-related mortality attributable to recent human-induced climate change. Nat. Clim. Change 11 , 492–500 (2021).

Article   CAS   ADS   Google Scholar  

Puvvula, J. et al. Estimating the burden of heat-related illness morbidity attributable to anthropogenic climate change in North Carolina. GeoHealth 6 , e2022GH000636 (2022).

Article   PubMed   PubMed Central   Google Scholar  

Anthoff, D. & Tol, R. J. The uncertainty about the social cost of carbon: a decomposition analysis using fund. Clim. Change 117 , 515–530 (2013).

Diaz, D. & Moore, F. Quantifying the economic risks of climate change. Nat. Clim. Change 7 , 774–782 (2017).

Goodess, C., Hanson, C., Hulme, M. & Osborn, T. Representing climate and extreme weather events in integrated assessment models: a review of existing methods and options for development. Integr. Assess. 4 , 145–171 (2003).

Bouwer, L. Have disaster losses increased due to anthropogenic climate change? Bull. Am. Meteorol. Soc. 92 , 39–46 (2011).

Tol, R. The marginal damage costs of carbon dioxide emissions: an assessment of the uncertainties. Energy Policy 33 , 2064–2074 (2005).

van den Bergh, J. Safe climate policy is affordable—12 reasons. Clim. Change 101 , 339–385 (2009).

Stott, P. et al. Attribution of weather and climate-related events. In Climate Science for Serving Society: Research, Modelling and Prediction Priorities . (eds Asrar, G. R. & Hurrell, J. W.) 307–337 (Springer, 2013).

Jézéquel, A. et al. Behind the veil of extreme event attribution. Clim. Change 149 , 367–383 (2018).

Hammitt, J. Valuing mortality risk:  theory and practice. Environ. Sci. Technol. 34 , 1396–1400 (2000).

Kniesner, T. & Viscusi, W. The value of a statistical life. SSRN https://doi.org/10.2139/ssrn.3379967 (2019).

Department of Transportation (US). Departmental Guidance on Valuation of a Statistical Value of Life in Economic Analysis . Retrieved from https://www.transportation.gov/office-policy/transportation-policy/revised-departmental-guidance-on-valuation-of-a-statistical-life-in-economic-analysis (2021).

Treasury, H. M. The Green Book: Central Government Guidance on Appraisal and Evaluation . Retrieved from https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/938046/The_Green_Book_2020.pdf (2018).

Scimago Research Group. Description of Scimago Journal Rank Indicator . Retrieved from https://www.scimagojr.com/SCImagoJournalRank.pdf (2007).

Viscusi, W. Best estimate selection bias in the value of a statistical life. J. Benefit-Cost. Anal. 9 , 205–246 (2018).

Zhang, W. et al. Influences of natural variability and anthropogenic forcing on the extreme 2015 accumulated cyclone energy in the western North Pacific. Bull. Am. Meteorol. Soc. 97 , S131–S135 (2016).

Disher, B., Edwards, B., Lawler, T. & Radford, D. Measuring heatwaves and their impacts. Air Pollut. Sour. Stati. Health Effects 455–475 (2021).

CarbonBrief. Attributing extreme weather to climate change. Retrieved from https://www.carbonbrief.org/mapped-how-climate-change-affects-extreme-weather-around-the-world (2021).

Download references

Acknowledgements

We are grateful for very useful discussions with Dave Frame, Luke Harrington, Fraser Lott, Fredi Otto, Sue Rosier, Dáithí Stone, Belinda Storey, Eric Strobl, Michael Wehner, and many audiences in seminars and public talks. This research was supported by funding from the Whakahura Endeavour grant.

Author information

Authors and affiliations.

Reserve Bank of New Zealand, 2 The Terrace, Wellington, 6140, New Zealand

Rebecca Newman

Victoria University of Wellington, 33 Bunny St., Wellington, 6011, New Zealand

You can also search for this author in PubMed   Google Scholar

Contributions

I.N. conceptualized the paper. R.N. led the data collection and empirical analysis. R.N. and I.N. jointly wrote the paper.

Corresponding author

Correspondence to Ilan Noy .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Peer review

Peer review information.

Nature Communications thanks Stephane Hallegatte, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary information, peer review file, reporting summary, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Newman, R., Noy, I. The global costs of extreme weather that are attributable to climate change. Nat Commun 14 , 6103 (2023). https://doi.org/10.1038/s41467-023-41888-1

Download citation

Received : 12 October 2022

Accepted : 22 September 2023

Published : 29 September 2023

DOI : https://doi.org/10.1038/s41467-023-41888-1

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Climate mental health — making connections.

Nature Mental Health (2024)

After millions of preventable deaths, climate change must be treated like a health emergency

• Colin J. Carlson

Nature Medicine (2024)

Unveiling the dynamics of sequential extreme precipitation-heatwave compounds in China

• Lijuan Miao

npj Climate and Atmospheric Science (2024)

COP28: ambitions, realities, and future

• Pranay Arora

Environmental Sustainability (2024)

A Systematic Analysis of Systems Approach and Flood Risk Management Research: Trends, Gaps, and Opportunities

• Lum Sonita Awah
• Johanes Amate Belle
• Israel Ropo Orimoloye

International Journal of Disaster Risk Science (2024)

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

• Global Assessment Report (GAR)
• PreventionWeb

Special Events

• Global Platform
• International Day for Disaster Reduction
• World Tsunami Awareness Day
• Sendai Framework Monitor
• Voluntary Commitments

Please help us improve PreventionWeb by taking this brief survey. Your input will allow us to better serve the needs of the DRR community.

See the survey

Four adaptation opportunities Myanmar can take to build resilience to climate change

By Georgina Wade

Myanmar has experienced a number of extreme weather events in recent years, many of which are likely to worsen with climate change. In 2019, the country suffered record pre-monsoon heat. The city of Yangon recorded 42 degrees Celsius in April –  a new record  for the city. Dozens of people were admitted to hospital with symptoms of heatstroke, and seven individuals died in the Yangon and Bago regions due to the  extreme temperatures . In 2008, Cyclonic Storm Nargis caused the  worst natural disaster  in the recorded history of Myanmar. 84,500 people were killed, and 53,800 went missing. In 2015,  severe flooding  affected twelve out of Myanmar’s 14 states, resulting in about 103 deaths and displacing up to a million people.The Earth is  getting warmer , meaning that extreme weather events are likely to become more severe.

Adaptation efforts are essential for Myanmar to build its resilience to climate change, and achieve its development objectives. As climate impacts become more severe, Myanmar cannot afford inaction. The country does have opportunities to build its resilience to climate change, many of which are low-cost and can result in positive co-benefits. Here are four adaptation opportunities that have the potential to make a significant difference to the lives and livelihoods of Myanmar’s population.  

1. Invest in Nature-based Solutions

Nature-based Solutions (NbS) are defined as “actions to protect, sustainably manage, and restore natural or modified ecosystems, that address societal challenges effectively and adaptively, simultaneously providing human well-being and biodiversity benefits”. [1]  NbS have  a variety of co-benefits  that can range from biodiversity conservation, decreases in water runoff, the creation of new jobs, and poverty reduction. For example, Myanmar’s mangroves can protect shorelines from damaging storms and hurricanes by dissipating the energy from waves and winds and preventing erosion by stabilising sediments with their tangled root systems. Mangroves provide  nursery habitats  for many wildlife species, including commercial fish and crustaceans, contributing to sustaining the local abundance of fish and shellfish populations. Restoring these natural habitats provides a valuable opportunity for building resilience and fisher livelihoods.

2. Ensure that new infrastructure is climate-resilient

Climate-resilient infrastructure is planned, designed, built, and operated in a way that anticipates, prepares for, and adapts to changing climate conditions. Myanmar needs to upgrade existing infrastructure to reduce vulnerabilities and adapt to a changing climate and reduce emissions through greener transport systems. Furthermore, ensuring the  resilience of infrastructure  protects them from further climate-related damages. For example, changing the composition of road surfaces can prevent them from deforming in high temperatures.

3. Strengthen community knowledge management

Knowing how to use the tools for change is just as important as having them. There is an urgent need to foster knowledge and design management techniques into planning. Technological innovation and establishing institutional arrangements can help strengthen Myanmar’s capacity to adapt. Providing vulnerable communities with the knowledge, skills, and resources to mitigate the risks of and recover from climate shocks, and stresses will empower them and strengthen their resilience to climate change impacts. Additionally, educating people about different adaptation practices, such as pursuing sustainable farming methods or using new resilience infrastructure, doesn’t require any physical infrastructure to be built, making for adaptation that is low-cost, high impact and long-lasting.

4. Secure finance

Myanmar needs finance and support to adapt to climate change and pursue low-carbon development. Whilst adaptation finance to developing countries is available, it needs to  increase significantly , and improvements need to be made to make it more accessible. At this year’s United Nations Climate Change Conference, discussions over delivering the $100 billion finance target will take place. With Myanmar on track to experience worsening impacts of climate change, the country will need to develop investable adaptation projects that can attract finance from the principal climate finance mechanisms such as the Green Climate Fund (GCF), Adaptation Fund (AF) and Global Environment Facility (GEF).

International co-operation is needed

Climate change has already resulted in loss of life and damage to Myanmar’s economy and put its renowned biodiversity and natural resources under increased pressure. The country’s exposure to climate impacts is high, meaning that swift adaptation efforts are necessary to protect the lives and livelihoods of its population. These efforts need to be made across all parts of Myanmar’s economy and society and become fully integrated into its development planning to ensure its development. Myanmar will need support from the international community to realise these goals, particularly in the form of climate finance, technology and capacity transfers from wealthy countries.

This year’s COP26 will be significant. Coming five years after the Paris Agreement came into force, it represents the deadline for countries to renew their carbon reduction pledges, and increase their ambition towards meeting the 1.5˚C temperature target. For countries like Myanmar, co-operation with other climate-vulnerable nations will be important to ensure they are able to hold wealthy to account, putting pressure on them to meet their obligations and limit the damage that will result from climate change.

Please note: Content is displayed as last posted by a PreventionWeb community member or editor. The views expressed therein are not necessarily those of UNDRR, PreventionWeb, or its sponsors. See our terms of use

Is this page useful?

Thank you. If you have 2 minutes, we would benefit from additional feedback (link opens in a new window) .

Explore historical and projected climate data, climate data by sector, impacts, key vulnerabilities and what adaptation measures are being taken. Explore the overview for a general context of how climate change is affecting Myanmar.

• Climate Change Overview
• Country Summary

Climatology

• Trends & Variability
• Mean Projections (CMIP6)
• Mean Projections (CMIP5)
• Extreme Events
• Historical Natural Hazards
• Sea Level Rise

This page presents Myanmar's climate context for the current climatology, 1991-2020, derived from observed, historical data. Information should be used to build a strong understanding of current climate conditions in order to appreciate future climate scenarios and projected change. You can visualize data for the current climatology through spatial variation, the seasonal cycle, or as a time series. Analysis is available for both annual and seasonal data. Data presentation defaults to national-scale aggregation, however sub-national data aggregations can be accessed by clicking within a country, on a sub-national unit.  Other historical climatologies can be selected from the Time Period dropdown list. 

Observed, historical data is produced by the  Climatic Research Unit (CRU)  of University of East Anglia.  Data  is presented at a 0.5º x 0.5º (50km x 50km) resolution.

Myanmar has a tropical to sub-tropical monsoon climate with three seasons: i) hot, dry inter-monsoonal (mid-February to mid-May); ii) rainy southwest monsoon (mid-May to late October); and iii) cool relatively dry northeast monsoon (late October to mid-February). Climate varies across Myanmar’s different ecological zones, controlled mainly by distance from the coast and altitude. The country’s southern regions in and around the Ayeyarwady Delta and along the Rakhine, Mon, and Tanintharyi coast lines experience a climate typical of Southeast Asia. Temperatures are high and relatively the same all year round and precipitation can be very high ranging between 2,500-5,500 mm per year. These regions experience the highest exposure to tropical cyclones. Myanmar’s central zone is drier (typically 500-1,000 mm of rain per year) and experiences greater temperature variation, but temperatures can still exceed 40°C. Myanmar’s more mountainous regions in the north and east are generally cooler and receive moderate rainfall in the range of 1,000-2,000 mm per year.

• VIEW BY MAP
• VIEW BY LIST

• International edition
• Australia edition
• Europe edition

Cyclone Mocha: three dead and 700 injured as storm pounds Myanmar

About 1,000 people trapped by seawater rescued amid damage to homes, electricity infrastructure and mobile phone masts

Rescuers have evacuated about 1,000 people trapped by seawater 3.6 metres (12ft) deep along western Myanmar’s coast after a powerful cyclone injured hundreds and cut off communications in one of Asia’s least developed countries.

Strong winds injured more than 700 of about 20,000 people who were sheltering in sturdier buildings on the highlands of Sittwe township such as monasteries, pagodas and schools, according to a leader of the Rakhine Youths Philanthropic Association in Sittwe.

Seawater raced into more than 10 low-lying wards near the shore as Cyclone Mocha made landfall in Rakhine state on Sunday afternoon, said the rescue group leader, who asked not to be named due to fear of reprisals from the authorities in the military-run country. Residents moved to roofs and higher floors, while the wind and storm surge prevented immediate rescue.

“After 4pm yesterday the storm weakened a bit but the water did not fall back,” the leader said. “Most of them sat on the roof and at the high places of their houses the whole night. The wind blew all night.”

Water was still about 1.5m (5ft) high in flooded areas on Monday morning but rescues were being made as the wind calmed. The leader asked civil society organisations and authorities to send aid and help evacuate residents.

At least three deaths had been reported earlier in Myanmar, and several injuries were reported in neighbouring Bangladesh, which was spared the predicted direct hit.

Mocha made landfall near Sittwe township with winds blowing up to 209km/h (130mph), Myanmar’s Meteorological Department said. By Monday morning it was downgraded from its severe status and was steadily weakening over land, according to the India Meteorological Department.

High winds crumpled cell phone towers during the day, cutting off communications. In videos collected by local media before communications were lost, deep water raced through streets and wind blew off roofs.

Myanmar’s military information office said the storm had damaged homes, electricity infrastructure, mobile phone masts, boats and lampposts in Sittwe, Kyaukpyu, and Gwa townships. It said the storm also tore roofs off sports facilities on the Coco Islands, about 260 miles (418km) south-west of the country’s largest city, Yangon.

Volunteers previously said shelters in Sittwe did not have enough food after more people arrived there seeking help.

Rakhine-based media reported that streets and the basements of houses in Sittwe’s low-lying areas had been flooded. Rakhine-based media reported that streets were flooded, trapping people in low-lying areas in their homes as worried relatives outside the township appealed for rescue.

More than 4,000 of Sittwe’s 300,000 residents were evacuated to other cities, and more than 20,000 people were sheltering in monasteries, pagodas and schools on higher ground in the city, said Tin Nyein Oo, who is volunteering in Sittwe’s shelters.

Mocha largely spared the Bangladeshi city of Cox’s Bazar, which initially had been in the storm’s predicted path. Authorities had said they evacuated about 1.27 million people before the cyclone veered east.

“The level of risk has reduced to a great extent in Bangladesh,” said Azizur Rahman, director of the country’s meteorological department.

Several deaths were reported as a result of the storm. A rescue team from eastern Shan state announced on its Facebook page that it had recovered the bodies of a couple buried when a landslide hit their house in Tachileik township. Local media reported that a man was crushed to death when a tree fell on him in Pyin Oo Lwin township in central Mandalay region.

Myanmar state television reported that the military government was preparing to send food, medicine and medical personnel to the storm-hit area. After battering Rakhine, the cyclone weakened and was forecast to hit the north-western state of Chin and the central regions on Monday.

Strong winds with rains continued in Saint Martin’s Island in the Bay of Bengal, it was reported, with leading Bengali-language daily Prothom Alo saying about a dozen islanders were injured and around 300 homes destroyed or damaged. One woman was critically wounded, it said.

UN agencies and aid workers in Bangladesh had positioned tons of dry food and dozens of ambulances with mobile medical teams in refugee camps that house more than a million Rohingya people who fled persecution in Myanmar.

In May 2008, Cyclone Nargis hit Myanmar with a storm surge that devastated populated areas around the Irrawaddy River delta. At least 138,000 people died and tens of thousands of homes and other buildings were washed away.

• South and central Asia
• Extreme weather

More on this story

Myanmar hands over junta-backed warlords to China in telecoms scam case

Myanmar junta and armed rebels agree ceasefire

Myanmar on ‘precipice of humanitarian crisis’ after military coup, UN warns

Myanmar fighting at its worst since 2021 coup, says UN

Myanmar: at least 29 people killed in attack on camp for displaced people

Myanmar junta extends state of emergency, forcing delay to elections

Villagers say 14 killed as Myanmar violence flares

Sexual violence is junta’s ‘modus operandi’, Myanmar activist tells UN

Last images taken by Japanese journalist killed in Myanmar in 2007 released

Myanmar airstrike on civilians sparks global outcry as witnesses describe attack

Most viewed.

Content Search

Climate change actions in conflict affected contexts: insights from myanmar after the military coup, attachments.

Helene Maria Kyed & Justine Chambers

Violent conflict and state oppression in Myanmar demonstrates the importance of placing conflict analysis and people-centred approaches at the centre of international programming on climate change and environmental protection.

In 2021, the United Nation’s Intergovernmental Panel on Climate Change (IPCC) warned that the impacts of the climate crisis will be particularly pronounced in poor and conflict-affected countries. Research also identifies climate change as a ‘threat multiplier’ that, in combination with socio-political factors like poverty, state incapacity and inequality, can intensify violent conflict. However, gaps remain in how to address the increase in climate change vulnerabilities in contexts with violent conflict and state oppression. This is evident in Myanmar, where a historically repressive military regime is threatening to cause longer-term ‘climate collapse’.

Since a military coup in February 2021, extractive activities and war economies are destroying the natural environment and placing communities at further risk of displacement, violent persecution and food shortages. These effects of conflict are reducing local people’s capacity to adapt to climate change and threatening civil society’s efforts to protect the environment. Under such conditions, climate change programming needs to place conflict analysis at its centre stage and substitute state-centric and purely technical approaches with people-centred ones, in alignment with the localisation of aid agenda.

* Climate change vulnerabilities in Myanmar

In the 2021 Global Climate Risk Index, out of 183 countries Myanmar is ranked the second most vulnerable to extreme weather events. With more frequent heatwaves, floods, cyclones, droughts and rising sea levels that impact production, food security and land scarcity, climate change poses a severe threat to livelihoods and sustainable development. Myanmar is simultaneously rich in natural resources and home to some of the largest remaining areas of contiguous biodiverse-rich rainforests in Southeast Asia, crucial for global climate stabilisation due to their absorption of carbon dioxide. For generations, indigenous communities have protected these forests using local ecological knowledge systems. However, these systems have been perpetually undermined by top-down conservation interventions, extractive activities and conflict dynamics.

Myanmar provides evidence that climate change vulnerabilities cannot be attributed to global changes in temperatures and weather patterns alone, but also to issues related to governance, natural resource use and conflict. The ability of local communities to mitigate and respond to climate change has been severely hampered by decades of authoritarian rule, agrarian land struggles and long running armed conflicts, which have worsened since the coup.

Escalations since the military coup

Research shows that, since the coup, the military has turned to the country’s vast natural wealth to fund its regime and violent operations. This reinforces a long history of military exploitation that was only partly tempered during a ten-year reform period. Satellite data reveals the depletion of large patches of rainforest since the coup. Civil society organisations (CSOs) also report a rapid increase in unregulated mining, which is polluting waterways, decimating forests, destroying mountains, and causing landslides and changes to fragile ecosystems.

Military-linked militias and businesses are behind much of this mining, but the escalating violent conflict is also fuelling a war economy where other armed actors engage in unregulated resource extraction. These activities are further degrading the environment and accelerating the longer-term impacts of climate change. Another concern is that the military junta’s plan to revive controversial hydropower dams and palm oil plantations will heavily disturb important riverine ecosystems and destroy natural forests, in addition to threatening local land rights and livelihoods.

Prior to the coup, some hydropower dams were stalled due to protests by local communities and environmental defenders. However, the violent reimposition of military rule has drastically undermined the civic space for environmental and climate justice actors, which during the 2011-2020 reform period provided some degree of protection to customary lands and the environment. The military’s brutal crackdown on civil society and environmental activists has also significantly undermined previous efforts to create public climate change awareness and to advocate for equitable climate actions.

Since the coup, regulatory and environmental oversight mechanisms have disappeared, meaning that local communities now have nowhere to take complaints about the effects of extractive projects on their land rights, local environment and livelihoods.

Top-down vs. people centered Top-down technical approaches to climate change typically involve investment in and introduction of agricultural techniques and infrastructures to adjust to climate change, which are developed external to local solutions, knowledge systems and context-specific socio-political relations (e.g. irrigation systems, satellite-based early warning systems, sea walls, drought-resistant crops, new seeds, etc.).

• people-centred approach adheres to the localisation of aid agenda, by involving local people and their knowledge in decision-making and planning of climate change programmes from the outset. This also includes incorporating context-specific understandings of climate change and drivers of vulnerability into programme design and solutions.*

In this context, local efforts to adapt to and mitigate climate change are hampered both by the challenges facing the operations of supportive NGOs and CSOs, and by ongoing violent conflicts and displacements. There is also a high risk that natural disaster relief – in the case of, for instance, cyclones, flooding and drought – will be undermined or be used as an oppressive political tool, with the military preventing humanitarian organisations from helping affected populations.

Pre-coup climate change policies in Myanmar

In the current situation in Myanmar there is an urgent need for international donors to rethink conventional climate change programming. This includes a critical reframing of the policies and approaches that were adopted by the civilian government prior to the military coup, based on international technical assistance, such as the 2019 interlinked Climate Change Policy, Strategy and Master Plan, which aimed to create a climate-resilient and low-carbon society.

While recognising the urgency of climate change actions, earlier policies focused predominantly on support through central government departments and on techno-managerial solutions, with a heavy focus on state regulations. These were by and large apolitical and conflict blind. There was no mention of armed conflicts in the border regions, agrarian land struggles, non-state-controlled areas, or the legacies of authoritarianism, let alone a recognition of how these realities affect the lives of people. Locally-driven climate change adaptation and indigenous natural resource protection were underprioritized in favour of state-centric and top-down solutions. This was evident in the design of several internationally sponsored adaptation projects, some of which were aborted after the coup due to the freezing of aid channelled through government departments. These projects reflected the centrality of technical solutions and involved very little local consultation. They also largely ignored conflict dynamics and failed to target vulnerable populations in areas controlled by non-state ethnic resistance organisations (EROs).

Research also shows that large-scale mitigation projects, like REDD+ ignored local concerns, contributing instead to indigenous communities’ vulnerability and a consolidation of central state power at the expense of local conservation initiatives. These projects also had conflict repercussions. There were some exceptions to this dominant trend, such as international support for community-led conservation initiatives. However, much of the climate-related programming failed to acknowledge the socio-political marginalisation and asymmetric power relations that lie at the root of Myanmar’s protracted conflicts and authoritarian governance structures.

Ways forward and entry points for programming

Since the February 2021 coup, many international donors have withdrawn their state-to-state aid, including for climate change, so as not to legitimise and finance the military regime. Many of the CSO partners of international NGOs have moved their environmental and climate change work underground. Under these conditions, and with the gaps in pre-coup climate change policies, there is an urgent need to adopt more conflict-sensitive, flexible and adaptive programming:

Conflict analysis should be integrated into the design of climate change programmes, with a focus on mapping the power relations, political contestations and pluralism of actors that are implicated both in environmental protection and in natural resource management and extraction. The analysis should be based on in-depth contextual and historically grounded understanding that climate-related challenges are deeply embedded in longer-term ethno-nationalist conflicts and the co-existence of state and non-state legal-institutional arrangements (e.g. for the manangement of land, forests and other natural resources). Particular attention should be paid to EROs like the Karen National Union, which for decades have engaged in natural resource governance in their areas of non-state control.

Localisation of programme implementation is important to ensure that support benefits and reflects the needs of local populations. This requires a shift in programme implementation from top-down, state centric technical solutions towards climate change actions that are people-centred and work from the ground up. Flexible funding and reporting requirements that are adjustable to a volatile and insecure context is important to this approach. Entry points for support could include: a) core costs to secure the continued activities of existing environmental CSOs and indigenous-led networks, and their research and policy advocacy for inclusive and community-led climate change mitigation and adaptation programmes and policies; b) funding for the ongoing documentation of indigenous and customary natural resource management and ecological knowledge systems as a basis for sustainable development; c) support for the documentation of environmentally harmful extractive projects; and d) integration of climate adaptation and environmental protection into humanitarian support to internally displaced people and the communities that host them (e.g. in terms of forestry, green energy and waste management).

Policy-related support to pro-democratic movements in developing climate change policies and initiatives that support sustainable environmental protection and equitable natural resource sharing, land rights and locally embedded solutions. The ongoing drafting process of a federal democratic charter by the National Unity Government (NUG), in collaboration with the National Unity Consultative Council (NUCC) and allied EROs, presents an opportunity to provide technical support within the area of climate change. Informed policy advice should support the inclusion of CSOs that have an existing track record for working with climate change and in-depth experiences with environmental protection and familiarity with indigenous ecological knowledge systems. Funds and technical advice should also be targeted to support these groups to engage in international climate-related forums such as the UN’s Conference of the Parties (COP) to assess progress and add to global conversations on climate-related programming in conflict affected areas.

While these recommendations are specific to the current situation in Myanmar, they also apply more broadly to climate change actions in other conflict-affected and authoritarian states.

Related Content

Myanmar 2024-2026 ifrc network country plan.

Myanmar + 2 more

Myanmar Humanitarian Needs Overview 2023 (January 2023)

Climate, environmental degradation and disaster risk in myanmar: a mimu analytical brief (may 2022), mimu 5w/overview of southeastern myanmar as of 25th march 2022 - bago (east), kayah, kayin, mon, shan (east), shan (south) and tanintharyi - may 2022.