A new UNHCR warning ahead of COP31 puts climate exposure among displaced people in focus, as a separate UN report reveals how far adaptation funding trails developing countries’ needs.

Nearly 90 million forcibly displaced people are living in countries facing high to extreme exposure to climate hazards, according to a UN refugee agency assessment released on 6 November 2026. The estimate, representing around three-quarters of the global population forced from home, presents a stark challenge ahead of the COP31 climate summit in Antalya, Türkiye: people who have escaped violence or persecution may find that the places offering safety are increasingly difficult to inhabit.

UNHCR identifies drought, flooding and extreme heat as major hazards. The assessment does not mean that 90 million people were displaced by climate change alone, or that every person in the group has experienced the same severity of damage. Rather, it measures how widely climate exposure overlaps with existing forced displacement. That distinction matters because UNHCR’s policy guidance treats climate stresses as interacting with conflict, insecurity and other causes of flight, not as an interchangeable explanation for them.

A new emergency after the first flight

People living in camps or temporary settlements often have limited control over where they can build shelter, how easily they can relocate, or whether they can afford water, cooling and flood protection. Exposure to a hazard therefore becomes dangerous through vulnerability as much as through weather itself. A heatwave may threaten a household with dependable electricity very differently from one without a reliable water source or a shaded building.

That vulnerability reaches well beyond the first months of an emergency. UNHCR reports that more than two-thirds of refugees who returned home in the first half of 2026 went to countries highly exposed to climate hazards. A return can end immediate exile without guaranteeing that livelihoods, housing and basic services are secure. The agency warns that repeated shocks could put those returns at risk.

Syria offers a concrete example. UNHCR says nearly four million displaced people have recently returned there, but falling groundwater levels and severe climate stresses threaten efforts to rebuild. Flooding along the Euphrates River in May 2026 damaged homes and forced thousands from their homes again. These figures demonstrate why measuring returns alone cannot capture whether communities have the resources to remain.

Heat, floods and the risk of repeated displacement

The danger is geographically uneven. In sub-Saharan Africa, UNHCR projects that by 2050 one in three refugee settlements could face at least 100 days of dangerous heat each year. More than half the locations it assessed could see at least 50 additional dangerous-heat days annually. In some Ethiopian settlements, including Helaweyn, the projected figure exceeds 200 days a year. These are scenarios about future conditions, not observations of temperatures already recorded in 2026.

Flood risks carry a different warning. UNHCR estimates weather-related disasters triggered more than 255 million internal displacements over the last decade, an average of roughly 70,000 movements a day. These are displacement events rather than necessarily 255 million distinct individuals; one person can be uprooted more than once. The figure illustrates the scale of mobility caused by disasters without equating it to the separate count of refugees and others displaced by conflict.

In East Africa, the agency says severe flooding associated with a possible very strong El Niño could displace between 770,000 and three million people across Somalia, Kenya, South Sudan and Uganda before the end of 2026, depending on how conditions develop. The upper end is a conditional risk estimate, not a prediction that three million people will inevitably move. UNHCR points to severe floods in 2023, when more than two million people were displaced in the region, as a reminder of the potential consequences.

The pressure extends to people without legally recognised nationality. Nearly 90% of the world’s 4.7 million reported stateless people live in countries with high to extreme exposure to climate hazards, according to UNHCR. Lack of documents or legal recognition can limit access to relief and other public services just when a disaster makes them most necessary. The agency frames inclusion in emergency systems as a protection issue, not simply an environmental one.

The finance gap confronting COP31

The 6 November UNHCR release says fragile and conflict-affected countries receive only about 10% of global climate finance, despite hosting many displaced people. That is a statement about the distribution of climate funding across countries, not a claim that refugees themselves receive 10% of funds. Financing mechanisms often operate through national planning, institutions and project pipelines that are difficult to access in unstable environments.

A separate UN Environment Programme assessment published on 4 November underscores the size of the wider problem. UNEP estimates that developing countries will need about $305 billion annually for adaptation by 2035, while international public adaptation finance from developed to developing countries totalled only $28.2 billion to $31.7 billion in 2024. The figures concern different years and the needs of developing countries as a whole, rather than a refugee-specific funding calculation.

Even achieving the aspiration to triple international public adaptation finance by 2035 would meet only 20–29% of estimated needs, UNEP says. The report also finds that flows increased by about $1 billion between 2023 and 2024, a slow pace compared with the scale of planned work. More money is essential, but it is not enough if vulnerable communities are excluded from decisions or funds cannot be converted into completed infrastructure.

The 2026 Adaptation Gap Report also shows the challenge is broader than preventive investment: initial requests to the Fund for responding to Loss and Damage amounted to $2.8 billion against an opening funding envelope of $342 million. Adaptation aims to reduce future harm, while loss-and-damage support responds to impacts that still occur; the two should not be merged into one accounting category. Both face demands that exceed available resources.

What protection looks like on the ground

UNHCR identifies practical responses already being pursued with host communities: solar-powered water systems in the Sahel, climate-resilient agriculture in Uganda, and land and mangrove restoration in Colombia. These are not substitutes for diplomacy or for reducing greenhouse-gas emissions. They are examples of how adaptation can improve security and livelihoods in places where displaced people and their neighbours rely on the same water, land and infrastructure.

The agency’s water, sanitation and hygiene programme demonstrates the tension between rapid relief and durable services. In a sudden crisis, emergency water delivery can save lives. Over time, maintaining groundwater sources, protecting pumping systems and integrating settlements into local supply networks can provide a stronger foundation. Planning has to account for host populations, which may already face shortages, and for the recurrent costs of operating infrastructure after initial funding has been spent.

Early-warning alerts, accessible shelters and disaster-preparedness programmes must also account for legal status, language, mobility and access to information. An alert that never reaches people in an informal settlement cannot avert danger. National adaptation plans that count displaced communities and include local authorities may be better placed to connect international finance with conditions on the ground; that is a policy implication of UNHCR’s warning rather than an announced COP31 agreement.

What governments will be tested on

COP31 is scheduled to convene in Antalya from 9 to 20 November. UNHCR is asking governments and climate-finance institutions to include displaced people in national climate plans, direct long-term resources to fragile areas, and support locally led solutions. The test will not be whether delegates acknowledge the crisis in principle. It will be whether applications, eligibility rules and funding decisions make those commitments reachable for people who may have little political representation.

Useful measures would include counting displacement-affected communities in national risk assessments, tracking whether adaptation money reaches refugee-hosting districts, and making service improvements verifiable over time. Such indicators cannot erase the causes of war or eliminate climate hazards, but they can distinguish a pledge from protection delivered. They would also help identify where a project benefits hosts and displaced residents together instead of shifting scarcity from one group to another.

The central message of the new UNHCR assessment is therefore not that climate change has replaced conflict as a driver of displacement. It is that safety after flight is increasingly conditional: on the water people can reach, the shelter they can maintain, and the institutions able to prepare for the next shock. With COP31 approaching, the gap between those basic needs and the finance available to meet them has become a defining international test.