Geospatial for good · UNICEF · climate risk
UNICEF's Children's Climate Risk Report 2026 is a mapping project. Nearly every child on Earth now sits inside at least one climate hazard, and for the first time we can see exactly where.
Most climate reporting answers how bad the situation is. The harder question for anyone planning a response is where, and which children are most at risk. A national average hides the village on the floodplain and the school with no shade. UNICEF's new report tackles that question with the tools many of us in the geospatial community use every day: gridded population data, hazard layers, and spatial overlay analysis, applied at a resolution fine enough to act on.
About 2.3 billion children, almost every child alive, live somewhere exposed to at least one climate hazard. Nearly half are exposed to three or more at once, and a small but growing group faces as many as six overlapping threats in the same place. What makes the 2026 edition different from previous climate-and-children reports is not the alarm, it is the geography behind it.
How the map was built
The analysis rests on the new Global Child Hazard Database, a UNICEF-led, open dataset that combines child-specific exposure to eight climate hazards, riverine and coastal floods, droughts, tropical storms, heatwaves, extreme heat, fires, and sand and dust storms, plus two climate-sensitive hazards, malaria and air pollution. The data ships in both tabular and spatial formats, which means it slots straight into a GIS workflow rather than living inside a PDF.
The piece that matters most for spatial analysts is the baseline. Earlier exposure studies counted total population and assumed children were distributed evenly. This report instead uses WorldPop's new Children's Population Layer, described as the first global, high-resolution gridded dataset of where people under 18 actually live. Pair a child-population raster with a hazard raster and you can measure child exposure directly, by grid cell, instead of inferring it from a national headcount.
Exposure is then counted down to roughly a 100 metre grid and aggregated up to whatever administrative level a decision-maker needs. That is the move that turns a global statistic into something operational: a district health office can pull the cells inside its own boundary and see how many children sit under one hazard, two, or five. A "multi-hazard count" per cell flags the places where threats stack on top of each other, the hotspots that rarely show up in country-level tables.
The report is unusually candid about its own limits, which is worth flagging for a technical readership. Some hazard layers are coarse, parts of the heat data sit near a 30 km cell size, and malaria is modelled from reported cases rather than full climatic vulnerability. Where ground stations are sparse, such as Papua New Guinea, there is little to calibrate satellite estimates against. None of this sinks the analysis; it just sets the confidence with which each layer should be read, and points to where better open data would help most.
Single-hazard exposure · children worldwide
Each tile is one hazard layer from the database. The figure is the number of children living in exposed areas; the share is their portion of all children.
Source: UNICEF Children's Climate Risk Report 2026
Multi-hazard overlap · grid-level analysis
Counting hazards one at a time misses the real threat: the same cell sitting under several at once. Because the database resolves exposure to a fine grid, it can count how many children face at least two, three, four hazards in the same place. Select a tier to see what stacks up there.
From exposure to risk
Exposure tells you where the extreme weather is dangerous. It does not tell you where a child is most likely to be harmed by it. The report's Children's Climate Risk Analysis closes that gap by pairing each place's hazard exposure with child vulnerability, built from access to health care, water and sanitation, nutrition, education, child protection and social protection. A moderate flood is survivable with a working clinic and clean water nearby; the same flood in a place without them can be catastrophic.
Plot the two against each other and every country lands in one of four quadrants. The one that should drive spending sits top-right: high hazard and high vulnerability, where the danger is greatest and the capacity to cope is weakest. This is a classic spatial decision matrix, and it reframes the problem from "where is the climate worst" to "where will it do the most damage to children."
Explore the data · GeoSight
UNICEF publishes the analysis on GeoSight, its open geospatial platform, so anyone can switch hazards, zoom to a country and read the underlying grid. For the geospatial community this is the real story: not another static infographic, but a live dataset built for planning and budgeting.
What the data is for
A dataset earns the phrase "for good" by what it changes on the ground. UNICEF designed the database as a decision tool, and the intended uses read like a brief for any applied geospatial team:
Why the geography matters
This is not just data on a map; the consequences are already real. The overlapping hazards described here are already reshaping children's lives, which is why pinning down exact locations isn't just an academic exercise.
That first sentence is a statement of human responsibility. The second is a matter of geography, and it marks a novelty. For most of the history of this work, saying "we know exactly where these children are" was just a hopeful wish. A child-specific population layer, a stack of open hazard rasters and a public platform to query them turn it into something a planner can act on this budget cycle.
It is also a reminder of where this community can be useful. The report's honest notes on coarse cells, sparse calibration data and missing hazard layers are an open invitation: better gridded population estimates, sharper remote-sensing inputs, and more open hazard data all translate, fairly directly, into children counted who would otherwise have been missed. Mapping where the children are was the hard first step. Keeping that map accurate is ongoing work, and a good use of the tools we already have.