How many people in Ghana live where rivers can flood?
About 1.18 million people in Ghana live where a 1-in-100-year river flood is modelled to reach, and about 421,700 of them are children. Keta, Ketu South and South Tongu have the most. The count covers river flooding only, so towns that flood when drains overflow are not in it.
- 60,000+
- 20,000–60,000
- 5,000–20,000
- 1,000–5,000
- under 1,000
How to read this: the brighter the district, the more people live inside the modelled 1-in-100-year river flood. A dark district is not a safe one: drainage flooding in towns is not modelled.
- people live inside the modelled 1-in-100-year river flood (median of 1,000 samples)
- 1.18million
- of them are children aged 0–14
- 421,753
- are aged 65 or over: about a fifth more than their share of Ghana’s population would suggest
- 54,151
- live in eight districts on the lower Volta and the Keta lagoon (derived from the model)
- 48%
Which districts have the most people living in the flood zone?
Keta, with about 112,300 people inside the modelled 1-in-100-year river flood, then Ketu South with 101,800 and South Tongu with 89,100. All ten districts with the most people exposed are on the lower Volta, the Keta lagoon or the edge of Greater Accra, and together they hold over half the national total. derived from the model
I counted the people living in every 90 m grid cell that the flood maps show under water, district by district, for five sizes of flood. The 1-in-100-year flood is the one plans most often use, so the figures here are for that flood unless I say otherwise.
The ten districts in the table hold about 665,500 people in the flood zone, 56% of the national total derived from the model. Every one of them is also an Act district, in the national top 20 for modelled damage to homes under every assumption I tested.
Keta has the most older people in the flood zone, about 8,300; Ketu South has the most children, about 37,400.
People in the river flood zone are concentrated on the lower Volta and around Accra
How to read this: each colour is a band of people living inside the modelled 1-in-100-year flood, as in the key. Tap a district, or a row of the table below, to read its figures; “Accra to Keta” zooms in, and the switch compares people with damage.
Modelled river flooding for all 260 districts. People are 2026 population estimates living inside the flood extent as mapped.
Sources
My independent consultancy (2025–26), also my Oxford MSc dissertation. Hazard: JRC global river flood hazard maps v2 (CC BY 4.0). People: WorldPop 2026 (CC BY 4.0) and the Ghana Statistical Service 2021 census. Boundaries: OCHA/HDX COD-AB (Ghana Statistical Service).
| District | Region | People in the 1-in-100-year flood extent | Children aged 0–14 | People aged 65+ | National rank for damage across 9 combinations | Class |
|---|---|---|---|---|---|---|
| Keta Municipal | Volta | 112,294 | 35,617 | 8,269 | 1st–3rd | Act |
| Ketu South | Volta | 101,768 | 37,410 | 4,647 | 4th–6th | Act |
| South Tongu | Volta | 89,113 | 31,788 | 4,978 | 2nd–4th | Act |
| North Tongu | Volta | 66,672 | 24,862 | 3,574 | 7th–8th | Act |
| Anloga | Volta | 63,481 | 20,119 | 4,680 | 4th–7th | Act |
| Ada East | Greater Accra | 63,149 | 24,339 | 2,670 | 5th–6th | Act |
| Weija Gbawe Municipal | Greater Accra | 51,164 | 16,126 | 1,542 | 1st–11th | Act |
| Ningo/Prampram | Greater Accra | 47,692 | 16,566 | 1,506 | 1st–3rd | Act |
| Ada West | Greater Accra | 42,492 | 16,406 | 1,785 | 8th–12th | Act |
| Central Tongu | Volta | 27,705 | 10,306 | 1,491 | 9th–17th | Act |
People, children and people aged 65+: medians of 1,000 samples, 1-in-100-year flood, flood extent as mapped. Rank: best and worst national rank for modelled damage to homes across nine combinations of assumptions. All 26 Act and Investigate districts: flood-prone districts in Ghana.
How many children and older people live in Ghana’s flood zones?
About 421,700 children aged 0–14 and 54,200 people aged 65 or over live inside the modelled 1-in-100-year river flood. Children are there roughly in proportion to their share of Ghana’s population. Older people are over-represented: 4.6% of those exposed against 3.8% of the population, about a fifth more.
Both groups matter for the same practical reason. They are the people who most need help to leave before a flood and who take longest to recover after one. A warning plan for Keta or South Tongu has to reach them.
The counts of children and older people come from WorldPop’s 2026 population estimates by age group and the Ghana Statistical Service’s 2021 census. Like the totals, they are counted on the 90 m grid, so they are sound for a district and rough for a single street.
| Group | Inside the 1-in-100-year flood extent (median) | Range, P5–P95 | Share of those exposed | Share of Ghana’s population |
|---|---|---|---|---|
| Everyone | 1,180,242 | 1,156,143–1,204,389 | 100% | – |
| Children aged 0–14 | 421,753 | 410,990–432,878 | 35.7% | 35.2% |
| People aged 65+ | 54,151 | 49,288–59,210 | 4.6% | 3.8% |
How sure is the figure of 1.18 million?
The population data move it very little: across 1,000 samples it stays between 1.16 and 1.20 million. Where the flood edge sits moves it a lot. Pull the modelled edge in by two map cells and the count falls to about 608,000; push it out by one cell and it rises to about 1.78 million.
Global flood maps are drawn on a 90 m grid, and many towns sit right at the flood edge. One cell either way decides whether a whole neighbourhood is in or out. I set the narrower and wider edges from a real test: the White Volta flood of September 2020, where the model’s misses and over-reaches were mostly within a few hundred metres.
The size of the flood matters less than you might expect. Going from a 1-in-10-year flood to a 1-in-500-year flood adds about 469,000 people derived from the model. Moving the flood edge by one or two cells changes the 1-in-100-year count by about 570,000 to 600,000 either way.
Where the flood edge sits changes the count more than the size of the flood
How to read this: each bar is one size of flood, from the frequent 1 in 10 years to the rare 1 in 500, in thousands of people. Switch the flood edge and watch every bar jump; then compare the top bar with the bottom one.
People inside the modelled river flood across Ghana; medians of 1,000 samples.
Show the data as a table
| Flood | Narrower edge, thousands | Edge as mapped, thousands | Wider edge, thousands |
|---|---|---|---|
| 1 in 10 years | 365.8 | 872.1 | 1,420.0 |
| 1 in 50 years | 558.0 | 1,106.0 | 1,694.0 |
| 1 in 100 years | 607.9 | 1,180.2 | 1,778.4 |
| 1 in 200 years | 649.8 | 1,249.4 | 1,862.8 |
| 1 in 500 years | 702.1 | 1,341.1 | 1,982.6 |
Why does the north matter more for people than for money?
Because the model values homes by their size and by night-time light, and homes in the poorer north carry less value. Mamprugu Moagduri in North East Region ranks 32nd in Ghana for people in the 1-in-100-year flood but 94th for damage in that flood. A ranking by money alone would not reach it.
The five northern regions hold about 76,700 people inside the 1-in-100-year flood, around 34,000 of them children, but only about 1% of the modelled damage to homes derived from the model. That gap is not a flaw in the data. It is what a damage ranking measures: the cost of repairs, which is lower where homes are cheaper to build.
The reverse happens near Accra. Ga South Municipal ranks 32nd for damage in a 1-in-100-year flood but only 73rd for people: fewer people, in homes that cost more to repair.
The two rankings agree at the top: the same ten districts lead both, in a different order. Keta comes first for people, Ningo/Prampram first for damage. Further down they part, and that is where a people-based criterion changes the list.
| District | Region | Pattern | People in the 1-in-100-year flood | Rank for people | Damage to homes in that flood, GHS m | Rank for damage |
|---|---|---|---|---|---|---|
| Mamprugu Moagduri | North East | Ranks higher for people | 6,496 | 32nd | 14.9 | 94th |
| Chereponi | North East | Ranks higher for people | 2,170 | 68th | 6.8 | 118th |
| Sene East | Bono East | Ranks higher for people | 2,705 | 61st | 8.9 | 110th |
| West Mamprusi Municipal | North East | Ranks higher for people | 4,600 | 41st | 18.8 | 88th |
| North East Gonja | Savannah | Ranks higher for people | 2,018 | 71st | 7.8 | 115th |
| New Juaben South Municipal | Eastern | Ranks higher for damage | 361 | 143rd | 41.4 | 67th |
| Awutu Senya East | Central | Ranks higher for damage | 756 | 107th | 70.3 | 51st |
| Ga Central Municipal | Greater Accra | Ranks higher for damage | 1,369 | 88th | 202.5 | 36th |
| Yilo Krobo | Eastern | Ranks higher for damage | 1,058 | 97th | 61.1 | 53rd |
| Ga South Municipal | Greater Accra | Ranks higher for damage | 1,979 | 73rd | 260.6 | 32nd |
Damage here is for one 1-in-100-year flood, not the yearly average used for the Act, Investigate and Monitor classes. Flood extent as mapped; medians of 1,000 samples.
How does this compare with other estimates?
It is in the same range. The Pacific Disaster Center’s 2024 baseline assessment for Ghana counts 887,111 people, 3% of the population, exposed to riverine flooding. My estimate is 1.18 million, about one Ghanaian in thirty, with different flood maps and population data. derived from the model
People living in a modelled flood zone are not the same as people affected by a real flood. After the floods of 29 June 2026, the Interior Minister reported 91,989 people affected across 77 districts, most of them in Greater Accra and Central regions (MyJoyOnline, 24 August 2026). The two numbers measure different things, and events like that one often include rain-fed flooding in towns, which a river-only screen leaves out.
Reported impacts and modelled exposure answer different questions. The first says what happened. The second says where the next river flood is most likely to find people, so preparedness can start there.
What do these numbers leave out?
River flooding onlyDrainage flooding in towns, the sea and dam releases are not in the count. Coastal people figures exist only as a preliminary run and are not published.
A zero is not “no risk”37 districts show nobody in the modelled river flood, Kumasi Metropolitan among them, and Accra Metropolis shows only 66. In both cities heavy rain and blocked drains cause floods the river model does not see.
Where people live is untestedThe uncertainty runs rescale the population but never move anyone, so errors in where homes are on the grid are not in the range.
Screening, not street levelThe 90 m grid is right for comparing districts, not for saying whether one family’s home floods.
Not officialIndependent consultancy, also my Oxford MSc dissertation, presented for discussion with the World Bank. Not official figures of the Government of Ghana or the World Bank.
Questions people ask
Are 1.18 million people flooded every year?
No. 1.18 million is the number of people whose homes sit inside the area a 1-in-100-year river flood is modelled to reach. A flood that size has a 1% chance in any year, and real floods rarely hit every river at once.
Does the count include people at risk from the sea?
No. The coastal screen maps where the sea could reach, but its people figures come from a preliminary run and are not published yet. Ghana’s coast to 2050.
Why does Accra Metropolis show so few people?
Because the model covers river flooding, and most flooding in central Accra comes from heavy rain overwhelming drains. Accra Metropolis shows only 66 people in the modelled river flood. Weija Gbawe and Ningo/Prampram rank high because they sit beside rivers and lagoons.
Where do the population figures come from?
From WorldPop’s 2026 gridded population estimates, about 35.4 million people in all, including its estimates by age group, and from the Ghana Statistical Service’s 2021 census.
What does a 1-in-100-year flood mean?
A flood with a 1% chance of happening in any year. It can happen two years running; over 30 years the chance of at least one is about one in four.
Can I get the figures for my district?
Tap your district on the map above, or choose it from the list, to read its people and damage figures. The 26 Act and Investigate districts are in plain tables on flood-prone districts in Ghana. For a closer look at one district, get in touch.
Sources
- Nuhu, H. (2026). National fluvial flood-risk assessment and priority stability under uncertainty in Ghana. MSc dissertation, University of Oxford. Independent consultancy, presented for discussion with the World Bank. Method and results.
- WorldPop, University of Southampton. Gridded population and age structures, 2026 (CC BY 4.0). worldpop.org.
- Ghana Statistical Service. 2021 Population and Housing Census: district age and functional-difficulty tables.
- European Commission Joint Research Centre. Global river flood hazard maps, version 2 (CC BY 4.0). JRC data catalogue.
- Google Research. Open Buildings v3 building footprints. Open Buildings.
- Ghana Statistical Service / OCHA. Ghana subnational administrative boundaries (COD-AB). Humanitarian Data Exchange.
- Pacific Disaster Center (2024). National Disaster Preparedness Baseline Assessment: Ghana. Report.
- MyJoyOnline (24 August 2026). June 29 floods killed 39, affected nearly 92,000 people. Article.
Work with me
Need to know who lives in the flood zone in your district or programme area?
I can count the people exposed, children and older people included, test how far the count moves, and set it beside the damage ranking. Remote or in-country.