Coastal extreme water levels and inundation screening
On Ghana’s 566 km coast, the 1-in-100-year sea level, with tide, storm surge and waves added together, is 1.06–1.20 m above model mean sea level. I worked it out for 57 sections, then mapped the low ground connected to the sea: 132.3 km² today, 182.4 km² with a +0.25 m sea-level scenario for 2050. It is a screening, not a 2D model.
- More than 1 km² of low ground connected to the sea at 1 in 100, today
How to read this: the coast from Accra to Keta. The outlined districts, from the Densu delta at Weija to the Volta at Ada and Anloga, each have more than 1 km² of low ground the sea can reach at today’s 1-in-100-year level. Static screening with provisional forcing, not a 2D flood model.
- sections of Ghana’s coast, about 10 km each, each with its own 32-year hourly water-level record
- 57
- highest 1-in-100-year sea level along the coast (lowest 1.06 m), above model mean sea level
- 1.20m
- of that level added by breaking waves (wave setup), depending on the section; the lower figure is 0.09 m
- up to0.16m
- of low ground connected to the sea at 1 in 100 today; 182.4 km² with a +0.25 m sea-level scenario for 2050
- 132.3km²
About the author. Hisham Nuhu is a UK-based flood and disaster risk specialist. He worked at England’s Environment Agency from 2019 to 2025 and holds an MSc in Water Science, Policy and Management from the University of Oxford. He works remotely or in-country with governments, development banks, UN agencies and consulting firms, and in England with councils, communities, landowners and developers.
How high can the sea reach on this coast?
Add tide, storm surge and wave setup hour by hour, then fit an extreme-value curve to the storm peaks. For Ghana I built a 32-year hourly record (1993–2024) for each of 57 sections. The 1-in-100-year level is 1.06–1.20 m above model mean sea level, and waves add 0.09–0.16 m of it.
Tide and surge come from the Global Tide and Surge Model, waves from ERA5. Before relying on the model I checked it against the Takoradi tide gauge, hour by hour. The two agree closely (r = 0.94), though in the biggest storms the model runs about 0.17 m low.
Waves matter because breaking waves pile water up against the beach. I added them hour by hour, before the extreme-value step, so they only count when they coincide with high water. Each return level carries a 5–95% band from 500 bootstrap resamples.
Along most of the coast the 1-in-100-year level varies by little more than a tenth of a metre. What varies far more is the land behind it: low-lying lagoons and deltas flood at levels a steeper coast shrugs off.
The sea level barely changes along the coast; the low ground behind it changes a lot
How to read this: west to east along the coast. Top bars: the 1-in-100-year sea level in each 10 km section. Bottom bars: the low ground connected to the sea there. Switch the sea level, or tap the chart, to move along the coast.
Static connected-inundation screening at 30 m with provisional forcing. Hazard only: coastal people and damage are not shown.
Sources
Water levels: GTSM v3 hourly reanalysis; waves: ERA5 (Copernicus Climate Change Service); terrain: FABDEM v1-2 (CC BY-NC-SA 4.0). Extreme-value analysis, sections and mapping: my own work (2026).
| Scenario | Sea level added | Low ground connected to the sea at 1 in 100 years, km² |
|---|---|---|
| Today | 0 m | 132.3 |
| 2050 | +0.25 m | 182.4 |
| Stress test | +0.75 m (0.25 m sea-level rise + 0.5 m wave stress test) | 963.8 |
Is a screening model enough, or do we need a 2D model?
Screening is enough to choose where to look. A 2D model is needed to design defences or set warning levels. A static screen shows which low ground the sea can reach. It cannot show how fast water arrives, how long it stays, or what defences and drainage do. Screen the whole coast first, then model the priority sites.
Even a screen has choices that change the answer. A simple “bathtub” floods every low spot, including lagoon basins with no path to the sea: nationally 709.8 km² at 1 in 100. Keeping only land connected to the sea gives 132.3 km².
Resolution matters as much. In the Volta delta window, the same connected flooding covers 87 km² at 30 m, 97 km² of 90 m cells and 239 km² of 1 km cells. Mapped at 1 km it looks about 2.7 times bigger than it is.
A full hydrodynamic model of 566 km of coast would cost more than the decision it supports at national scale. Match the method to the decision.
How is sea-level rise included?
Add it to the extreme level and map again. I used +0.25 m for 2050, then a further +0.5 m as a stress test, not a forecast. Connected low ground at 1 in 100 grows from 132.3 km² today to 182.4 km² in 2050, and to 963.8 km² under the stress test, as the Keta lagoon connects.
Today most of the connected low ground lies around the Songor lagoon at Ada and the Densu delta at Weija. Switch Fig. 1 to 2050 and the same places grow; switch to the stress test and the Keta lagoon fills.
The jump under the stress test says more about the static method than about the future: once the lagoon connects in a static model, all of it floods. A 2D model would show how much water could actually get in during a storm.
Show connected low ground by return period and scenario
| Sea level | Today | 2050 (+0.25 m) | Stress test (+0.75 m) |
|---|---|---|---|
| 1 in 1 year | 119.8 | 153.5 | 909.5 |
| 1 in 10 years | 128.0 | 173.3 | 944.8 |
| 1 in 50 years | 131.2 | 178.7 | 959.3 |
| 1 in 100 years | 132.3 | 182.4 | 963.8 |
What happens after the screening?
Pick the priority sites and model them in 2D. For Ghana those are the Keta lagoon, the Songor lagoon at Ada and the Densu delta at Weija. A 2D model at 10–30 m, driven by the same hourly tide, surge and waves, with surveyed defences and lagoon inlets, gives design levels and warning triggers. That stage has not started.
From a national screen to design levels: only the first stage is built
National screening: built
57 sections; hourly tide, surge and waves; extreme-value analysis; connected flooding at 30 m for three scenarios.
Priority sites: next
Where today’s and 2050 flooding concentrate. Ghana: Keta lagoon, Songor lagoon at Ada, Densu delta at Weija.
2D hydrodynamic models: next
TUFLOW- or SFINCS-type models at 10–30 m, with surveyed defences, lagoon inlets and drainage.
Design levels and triggers: next
Design levels with sea-level allowances for investment, and trigger levels for coastal early warning.
How to read this: the stages light up in turn; tap, or tab to, any stage to read it. Only stage 1 has been carried out for Ghana.
I build full 2D hydrodynamic models at 10–30 m resolution. I used TUFLOW and Flood Modeller 2D at Hamstall; HEC-RAS 2D and SFINCS are other options for river and coastal work. My completed 2D work is on rivers: river hydrology and 1D–2D modelling.
What does connected-inundation screening not show?
It shows where low ground is connected to the sea at a given still-water level. It does not show flow, timing, overtopping, wave run-up, breaching or drainage. In Ghana the forcing for all 57 sections is still provisional. Treat the result as a map of where to look, not of what will flood in the next storm.
Provisional forcingAll 57 section assignments await review; 43 are flagged because the nearest model point is more than 15 km away.
No datum reconciliationThe model’s mean sea level, the tide-gauge zero and the terrain datum have not been reconciled, so absolute levels carry an offset uncertainty.
Hazard onlyThe screen maps flooded area. Coastal people and damage figures are preliminary and not published.
What data and licences does coastal screening need?
Hourly tide and surge, hourly waves, a tide gauge to check them against, and good terrain. For Ghana: GTSM and ERA5 from the Copernicus Climate Change Service, the Takoradi record from the University of Hawaii Sea Level Center, and FABDEM 30 m terrain. FABDEM is licensed CC BY-NC-SA 4.0; commercial use needs a FABDEM licence.
Terrain is a common weak point in coastal screening. Raw satellite elevation models include trees and roofs, which hide low ground. FABDEM removes them, which is why I used it. Coastal outputs from it can be shown in part, as on this site; a commercial project needs its own licence.
Questions people ask
Is the Ghana coastal work a 2D model?
No. It is a static screening of the low ground connected to the sea, with provisional forcing. Its job is to find the places that deserve a 2D hydrodynamic model, which is the next stage.
What is wave setup?
The rise in water level at the shore caused by breaking waves. On Ghana’s coast it adds 0.09–0.16 m to the 1-in-100-year level, depending on the section.
Are people and damage figures available for Ghana’s coast?
Not yet. Coastal exposure has been run only at a preliminary stage, so the published results are hazard only: sea levels and flooded area.
Which parts of Ghana’s coast flood first?
In the screen, the Songor lagoon at Ada and the Densu delta at Weija today, with the Keta lagoon joining under the stress test. More in Ghana coastal flooding and sea-level rise.
Sources
- Copernicus Climate Change Service. Global sea level change time series: hourly tide and surge from the Global Tide and Surge Model (GTSM v3).
- Copernicus Climate Change Service. ERA5 hourly data on single levels: wave height and period.
- University of Hawaii Sea Level Center. Research-quality hourly tide gauge data, Takoradi.
- Muis, S. et al. (2020). A high-resolution global dataset of extreme sea levels, tides, and storm surges, including future projections. Frontiers in Marine Science 7, 263. doi:10.3389/fmars.2020.00263.
- Fathom and University of Bristol. FABDEM v1-2, 30 m terrain (CC BY-NC-SA 4.0).
- Deltares. SFINCS; TUFLOW: 2D models for the next stage.
- Ghana Statistical Service and OCHA. Ghana subnational administrative boundaries (COD-AB).
Work with me
Need to know how high the sea can reach on your coast?
Send the terms of reference or the stretch of coast. I will reply with how I would screen it, which sites would need 2D models and when I am available, remotely or in-country.