The heaviest rain since Ghana began keeping records fell in June 2026 on a city that has drowned in the same streets for sixty-six years. The problem has never been that our engineers lacked answers. It is that we kept asking them the wrong question.
AZOSKA SIMEONE, & ENGR. WORMENOR EMMANUEL
When the rains came to Accra
On 18 April 1960, the front page of the Daily Graphic read: When the rains came to Accra. Sixty-six years later, they came again. On 29 June 2026 the Odaw overflowed into Odawna, Alajo, Nima and the Kwame Nkrumah Circle; Kaneshie market went under; at least twelve people died, seven vanished, and nearly forty thousand were displaced. Across the border in Côte d’Ivoire the same weather system killed around twenty more.
The striking thing this time was the familiarity not the scale. The flooded communities of 2026 ;Kaneshie, Odawna, Adabraka, Alajo, Circle, Weija ,are the same names that fill the flood reports of 2023, 2016, 2015, 1997, 1968. Asked how he felt, a resident did not reach for surprise: when it rains like this, he said, we already know the trouble that is coming. A city may be forgiven a disaster it did not see coming. Accra sees this one every year, has been handed the diagnosis by its own engineers, has signed the contracts and buried the dead,and is still found, each June, almost as unprepared as the year before.
We sat in this room with a good friend from the Presidential Family, like we always do, model plans to solve pentinent issues such as galamsey, hunger, trade, etc and pitching them with our friends who are in official positions to do so or hear them at least. On the accra floods, as worried as we were, we triggered an intuitive inquiry to find solutuons, we came up with a modest model available on Research Gate (https://doi.org/10.13140/RG.2.2.19813.77280) , a non engineering perspective, the value of which, is put in this essay.
This work began not as an engineering exercise but as a structured inquiry into one question: would holding storm water in large upstream basins actually stop Accra from flooding? The intuitive hypothesis was tested two ways — against the published, calibrated modelling of the Odaw catchment (the HEC-HMS/HEC-RAS study of Acheampong, Gyamfi and Arthur, 2023) and against first-principles hydrology, using the standard SCS Curve-Number method to estimate storm runoff and a reservoir-routing relationship to link storage to peak reduction. Both pointed the same way: upstream detention trims the flood only marginally, and least in the largest storms, because the storage required to cut a peak rises with the square of that cut, and because the real bottleneck is the choked exit to the sea rather than any shortage of storage upstream. That result redirected our inquiry from how large the ponds must be to what actually limits the flood, and across how much of the city — expanding the scope from the single Odaw corridor to all five of Greater Accra’s drainage systems (some 458 km²), sizing a distributed network of about a hundred and twenty small basins against the same runoff arithmetic, siting them on the natural depressions and flow-blocking points mapped in a 2024 digital-terrain screening, and restoring the coastal wetlands as the large, self-cleansing buffer. Throughout, the figures were drawn from published basin data, official records and contemporaneous reporting and kept deliberately planning-grade, the aim being to frame the problem correctly and set the target, so that the detailed design, which belongs to the government’s calibrated models and to qualified engineers, is aimed at the right thing.
This essay makes three claims: that Accra’s flooding, though triggered by rain, is a solvable systems problem, not a weather problem; that the dominant remedy, two great detention ponds high in the hills to delay the flood is, on the evidence, the wrong one, and that the answer lies nearer its opposite, in a hundred small basins spread across the city with its natural wetlands restored; and, least comfortably, that it should not have taken an outsider to say so. Everything here is planning-grade: derived transparently, and to be checked against the government’s calibrated models before a spade breaks ground. The value is in getting the shape of the problem right, so that considerable talent and money are aimed at the right target,and not at the wrong one.
An honest accounting of the deluge
Begin with the rain, because June 2026 was genuinely extreme. Ghana recorded 593.2 millimetres that month, the highest since national records began in 1995, and 169.2 millimetres on the 29th alone, the fourth-wettest day in that record.
Many have drawn permissible general conclusions, rightly so: even if everything had been done right, rain like this would still have overwhelmed the city. In its narrow form, this is true. No drainage system on earth makes a third of a normal year’s rain vanish in a single month.
But that is the wrong test, and quietly substituting it is how the force-majeure defense does its damage. Three facts defeat this permissible narrow conclusion. First, resilience never meant prevention: a resilient system is one that fails gracefully , draining fast enough to shorten the flood, holding enough to blunt its peak, keeping the water shallow enough to take property rather than lives. Second, Accra floods in ordinary storms too. In May 2025 a 132-millimetre downpour ,heavy, but nothing like June , killed four, including a child swept into a gutter in Nima, and displaced more than three thousand. A system that drowns at ordinary rain cannot hide behind the extraordinary. Third, the extremes are becoming ordinary, and a standard built for the storms of the 1990s is already obsolete. Taken to its logical end, “the rain was unprecedented” is not a reason to do less. It is the strongest possible argument for building, deliberately and now, for the deluge.
The seduction of the great pond
This inquiry did not begin with its conclusion. It began with the intuition almost everyone reaches for: too much water arrives at the bottom of the catchment too fast, so intercept it at the top, one or two large basins in the hills, releasing slowly, flattening the peak that hits the streets. It is elegant, monumental, and it feels like decisive action. Myself and my colleague have discussed such options in tackling the galamsey polluted river systems, but that’s a study to share another day.
In Accra it has a concrete form: two proposed detention basins in the upper Odaw at Kwabenya, in the Atomic area, named due to its proximity to the Ghana Atomic Energy Commission that sits there. Their history is the history of the modern response to these floods. They were conceived after the 3 June 2015 flood-and-fire that killed more than 150 people at Circle; from that catastrophe came the World Bank-backed Greater Accra Resilient and Integrated Development project (GARID), prepared from 2016, approved in 2019, meant to run to 2025 and now extended to 2027. Within it, the two Atomic ponds were the flagship storage element, valued at roughly US$44 million. And it is only fair to say that Accra has not flooded for want of thought: its engineers have understood the mechanics since drainage master plans of 1963, 1991 and 1995; GARID has done real work on dredging, waste and early warning; and the science this argument leans on is Ghanaian and recent. The trouble is not a failure of intelligence. It is a failure of framing ; and the frame was the great pond.
Here is the first hint of that failure, after a decade, the ponds are not built. They reached design and procurement, then largely stalled. But set the delay aside and ask the deeper question this inquiry eventually learned to ask, not whether the ponds can be delivered, but whether they would work.
How the question changed
Two findings answer it; one from Ghana’s calibrated science, one from physics.
The first is empirical. In their 2023 model of the Odaw, Acheampong, Gyamfi and Arthur did not assert that detention would help; they simulated the two ponds across storms of rising severity and measured the result. The ponds cut the flood peak by only about 8–13 per cent and its volume by 2–4 per cent. Worse, the benefit shrank as the storms grew ;from around 13 per cent in a modest event to about 8 in a severe one ;because a fixed basin fills, and then every extra cubic metre pours straight over it. The great pond delivers its weakest performance in exactly the catastrophic storms it exists to defend against. A flood defense that fades as the flood worsens is not a defense at all.
The second finding explains why, and it is the crux. To cut a flood peak by a given fraction, the storage required rises with the square of that fraction, shaving 30 per cent off the peak means storing about 12 per cent of the whole storm; half means a third; 70 per cent means nearly two-thirds. Large cuts are punishingly, non-linearly expensive. Concentrate all the water at one upstream point and it is measured against the entire basin’s runoff, some twenty-two million cubic metres in June’s storm, so even a 30 per cent cut demands about 2.5 million cubic metres of storage there. The two ponds together offer around 0.2 million ,under a tenth of what a serious cut needs. The physics had ruled them out before the first spadeful of earth.
But the square law contains its own escape: the penalty applies only to whatever passes through a given basin. Spread the storage ,many small basins, each holding a small local volume, each clipping a local peak, and you pay the gentle end of the curve many times over instead of the savage end of it once. And the small basins sit where the city actually floods, above Circle and Kaneshie and Alajo, not in the thinly settled hills.
A third realisation reframes everything, even a perfectly delayed flood still arrives. If the channel and lagoon at the bottom cannot carry the water out, delay merely postpones the disaster by hours. The Odaw ends at one small, silted, waste-choked lagoon behind a tidal sandbar; Ghana’s own terrain studies show that a modelled trash barrier at that outlet reproduces the upstream flooding all by itself. The binding constraint is not upstream storage. It is conveyance, the capacity to move water out at the bottom. So the question changed: from how big must the ponds be? to what actually limits the flood, and what shape of solution addresses it? Asked that way, the answer nearly writes itself.
The scrutiny GARID deserves
To argue for a different approach is to scrutinise the dominant one, fairly, because GARID has done real work, but without flinching. Begin with scope: GARID centres on the Odaw, which carries about 60 per cent of Accra’s storm water, leaving the other 40 per cent, Densu and Weija to the west, Kpeshie, Sakumo and Ashaiman, Chemu and Songo–Mokwe to the east, largely outside the flagship. Weija, Tse Addo, Ashaiman and Kpone flood too. Then the storage itself, to have built a US$44 million marquee around an instrument the science shows to be marginal , and marginal in a way that worsens when the stakes are highest, is the central analytical weakness.
Then execution, on which We are deliberately even-handed, because it has become partisan and my point is structural. After the flood, a bitter dispute erupted over GARID’s money. Some credible Analysts argued from the data that under the previous finance ministry roughly nine of every ten GARID cedis went to purposes other than flood control; Others countered — citing a World Bank report — that they had secured the funding and completed about 40 per cent of the works, while the current government left available funds idle. We take no side. What matters is that both accounts describe the same pathology: a large, capital-concentrated, single-programme defense is acutely vulnerable to delay, to contested spending, and to the friction between administrations. When your flood defense is one big thing, one failure of will stops all of it, witness the roughly US$7 million early-warning system, built for this very region and confirmed not operational when the 2026 rains came, so that residents received no alert. Add that a 2024 assessment reportedly priced an upgraded Odaw system for a one-in-25-year storm at around US$675 million ,triple GARID’s original budget, for the central river alone ;while the Odaw’s own dredging sits about 40 per cent done, and the shape of the problem is clear. GARID should not be abandoned. But its centre of gravity; one corridor, two great ponds, concentrated capital, perpetual maintenance- is misplaced.
The model, and where the basins go
A modest proposal is one system of four layers, applied across all five basins and listed in the order in which they bind.
Conveyance comes first, because nothing works behind a blocked exit: finish dredging the Odaw and the primary drains, remove the channel encroachment (Old Fadama and Agbogbloshie the starkest case), enlarge the culverts at the about 163 chokepoints terrain that can be mapped , and keep the lagoon mouths open against the sand. Unglamorous, and the highest-value work there is.
Distributed storage replaces the two ponds with about 120 small detention basins on the natural low points across every sub-catchment, each roughly a hectare and a few metres deep, cheap, quick, buildable in parallel, and redundant, so the system keeps working even when any one silts up. This is the square law turned to the city’s advantage.
Terminal storage restores the buffers that clean themselves: the Korle, Densu Delta, Sakumo, Kpeshie and Chemu wetlands, returning on the order of 22 million cubic metres, about a whole central-basin storm, and, unlike concrete tanks, self-cleansing by tide and season, without the perpetual dredging Accra has never managed to sustain.
Source control, permeable surfaces, retained green space, an end to paving the last absorbent ground , shrinks the runoff at its origin. Wrapped around all four: a warning system actually switched on, and the enforcement that keeps cleared boundaries clear.
The number is not arbitrary. Across the city’s 458 square kilometres, June’s storm yields about 36.6 million cubic metres of runoff; a 30 per cent peak cut calls for about 4.3 million cubic metres of distributed storage; at roughly 35,000 cubic metres a basin, that is about 120, as few as 70 if basins can be made larger, up to about 214 if only smaller ones fit the dense core. Land is not the constraint: the 2024 screening mapped 5,018 natural depressions and 163 chokepoints, so 120 basins use under 2.5 per cent of available sinks. The rule is to start at the chokepoints, then the depressions just above the neighbourhoods that flood every year, then fill out the network for coverage. The table below names candidate localities ,every one a place in the flood record , at the level a budget is built; fixing the exact pond within each is the survey’s job.
Drainage system (share of runoff) Basins Candidate localities — sited on mapped low points and chokepoints
Odaw–Korle (60%) 74 Odawna, Circle, Alajo, Avenor, Agbogbloshie, Abossey Okai, Adabraka, Kaneshie; Nima, Aboabo, Maamobi, Kotobabi; Odorkor, Darkuman, Lapaz, Abeka, Tesano; Achimota, Dzorwulu; Ofankor, Dome, Taifa, Ashongman; Kwabenya/Atomic, Haatso, Legon, Pantang
Densu–Weija (15%) 19 Weija, Oblogo, Tetegu; Gbawe, Mallam, McCarthy Hill; Ngleshie Amanfro, Domeabra
Sakumo–Ashaiman (11%) 13 Ashaiman, Nmai Dzorn, Zenu; Sakumono, Lashibi–Klagon, Tema Community 18; Afienya
Kpeshie (8%) 9 Spintex, Baatsona, Manet; Tse Addo, La (Labadi), Teshie, Nungua
Songo–Mokwe–Chemu (6%) 7 Tema industrial area, Community 25; Kpone, Prampram, Dawhenya
Total 122 across five systems and roughly forty named localities, each with a restored terminal wetland behind it
Why it took an outsider to say so
Consider how a competent person engages a lawyer. She does not describe a vague unease about a piece of land and wait for the lawyer to divine her aim; she says, I want to acquire this land — give me the strongest legal justification for it. That instruction is not an insult to the lawyer’s expertise. It is the precondition for deploying it. A clearly posed objective is what frees a professional to bring the full force of their craft to bear; hand the same lawyer a fog, and even genius returns only more fog.
There are two different jobs in solving a problem like Accra’s, and we have been conflating them. One is framing: deciding what the system should optimise for, what the binding constraint truly is, what counts as success. The other is execution: the hydraulics, structures and survey that meet the frame. Accra has handed its engineers both, “flooding is a technical matter, so engineers, go and solve flooding” and then blamed the profession when the floods returned. But an engineer told to “solve flooding” and pointed at the Odaw will, quite rationally, optimise within that frame: a bigger channel, two ponds on the river he was named. What the mandate implicitly forbids is challenging the frame itself, saying the river is only 60 per cent of the problem, the storage instrument is the wrong one, the true constraint is the blocked exit and the encroached wetland and the institution that will not maintain what it builds. Reframing is a prior act, and when no one performs it, the most brilliant engineering in the world is aimed with precision at the wrong target.
This is where an outsider is an asset rather than a liability, provided he is honest about the division of labour. His job is to pose the question correctly: the constraint is conveyance and distributed local storage across five systems; the target is to clip the peak by a third with a hundred basins and restored wetlands; now bring your expertise and make it real, better and more cleverly than I ever could. That does not diminish the engineer; it elevates him, by handing him a well-posed problem worthy of his talent and making its value legible to the public that will pay for it. And it is democratic, what a city should optimise for, cost against safety, concrete against wetland, monument against network- is not a technical question to be surrendered wholesale to technocrats, but a public one. The limit is the mirror image of the point: We do not design the basins. Final sizing belongs to the calibrated models, placement to survey, the structures to qualified engineers. This essay stops at the frame and hands execution back, which is the whole thesis in practice: framing and execution are different jobs, both indispensable, and a city gets into trouble when it forgets which is which.
Designed for the deluge
The general commentators’ assertion was right that June’s storm would overwhelm any system’s power to make it vanish. But that was never the standard. A seawall is not a failure because a great wave wets the promenade; it succeeds if the town behind it is still standing. The measure of a flood system is not whether the extraordinary storm passes unnoticed, but whether, when it comes- and it will come more often now , the city bends instead of breaking: the water shallower, the flood shorter, the exits open, the warning given, the lives kept. Resilient systems are designed, deliberately and specifically, for the deluge.
Accra has spent sixty-six years proving that it knows why it floods and cannot, somehow, stop it. The barrier has never been diagnosis, and it has never truly been money, the country’s reserves dwarf the cost of a permanent fix. It has been a mistaken picture of the solution, and a confusion about whose job it is to draw that picture. The mistaken picture is the great pond: monumental, satisfying, and marginal. The truer one is humbler and stronger ,a hundred small basins where the water actually gathers, the living wetlands restored to the work they once did for free, the channels finally cleared to the sea, the warning switched on, the boundaries held. It is less photogenic than a single great dam. It is far more likely to keep a child out of a gutter in Nima. The rains will return to Accra; they always have. The only question the city has never quite answered is whether knowing that will finally be enough to build for it.
Acknowlegeement:
This synthesis rests on our simple inquiry study deposited on Research Gate which also relied on the calibrated Odaw model of Acheampong, Gyamfi and Arthur (Journal of Hydrology: Regional Studies, 2023), the source of the 8–13% peak and 2–4% volume figures for upstream detention and their decline with storm severity; a 2024 digital-terrain screening of Greater Accra (GeoHazards) that mapped 5,018 natural depressions and 163 flow-blocking points; the World Bank’s GARID programme; and contemporaneous reporting, including the Interior Minister’s statements to Parliament, Reuters, UNDP Ghana, MyJoyOnline, ModernGhana and GhanaWeb. Runoff and storage figures use the standard SCS Curve-Number method and reservoir routing and are the author’s own planning-grade derivations, to be verified against the government’s calibrated models before any design decision. The competing GARID-spending claims of Franklin Cudjoe and of the previous administration’s representatives are presented without endorsement, to illustrate the governance fragility of concentrated infrastructure.
source:www.senaradioonline.com
