The morning of 29 June

Around 140 millimetres of rain fell on Accra within a few hours on the morning of Monday 29 June 2026 — the heaviest single downpour the city had seen in years, and more than double the 56 millimetres recorded in a comparable event the previous year. It arrived on a Monday. People were on the road to work.

The neighbourhoods that go under went under. Kaneshie. Alajo. Odawna. Weija. The Kwame Nkrumah Circle enclave. Adabraka. Roads became channels within the hour. Rescue teams pulled more than four hundred people from submerged homes. In the Achimota–Agbogbloshie area a mother and her child were swept away. At least twelve people died in the capital.

The national picture was worse than the Accra picture, which is a thing the coverage largely missed. By 7 July the Interior Minister reported thirty-five deaths in rain-related incidents during June — eighteen in the Central Region, thirteen in Greater Accra, four in Volta — with six people still missing. Seven regions were affected. More than ninety-one thousand people were displaced nationwide. Eighteen Metropolitan, Municipal and District Assemblies in Greater Accra alone reported impact.

Then came the part Ghanaians can now recite from memory. Emergency response activated. Relief items distributed. Assessments commissioned. The Ghana Armed Forces tasked from 3 July to lead post-flood recovery in Accra and parts of Volta and Central — clearing debris, removing silt and refuse from roads, drains and waterways. A task force stood up. Demolitions announced. A budget commitment made for next year.

Every element of the response was competent. Every element of it was also a response — which is the entire problem.

The number that explains everything

Buried in academic research on Accra’s waterways conducted between 2020 and 2022 is a single operational fact that does more to explain the June floods than any quantity of rainfall data.

Thirteen people. For a metropolitan assembly serving a population in the millions, across a drainage network of primary channels, secondary drains and tertiary gutters that runs to hundreds of kilometres. That is not a maintenance regime. It is a gesture.

Consider what it means in practice. A secondary drain in a residential neighbourhood, open-topped as most of them are, collects roadside dust, loose sand and windblown litter continuously through the dry season. By the time the first rain falls, the channel has lost a substantial fraction of its capacity before a single drop of stormwater enters it. Multiply that across every uncovered gutter in the city and Accra begins each rainy season with a drainage system already partly full of solid material — and thirteen people with which to empty it.

This is why annual desilting campaigns fail even when they are genuinely executed. They are not maintenance. They are triage performed once a year on a system that needs continuous attention, by a workforce roughly the size of a football team.

To their credit, both the Ministry and the Presidency have now identified this. The Works, Housing and Water Resources Minister has said that on the President’s directive, desilting nationwide is to become a continuous, year-round exercise rather than a seasonal one. That is precisely the correct diagnosis. Whether it is accompanied by the several hundred additional staff, vehicles and disposal contracts that “continuous” actually requires is the question nobody has answered, and it is the only question that matters.

Ninety-one years of the same lesson

Accra has recorded flooding almost every year since 1935. That is not a rhetorical flourish; it is the documented baseline. Ninety-one years of annual evidence, across colonial administration and nine decades of independent government, through every planning regime the city has ever had.

The President of the Ghana Institution of Engineers put the professional verdict in one sentence: we have done things the wrong way for over thirty to forty years.

What “the wrong way” means specifically is well established and not seriously contested by anyone who studies it. Natural water retention areas that once absorbed rainfall have been built over. Wetlands have been filled and sold. Permeable ground has been replaced by concrete and roofing across an enormous and rapidly expanding footprint, so rainfall that once soaked in now runs off within minutes. Drainage channels are undersized for the catchment they now serve, and blocked besides. And structures have been erected directly on and over watercourses, with planning permission granted or simply never sought.

The rainfall did not change. The city changed. A storm that Accra absorbed in 1975 now overwhelms it before lunchtime, because the surface that used to do the absorbing has been paved and the channels that used to do the carrying have been narrowed, silted and built upon.

What the task force is actually doing

Contrary to a widespread impression on social media that nothing is happening, a great deal is happening — and it is worth setting out precisely, because the work is real and the criticism should be aimed at the right target.

The task force is led by the Deputy Chief of Staff for Operations, Stan Xoese Dogbe, with the Works, Housing and Water Resources Minister Kenneth Adjei and coordinated by Brigadier General Forster Okae-Yeboah, supported by the Engineer Brigade. The operations are specific and documented: at Oyarifa a block of apartments built directly over a natural waterway was removed, along with a watercourse a developer had diverted entirely; at Mayfair Estates a concrete-reinforced bridge built by residents and developers across a natural waterway was pulled down.

That last detail deserves emphasis, because it identifies who is actually being cleared. An apartment block over a watercourse and a reinforced concrete bridge at a gated estate are not the works of squatters. They are capital projects. Somebody financed them, somebody built them, and in most cases somebody in an office signed something. The demolition removes the obstruction. It does not touch the permission.

Ghana is very good at removing buildings from waterways and has never once removed the official who let them be built there.

The money, and where it goes

GARID is a 150-million-dollar World Bank effort reaching more than 2.5 million residents of the Odaw River Basin, built to address drainage and solid waste together. It is the largest single instrument Ghana has for this problem, and its existence undercuts any claim that the capital lacks funded plans.

There is one item in that list worth watching closely. At Agbogbloshie, the Assembly has stated it is awaiting compensation funds from the World Bank in order to pay residents whose structures will be affected by demolition — and that this compensation is what will allow full implementation of the Korle drain desilting project. In other words, a flood-prevention scheme in the most flood-critical channel in Accra is gated on a compensation disbursement.

Note the contrast, because it is stark and it is the throughline of this whole series. Where the demolition is for a drain, compensation is a prerequisite and the work waits. Where the demolition has been for a market development elsewhere in the country, traders have repeatedly described structures coming down first and compensation being discussed afterwards, or not at all.

The honest version of the fix

Ghana does not need a new strategy for Accra’s floods. It has had the correct strategy for years, stated repeatedly by its own ministers: expand drainage, restore wetlands, enforce planning law, maintain channels continuously, manage waste at source. The Works Minister himself has said the country already possesses the laws and the technical expertise, and has fallen short on enforcement and cooperation. That is an accurate self-assessment and it should be quoted back at the state annually.

Publish the maintenance headcount

If desilting is now continuous and year-round, then every metropolitan and municipal assembly in Greater Accra should publish, quarterly, the number of staff assigned to drain maintenance, the kilometres cleared and the tonnage removed. The thirteen-person figure became public through academic research, not disclosure. It should not have taken a university to find it.

Name the permissions, not just the structures

Every demolished obstruction should be published alongside the answer to one question: was it permitted, and if so by whom? Until that answer appears, demolition is a treatment of symptoms carried out at the expense of whoever bought the property, while the office that enabled it continues issuing permits.

Protect what remains before restoring what is gone

Wetland restoration is expensive and slow. A moratorium on further development in the remaining natural retention areas of the Densu and Odaw catchments costs nothing and can be issued this week. Restoring degraded catchments upstream improves water retention and slows peak flows before they ever reach the city — but only if there is anything left upstream to restore.

Fix the waste, or accept the flood

Bigger drains filled with the same plastic flood the same way. Until the volume of material entering the channels falls, every cedi spent on dredging buys a diminishing number of months. GARID’s designers understood this a decade ago. The implementation has not caught up with the design.

Facts verified: 22 August 2026.
Rainfall, casualty and displacement figures from Ghana’s Ministry of the Interior, the National Disaster Management Organisation and the Copernicus Emergency Management Service record of the June 2026 floods. Task force operations, locations and leadership as reported by Ghanaian media in July and August 2026. GARID scope per World Bank project description. The thirteen-person drain-clearing figure derives from peer-reviewed research on Accra’s waterways covering the period 2020–2022 and reflects that period, not necessarily current staffing. The Ghana Institution of Engineers assessment is as publicly stated by its President. Budget commitments announced in the 2026 Mid-Year Review had not been costed in published form at the time of writing.