Terra Rhythm

Terra Rhythm · A scroll-driven data essay

The rhythm
of rain

West Africa's year doesn't turn on temperature. It turns on water — a rain belt that breathes north and south and sets the tempo for everything beneath it. Scroll slowly; the charts keep time.


⚑ All figures illustrative — read the method note
Begin

The Wheel

Before you can hear a rhythm, you need an instrument that can hold a whole year in one glance. Meet the year-wheel.

Read it like a clock: January at the top, the year runs clockwise, and rain will grow outward from the hub.

The instrument

A year, bent into a circle

Most rainfall charts are laid out like a sentence — January on the left, December on the right, as if the year ended. But a year doesn't end. December leans into January; the dry season wraps around the calendar's seam like thread around a spool.

So we'll bend the chart into a circle. Twelve spokes, one per month, January at twelve o'clock. Rain, when it comes, will grow outward from the hub like rings in wet wood.

The first rains

Then the sky opens

Here is a year of rain over Kumasi, in Ghana's forest belt — modelled on the region's long-term pattern, and clearly labeled illustrative. Watch the bars reach for the rim: a quiet January, a stirring March, and then June — around 210 mm in a single month, more rain than some cities see in a year.

Add it up and the wheel holds roughly 1,400 mm — a metre and a half of sky, delivered on schedule.

Two peaks

Not one wet season — two

Now squint at the shape. This is not one hump of rain: it's two. A long first crescendo from March to July, then — strangely — a hush in August, right in what should be the heart of the wet season. Then a second, shorter movement from September to November.

Locals don't need the chart. The August lull is planned around, cooked around, even named around. The wheel simply makes visible what every farmer in the forest belt already keeps in their head.

480 km north

Tamale keeps different time

Drive north from Kumasi for a day — forest thinning to savanna — and the rhythm changes entirely. The green line is Tamale's year on the same wheel: silent through the long dry season, then one great wave that builds all summer and crests in August–September, exactly when Kumasi is pausing.

Two cities, one country, one sky. Different drums.

The question

Same sky, different drum

Why does the south's year beat twice while the north's beats once? Nothing about the ground explains it. The answer is above: the rain isn't really seasonal at all. It's positional — a belt of weather that travels, and your rhythm depends on where you stand along its road.

To see that, we need a map.

The Front

The whole region's weather leans on one slow, magnificent commute — a wall of moist air that spends the year walking north, and then walking home.

A stylised map: the Gulf of Guinea below, the arid north above — and every city living somewhere on the rain's road.

The stage

One coastline, many rhythms

Here is West Africa, simplified to its essentials: the Atlantic curling around the bulge, the Gulf of Guinea along the bottom, the arid interior hatched at the top. Kumasi and Tamale — our two drummers from Act I — sit 480 km apart on the same meridian.

Two air masses share this stage. Cool, moist air off the ocean. Hot, bone-dry air off the desert. Where they meet, it rains.

January

The dry season's grip

In January the meeting line — call it the front — is pressed right down against the coast, near six degrees north. Nearly the whole map sits on the desert's side of it. This is harmattan weather: hazy skies, cracked lips, dust from the interior riding a northeast wind all the way to the sea.

Only a thin ribbon along the Gulf keeps any humidity at all. Everyone else waits.

The climb

Eight months of walking north

As the sun climbs through the year, the moist ocean air shoulders the dry air backwards. Watch the front go: seven degrees by March, eleven by May, and by August it stands near eighteen degrees north — deep in the Sahel, over a thousand kilometres from where it started.

The rain doesn't fall at the front. It hangs a little behind it, a belt of storms trailing to the south like a curtain hemmed to the sky. Wherever that curtain is — that's where the wet season is happening.

The retreat

And then it turns around

The front's retreat is faster than its advance — back through the Sahel in September, past Tamale in October, pressed against the coast again by December.

And here is the trick of the whole system: on the way south, the rain curtain passes over Kumasi and Accra a second time. The south doesn't get two wet seasons because the sky is generous. It gets two because it stands on a road the rain walks down twice.

The reveal

The wheel, explained

Now Act I resolves like a chord. Kumasi's two peaks are the belt's northbound and southbound passes. The August pause is the gap in between, when the rains are busy elsewhere. And Tamale, standing near the belt's northern turning point, is grazed just once — a single long wave, cresting in August.

One belt. One commute. Every rhythm on the wheel is just a question of address.

The Farm Calendar

Where rain is a schedule, farming is choreography — every task on the land is a note played against the sky's tempo.

The same Kumasi curve, laid flat — and beneath it, the farm's year snapping to its shape.

The layout

Hang the year on a line

For this act we unbend the wheel. The blue curve along the top is Kumasi's rain again — the same two-peaked figure, laid flat so we can hang a working calendar beneath it, task by task.

What follows is a composite maize year for Ghana's forest belt: not any one farm's diary, but the general choreography the rains impose on all of them.

First season

Everything follows onset

The first bars drop in, and notice how each one leans on the curve above it. Land is cleared in the dry months, when cut vegetation actually dries. Planting waits for onset — the moment the rains stop teasing and hold, usually late March. Weeding peaks with the deluge, because weeds love rain as much as maize does. And the first harvest arrives in July, timed to land just as the sky begins to ease.

Nothing on this calendar is arbitrary. Every date is the rain, translated.

Second chance

The pause is a gift

Remember the August hush from Act I — the odd quiet between the peaks? On the farm it is anything but dead time. Dry weeks right after a harvest are exactly what grain needs: maize dries in the sun, gets shelled, gets stored, gets to market on passable roads.

Then the second rains arrive, and the whole dance runs again in miniature — a second planting in late August, a second harvest by December. The south's double-beat rhythm becomes a double-cropping economy. Two peaks, two harvests.

Meanwhile, north

One season, one chance

Now recall Tamale's single wave. North of roughly nine degrees, this whole calendar compresses into one pass: one clearing, one planting, one harvest, and a long dry season to plan around it. Sorghum and millet — crops bred for patience — replace the forest belt's second maize crop.

It's the same choreography with half the music. Which is why, in the north, reading the sky correctly isn't just good practice. It's the year's single most important decision.

The City

Accra doesn't farm the rain. It has to swallow it — all of it, fast, through pipes and gutters sized by arithmetic.

Coastal squalls arrive fast and heavy — here is a modelled two-hour storm, sliced into ten-minute pieces.

Scene

A June morning, dark at nine

When the rain belt sits over the coast, Accra's storms don't drizzle in — they arrive like a decision. The sky goes charcoal at breakfast, the first fat drops hit like thrown coins, and within minutes the city is inside a waterfall.

To understand what happens next, meteorologists slice a storm the way we've been slicing the year: into pieces. Here is a modelled two-hour squall, ten minutes at a time.

The peak

Ten violent minutes

The storm's whole character lives in its third slice: around 22 mm of rain in ten minutes. For scale, that single slice is more water than Kumasi's entire January — the month the wheel showed as a sliver — delivered before a kettle boils.

Cities are rarely broken by how much rain falls. They're broken by how fast.

The line

What the drain can swallow

Every gutter and culvert in a city embodies one number: the heaviest rain it was designed to pass. Draw that number on the storm — the dashed line — and the morning's story tells itself.

Five slices of this storm land above the line. The water above it doesn't disappear or wait politely; it queues in the only reservoir left, which is the street. Junctions become ponds. The flood isn't an invasion — it's an overflow, exactly as deep as the gap between two lines on this chart.

The ledger

Where a flood comes from

Total the slices and the storm delivered about 94 mm; the drains, running flat out, could pass roughly 74 in the same window. The remainder stood in the city until the sky eased and the pipes caught up.

This is the sober wonder of urban hydrology: floods are rarely mysterious. They are subtraction — a fast sky minus a fixed pipe. And as the next act shows, one side of that subtraction is beginning to move.

The Future

The rhythm holds. What's starting to swing is the timing — and on a calendar built from rain, timing is everything.

Thirty modelled years, one dot each: the date the first rains finally settled in.

The bet

Every year, one guess

Act III hid a hard question inside one soft word: onset. Plant too early and a false start strands your seed in dust; too late and the crop runs out of season. Every farming family, every year, makes this one bet against the sky.

So let's chart the bet itself. Thirty modelled years, one dot per year: the date the first rains genuinely settled in over southern Ghana.

The spread

A widening question

Watch the band, not the dots. In the early years the onset kept beautiful time — nearly every dot within a week of late March. The middle of the band barely drifts across three decades. What changes is its width: by the 2020s the same bet spans more than five weeks.

The average year looks reassuringly normal. The average is not the problem.

The outliers

The edges do the damage

It's the extreme years that write themselves into memory. A start a month late means seed re-bought and a season squeezed. A start three weeks early tempts a planting the sky then abandons. Neither extreme shows up in an annual-total chart — the year's volume of rain can be perfectly ordinary while its timing quietly ruins the choreography.

Rhythm, it turns out, is a resource of its own. You can lose it while losing nothing else.

Coda

Listen anyway

None of this is a story about the rain stopping. The belt still makes its great commute; the wheel still fills twice in the south and once in the north; the pause still dries the harvest. The music is intact — it's the metronome that's loosening.

And so the response, everywhere on this map, is the same one musicians use: listen harder. Forecast onset instead of assuming it. Size drains for the storm that's coming, not the one from the old tables. Keep both hands on a rhythm that has carried this region for as long as anyone has farmed it — and will carry it still, a little less predictably, from here.

Method note

About the numbers

Every figure in this essay is illustrative. The charts were built from small, hand-authored datasets, modelled on the well-documented broad patterns of West African seasonal climate — the bimodal south and unimodal north, the seasonal migration of the intertropical front, the double-cropping calendar of the forest belt, and the flash-flood arithmetic of coastal squalls.

Please don't cite this. The numbers were designed to teach the shape of these phenomena, not to report measurements. No dataset here is real: station totals, front latitudes, the storm, the drain capacity and the onset dates are all invented, plausible-by-construction values. For real data, national meteorological agencies and the global climate-research community publish the genuine article.

Why draw invented data at all? Because the patterns themselves are real, widely described, and — the essay's whole premise — genuinely wonderful. Illustrative data lets the charts stay simple enough to read at a glance while remaining honest about what they are. Every chart carries the label, and every value used is published below.

Table 1 — Monthly rainfall, illustrative stations (Acts I & III)
mm per month; invented values echoing typical bimodal (Kumasi) and unimodal (Tamale) shapes.
StationJanFebMarAprMayJunJulAugSepOctNovDecYear
Kumasi (south)20551201451802101207517019090301,405
Tamale (north)363070105130150200220951241,025
Table 2 — Seasonal front, approximate surface latitude by month (Act II)
Degrees north; invented positions sketching the front's annual migration.
JanFebMarAprMayJunJulAugSepOctNovDec
Front (°N)5.567.591113.51618151186
Table 3 — Modelled Accra squall, 10-minute slices (Act IV)
mm per 10 minutes; invented storm against an invented 10 mm/10 min design capacity.
Minute0102030405060708090100110Total
Rain (mm)261422181285321194
Table 4 — Modelled first-season onset dates, 1995–2024 (Act V)
Day of year on which the modelled first rains “settled in”; spread widens by construction.
Year'95'96'97'98'99'00'01'02'03'04'05'06'07'08'09
Onset (day)768274857871837786738879699075
Year'10'11'12'13'14'15'16'17'18'19'20'21'22'23'24
Onset (day)8466927295638799689461977110365

The maps are stylised, not surveyed; the drainage story describes engineering arithmetic in general, not any specific network or event; and the essay is deliberately apolitical — it is about wonder at a system, written from within the region that lives by it.