✻Forecast FarmClimate · Cultivation · Consequence
Commodity desk · ICE US Cocoa (CC) · ICE London (C)

Cocoa · West Africa

Côte d'Ivoire and Ghana grow roughly six of every ten tonnes of the world's cocoa, and the belt is rain-fed smallholder forest agriculture: the market has no irrigation hedge. July–September rain builds the October main crop, and the harmattan's arrival decides how it dries.

ENSO state · through Sep 2026
very strong El Niño +3.40 · strengthening · Oct 2026 provisional +3.83
The Forecast Network's ERA5 multivariate ENSO index (the series behind celsius.earth/enso); positive is El Niño-like. Monthly means arrive about six weeks after the month; the provisional month is ERA5T and may still move.
Current stage
Main-crop harvest ★
Temp, next 2 wk
-0.2°C vs normal
Rain, next 2 wk
66 % of normal
SEAS5, next 3 mo
+1.3°C vs hindcast
Obs rain, 3 mo
157 % of normal
ENSO
very strong El Niño +3.4

The daily sheet · 2026-10-21 to 2026-11-03

Temperature anomaly · °C vs 1991–2025 normal
+0.7°−0.7°2123252729312
Rainfall · mm per day, brass dash = normal
2123252729312

Day-by-day ensemble means from the same ECMWF extended run as the weekly ledger — the short-term sheet a desk reads each morning. Days are UTC.

The six-week ledger · ECMWF extended ensemble, 2026-10-21 00z

Temperature anomaly · °C vs 1991–2025 normal
+1.5°C0−1.5°C+0.0W1-0.5W2-0.4W3-0.3W4-0.1W5+0.1W6
Rainfall · mm per week, brass dash = normal
24W1 · 52%29W2 · 85%21W3 · 108%13W4 · 68%8W5 · 43%7W6 · 40%
The belts, first fortnight
RegionTemp anomalyRain% of normal
Côte d'Ivoire cocoa belt-0.3°C53 mm74%
Ghana cocoa belt+0.2°C48 mm68%
Nigeria–Cameroon belt-0.7°C60 mm55%

Ensemble-mean weekly aggregates over the desk's growing belts (land cells, production-weighted); whiskers show the ensemble spread carried with the mean — an upper bound for a weekly average. Weeks run from the initialisation date.

The plates · the fortnight on the map

Temperature anomaly, week 1, over the Cocoa · West Africa belts
Temperature anomaly · Week 1 · 2026-10-21 → 2026-10-27
Rainfall, percent of normal, week 1, over the Cocoa · West Africa belts
Rainfall, percent of normal · Week 1 · 2026-10-21 → 2026-10-27
Temperature anomaly, week 2, over the Cocoa · West Africa belts
Temperature anomaly · Week 2 · 2026-10-28 → 2026-11-03
Rainfall, percent of normal, week 2, over the Cocoa · West Africa belts
Rainfall, percent of normal · Week 2 · 2026-10-28 → 2026-11-03
Temperature anomaly (°C vs normal)≤ −4−4 to −2−2 to −1−1 to −½near normal+½ to +1+1 to +2+2 to +4> +4
Rainfall, percent of normal (% of normal rain)< 4040–6060–80near normal120–150150–200200–300> 300

EC-46 ensemble-mean weekly fields at 0.25°, painted in the Herbarium stepped ramps over the desk's belts (the oxblood rectangles). Near-normal stays paper; sea and no-data carry no ink; rainfall percent is unpainted where the weekly normal is under 2 mm. Cells are honest — drawn at the data's own resolution.

The seasonal ledger · SEAS5, initialised Oct 2026

Run-to-run · the last 6 initialisations
+2.50MayJunJulAugSepOctNovDecJanFebMarApr2 m temperature anomaly, °C · dark line = latest run
A forecast that holds its shape across initialisations deserves more weight than one that just appeared.
Latest run, month by month
MonthTemp anomRain anom% of normalTopsoil moisture anom m³/m³
Oct 2026 ◔+0.9°C+6 mm103%-0.002
Nov 2026+1.0°C+5 mm106%+0.006
Dec 2026+1.2°C-6 mm81%-0.007
Jan 2027+1.7°C-5 mm79%-0.009
Feb 2027+2.0°C-9 mm81%-0.014
Mar 2027+2.2°C-15 mm87%-0.017
Apr 2027+2.1°C-4 mm97%-0.015
◔ part elapsed. Anomalies vs the store's ERA5 hindcast climatology; percent-of-normal vs the AgroClim 1991–2025 normals. Topsoil moisture is volumetric water in the top 7 cm — negative is drier than the hindcast normal.

Rain that actually fell · CHIRPS, through Dec 2025

Monthly rainfall over the belts · mm, brass dash = 1991–2020 normal
Last 3 months 157% of normal
JanFebMarAprMayJunJulAugSepOctNovDec

CHIRPS v3.0 satellite-gauge rainfall sampled at a lattice of land points across the desk's belts, production-weighted. Observed data, not a forecast.

The crop calendar · what this weather means now

JanFebMarAprMayJunJulAugSepOctNovDec
Flowering & pod set
Main-crop pod fill ★
Main-crop harvest ★
Harmattan watch
Mid-crop

★ market-critical stage · the outlined column is the current month.

What the desk watches
  • Rainfall percent-of-normal over the belt: pod fill and bean size follow the July–September rains with a one-to-three-month lag.
  • Dry-season severity: an early or harsh harmattan (December–February) cuts the tail of the main crop and stresses the trees ahead of the mid crop.
  • Sustained wet spells at harvest slow drying and raise quality discounts.
Climate drivers

El Niño years tilt West Africa toward a drier main-crop season and a hotter, dustier harmattan; the 2015–16 and 2023–24 events both preceded sharp production shortfalls. The desk shows the current ENSO state next to the rainfall ledger so that exposure is read, not assumed.

Go deeper