Latin America · crops to dekad 11 August · fires week to 29 August

Eight neighbouring countries are each having one of their four worst crop years in 26.

They run from Belize to Colombia without a gap: Belize, Guatemala, El Salvador, Honduras, Nicaragua, Costa Rica, Panama and Colombia. Each is measured only against its own record. At the same moment the southern half of the continent is having an unusually QUIET fire week: Bolivia and Argentina both sit below half their normal for this week of the year. Both are findings. Neither is visible on a channel page, which shows only what qualified, or on a country page, which shows one place.

Four instruments, one region, one week

Brazil against its own slowest year, at the same week

2015 completed less of its season by this week than any year on record, 22.8% of its eventual total, and still finished second of 14. Measured in hectares burned so far, 2026 is lower than 2015 was at the same point. Those are two different measures. On hectares alone 2015 ranks 5th of 14 at this week rather than first; its claim to the slowest start is about the SHARE of its own season completed, which cannot be computed for 2026 until 2026 ends.

Cumulative burnt area by Copernicus week 34: 2026 has 4.1 million hectares against 2015’s 9.4 million, 0.44 times it, and ranks 15 of 15 years including this one. The next slowest share after 2015 is 25.4%.

Burnt area, not detections. The two measure different things and are never converted into each other, so no figure here is a fire count. Where 2026 finishes is not forecast from this: a slow start is what 2015 also had.

This column counts thermal detections against each country’s own normal week. It does not distinguish agricultural burning from wildfire. In this region that difference is large and measured: Cuba’s detections fall on cropland 2.4 times more often than chance, and Brazil’s 5.0 times LESS often. Same instrument, same week, opposite phenomena. Share of this week's detections falling on cropland, against the share of the country that IS cropland. Above 1.3 the fires sit on farmland more often than chance; below 0.77 less often. It says WHERE detections fall, not what is burning: the instrument sees a thermal anomaly, not a substrate.

Every country we measure, and every one we do not

Heat is not measured anywhere in Latin America. All 45 cities on that channel are European. This row is empty because the instrument does not reach here, not because the nights are ordinary.

Fires reaches 11 of these 24 countries. 7 of the 8 countries in the run above are outside it, so some of the worst crop stress on this page sits where the fire instrument cannot see. That is a limit of our coverage and not a finding about those countries.

Crops measures all 24 and places 20. Belize, Dominican Republic, Jamaica, Suriname have a rank against their own 26 years, but are measured on too few regional readings to sit on the map’s scale. Measured and placeable are different things, and the second is a limit of the drawing rather than of the data.

Two flood pieces cover this region: Northern Chile and the Atacama, Southern Peru and the Altiplano. They are not in this table because floods publishes by catchment rather than by country, and because each measures RAINFALL rather than flooding.