10 August 2026 · Magnitude 7.4 · Western Colombia
Shaking intensity and homes destroyed or unsafe to live in, by municipality.
Colour shows the share of each municipality's homes registered as destroyed or unsafe to live in. Blue lines show shaking intensity. Select a municipality for details, or zoom into a city to see buildings mapped as damaged from satellite imagery.
Shaking intensity (Modified Mercalli scale): 5 felt by everyone; objects fall from shelves 6 furniture moves; weak buildings crack 7 considerable damage to poorly built buildings
Colour
Show
Go to
Use Ctrl + scroll to zoom, or the + and − buttons
Homes destroyed or unsafe: registered as destroyed or unsafe to live in, as a share of all homes. Sources: Servicio Geológico Colombiano (intensity), UNGRD disaster register, 12 September (damage), DANE 2026 estimate (homes), Copernicus EMSR916 (satellite), geoBoundaries CC BY 4.0 (boundaries).
Intensity vs damage
Damage rose with shaking intensity, but varied widely between municipalities at the same intensity.
Building age
On two university campuses, all 7 badly damaged buildings with a documented date were built before 1999.
Universidad Nacional rated 29 buildings with documented dates on its Manizales and Palmira campuses (–). Buildings are grouped by independently documented construction or opening dates, not by the dataset's own year field, which is wrong or unconfirmed for 9 of the 29 buildings. The split is at 1999 because Colombia's NSR-98 seismic code took effect during 1998, so a building started that year could fall under either the old or the new code. The sample is small and selected, several older buildings had been strengthened and several newer ones only have design records, so this does not show whether the code was followed.
Armenia · 2,419 inspection records
In Armenia, 24% of inspection records were tagged unsafe.
Each dot is one inspection record, made after damage was reported; the same building can appear more than once. Green: safe to use. Yellow: restricted use. Red: unsafe. Buildings without reported damage were not inspected.
By building use
Schools were tagged unsafe less often than inspection records overall.
Share of inspection records by tag.
Failure types
How buildings failed
Each failure type is illustrated with documented cases. The cases show how failures occur, not how often.
Case studies
12 documented cases
Selected for the quality of their documentation, not as a representative sample.
Data
All figures
Each figure links to its source.
Implications
What the data suggest
Older buildings
On two university campuses, every badly damaged building with a documented date predated 1999. The buildings that collapsed in Cali and Pereira haven't been documented the same way.
Walls, ceilings and parapets
Many buildings whose main structure stayed up were still closed because parapets, ceilings, walls or façades fell.