Juba and Shabelle basins · 1 to 21 October 2026 · published 25 October 2026
Satellites saw 2.1 times the usual standing water across the Juba and Shabelle basins from 1 to 21 October 2026, the 6th most in 24 years of the same 21 days.
Whether this water is a flood has not been established. Jowhar read 5.37 to 5.38 m on 13 to 15 Oct 2026, above SWALIM's high flood-risk level (5.25 m) and 12 cm below bankfull (5.50 m); no primary report of flooding found as of 23 Oct. Our own instruments measure rainfall and flood extent, and neither can answer it.
Standing water outside the usual rivers and lakes, measured by MODIS over 1 to 21 October 2026, against the same 21 days in every year since 2003. That is 2.07 times the median and the 6th highest of the 24 years compared. The measure is flooded pixel-days: each 250 m square counts once for every day it shows water, so it combines how much water there was and how long it stayed. This measures water seen from orbit, not flood damage, and cloud hid part of the region at the peak.
The flooded area is likely an undercount: on 13 to 19 October, when the Shabelle at Jowhar peaked, cloud let the satellite see only 52 to 74 percent of the region, and those days read almost no flooding. A day the satellite could not see is not a dry day.
Rainfall is not ranked here: satellite rainfall may only point where to look (D-322), and no same-window rainfall baseline exists for this region.
Attribution: ENSO-loaded window, per the science desk's ruling of 23 Oct 2026 (October to December short rains; IRI El Nino rainfall composites). The tag claims the season's odds, not this flood.
| Flood extentNASA MODIS MCDWD 3-Day composite | 2.07× median | 6 of 24 |
|---|---|---|
| RainfallGPM IMERG Late Run v07 | no baseline | |
| the rainfall baseline file holds no year with a complete window, so there is nothing to compare this period against | ||
the median for these 21 days
Standing water over 1 to 21 October 2026, by year
ENSO-loaded windowWhat this is
Standing water outside a reference water mask, from the NASA MODIS 3-day flood composite, over 1 to 21 October 2026, compared with the same 21 days in each year since 2003. Each value is flooded 250 m pixels, summed over the 21 days of the window (MODIS MCDWD 3-day composite, values 2 and 3); a pixel flooded on several days counts once per day, so this is pixel-days, not area.
What this is not
Not a measurement of flood damage or of who was affected, and not proof a flood occurred. A day cloud hid is not a dry day.
Source. NASA MODIS MCDWD 3-Day composite, standing water outside the reference water mask. As of 21 October 2026.