Morocco's seven-year drought ended when the Atlantic jet stream dipped south

The mechanism is unglamorous and entirely local to the Atlantic. The pressure difference between the Azores and Iceland sets how fast the jet stream runs and how far north it sits, and weakening it lets the jet drop toward the subtropics ^[6]. That is what the winter of 2025-26 did: pressure falling near the Azores, rising near Iceland, and the jet's fastest core sitting off the Moroccan coast rather than near the British Isles ^[7]. Storms followed it, the Rif and Atlas ranges lifted their moist air, and the heaviest rain landed over the Sebou basin in the northeast, each system arriving on ground the last one had already soaked ^[9].

The claim about the Pacific is narrower than the El Nino headline suggests. During that winter, according to the analysis, the tropical Pacific and the North Atlantic Oscillation were effectively unrelated ^[13]. A seasonal outlook keyed to Pacific sea surface temperatures would not have anticipated the season that broke the drought. The same authors note that strong El Nino winters tend to nudge the Oscillation negative late in the season ^[11] and that wetter early winters over Iberia have been linked to El Nino ^[12], so they allow that a strong event could raise the odds of another wet Moroccan winter while saying plainly they cannot yet call it ^[14].

No one in the piece treats the Pacific as the controlling variable for northwest African rainfall. It reports the opposite of a clean signal: decades of research has failed to pin a simple, reliable El Nino fingerprint over the Mediterranean ^[15], because the Pacific's influence arrives indirectly and inconsistently and competes with the Atlantic's own variability ^[16]. The argument on offer is against determinism, in either direction.

There is an arithmetic wrinkle worth naming. Six consecutive below-average rainfall years, 2019 to 2024, is the run described as the longest on record ^[1], while the drought that ended is described as seven years long ^[3]. 2024 minus 2019 plus one is six, so the seventh year has to be the 2024-25 hydrological year ^[18]. That discrepancy matters for anyone writing a drought trigger into a contract.

This analysis leaves out how much rain actually fell. No national total for 2025-26 is given to set against the 106.4 mm of 2023-24 ^[2], no reservoir percentage sits behind "filling rapidly" ^[3], and no statistic sits behind "effectively unrelated" ^[13]. It is one winter, described in a first-person article by one of the researchers involved ^[19]. Full reservoirs are also not recharged aquifers: levels in several Moroccan aquifers fell over three decades of overuse ^[5], and per capita renewable water availability is about 620 cubic meters a year ^[4], which one wet winter does not change.

For water planning, the operative variable is the Azores-Iceland gradient itself, and NOAA's better-than-90-percent odds on a very strong El Nino ^[10] do not settle where it lands.