The Trade Winds Don't Care About Your Calendar
Upwelling isn't some on/off switch. It's a continuous process driven by wind stress and the Coriolis effect, and El Niño and La Niña are just deviations from the baseline state. People treat it like a mystery because the terminology gets muddy fast. What I'm going to lay out here is how it actually functions in the Pacific, not the simplified version you'll find in most textbooks. At its core, upwelling in the equatorial Pacific is caused by the trades pushing surface water westward. That displaced water has to be replaced. Cold, nutrient-rich water from below rises to fill the gap. That's the engine. Everything else is modulation of that engine's output. During La Niña, the trades are stronger than normal. Stronger trades mean more westward surface transport, more divergence near the equator, and therefore enhanced upwelling. The thermocline tilts steeply, deeper on the western side and shallower on the eastern side. What this means practically is that the cold tongue along the South American coast is sharper, more persistent, and richer in nutrients. Anchovy populations in Peru usually respond well to this. Fisher folks know this intimately because the catches track these phases like a clock.
During El Niño, the trades weaken or even reverse. The westward push fades, the divergence weakens, and the upwelling cell essentially takes a breath. The thermocline flattens out or tilts in reverse, warm water pools along the eastern Pacific, and the cold upwelled water gets pushed deeper or suppressed entirely. This is why you see mass fish die-offs and seabird collapses during strong El Niño years. The nutrient supply gets cut off. I ran a coastal model once for a research group tracking Peruvian anchoveta, and we had a stubborn bug where the upwelling index kept showing strong signals during a moderate El Niño that everyone onshore was seeing as biologically dead. The issue turned out to be that we were using surface wind stress alone without accounting for the subsurface thermal advection lag. Once I added the depth-integrated heat content term and switched from raw wind data to a corrected Ekman transport calculation, the model output matched the field observations. It cost us three weeks of recalibration but saved the paper. One thing beginners consistently get wrong is assuming upwelling stops completely during El Niño. It doesn't stop. It weakens significantly, but local wind patterns, coastal geometry, and atmospheric Kelvin waves can still drive pockets of upwelling even in the middle of a strong El Niño event. The 1997-98 El Niño had several localized upwelling events off the Galápagos that caught researchers off guard because the regional signal was so overwhelmingly suppressed. If you're working with real data and your upwelling index reads zero everywhere during an El Niño, check your data source. You're probably looking at a smoothed composite, not point measurements.
Another nuance is the role of the Madden-Julian Oscillation. It modulates the trades on a 30 to 60 day timescale, and those pulses can temporarily reignite upwelling even during El Niño conditions. If you're monitoring this for aquaculture or fisheries management and you only look at monthly averages, you'll miss these pulses entirely. Daily resolution matters more than people realize. The southern boundary of the upwelling zone also shifts. During La Niña it extends further north along the South American coast. During El Niño it contracts southward. This has cascading effects on sediment transport, marine productivity gradients, and even the distribution of sardine and mackerel species. It's not just about temperature. Here's the uncomfortable part: current climate models still struggle to accurately simulate the exact magnitude of upwelling changes during ENSO transitions. The resolution required to capture coastal upwelling dynamics properly is somewhere around one-kilometer grid spacing, and most global climate models run at 50 to 100 kilometers. That gap means projections about how ENSO-driven upwelling will change under global warming carry substantial uncertainty. Downscaling helps, but it's computationally expensive and introduces its own errors.
Get the Full Details
If you need real-time upwelling estimates for operational purposes, satellite-derived sea surface temperature anomalies combined with OSCAR ocean surface current data give you something reasonably close to real-time. It won't capture subsurface thermocline dynamics perfectly, but for tracking the broad ENSO signal it's practical. For precise work, you need mooring arrays like the TAO/TRITON buoys, and even those have maintenance issues and data gaps that can stretch for months. What I've found working with this stuff for years is that the most reliable approach combines multiple data streams. Satellite SST, buoy thermocline depth readings, altimetry data for sea surface height anomalies, and wind stress from reanalysis products. Cross-referencing them catches the cases where any single source misleads you. A strong El Niño might show weak upwelling in the satellite data but the buoy network reveals that subsurface processes are still active enough to matter ecologically. Missing that connection is an easy mistake to make if you're relying on one dataset alone. The takeaway isn't complicated. Upwelling during ENSO events is fundamentally about how the trades modulate Ekman transport and thermocline depth. La Niña amplifies the system. El Niño dampens it. The devil is in the spatial and temporal details, and those details are where most people stumble because the signals aren't uniform across the basin.
