Surface sea temperature gets all the headlines, but some of the most useful, and earliest, El Niño signals come from beneath the surface, in a quantity called subsurface heat content.
What's Actually Being Measured
Subsurface heat content tracks the temperature of ocean water down to roughly 300 meters depth across the equatorial Pacific, rather than just at the surface. Under normal conditions, a layer of unusually warm water, often called the warm pool, sits in the western equatorial Pacific, held in place by steady east-to-west trade winds. When those trade winds weaken or reverse (often triggered by an active MJO and associated westerly wind bursts), that stored heat doesn't stay put.
The standard way to summarise this is the depth of the 20°C isotherm, the level at which water hits 20 degrees. That depth is a practical proxy for the thermocline, the boundary between the warm surface layer and the cold deep ocean. When the thermocline deepens in the east, warm water has arrived. When it shoals, cold water is closer to the surface.
CPC also publishes a thermocline slope index, which is the difference in anomalous 20°C isotherm depth between the western Pacific (160°E to 150°W) and the eastern Pacific (90° to 140°W). A flatter than normal slope, which shows up as a negative index, means the usual west-high east-low tilt has relaxed. That relaxation is one of the more reliable structural signs of a developing warm event, and it is currently below average while upper-ocean heat anomalies run above average.
The observations themselves come largely from the TAO/TRITON moored buoy array along the equator, supplemented by Argo floats. This is one of the few places in oceanography where a genuinely dense, continuous, decades-long subsurface record exists, which is why the diagnostic is trusted.
Why It's an Early Indicator
This heat doesn't just disappear; it propagates eastward as a slow-moving subsurface pulse called a downwelling Kelvin wave, often taking weeks to cross the Pacific before that warm water finally reaches the surface in the central and eastern Niño regions. CPC's July 9 Diagnostic Discussion described exactly this in the current event: a recent downwelling Kelvin wave deepened the thermocline and raised temperatures in the eastern Pacific.
Kelvin waves come in both signs, and the distinction matters. Downwelling waves push the thermocline deeper and warm the east. Upwelling waves do the reverse and can stall or reverse a developing event. During 2025 several upwelling waves were initiated across July, August and October; the sequence flipped to downwelling from December 2025 onward, which is the transition that set up the current event.
One caution on the big numbers. Subsurface anomalies quoted at +6 to +8°C are point maxima at a specific depth and longitude, not basin averages. They are real, and they are also the warmest pixel in a field rather than a description of the whole upper ocean. The basin-wide 0 to 300 metre heat content anomaly, which is the number the diagnostic actually rests on, is far smaller. Quoting the point maximum as though it were the index is the same category of error as reading a single warm patch off an SST map and calling it a region-wide anomaly.
This lag is precisely why subsurface heat content functions as an early-warning signal. By the time surface SST anomalies cross the official El Niño threshold, the subsurface signal has often already been elevated for one to three months, giving forecasters a meaningful head start on calling event onset and, to a lesser degree, eventual strength.
This site's subsurface heat panel and the broader SST spaghetti chart are both built around this principle, tracking not just where the surface temperature sits today, but what the underlying heat reservoir suggests is still working its way toward the surface.