Why there's no one global seasonal sea level curve
Seasonal sea level is driven by several different physical processes, and they don't all peak at the same time or in the same place. In the China Seas, researchers identified four distinct shapes of seasonal cycle: an asymmetric wave, a sharp peak on a flat background, a relatively flat signal, and a symmetric wave [3]. The asymmetric type, found in northern China, Taiwan, Korea, Japan and the Philippines, is dominated by the 'inverse barometer' effect—where higher atmospheric pressure pushes the sea surface down and lower pressure lets it rise—along with a steric contribution, meaning water expanding or contracting as it warms or cools [3]. The sharp-peak type along eastern and southern Chinese coasts is linked to the seasonal monsoon and freshwater exchange [3]. Because different regions are governed by different processes, their seasonal peaks land at different times of year, which blurs any single global pattern.
The scale of these regional cycles is substantial where they occur. The 18–59 cm range in the China Seas is comparable to the kind of water-level anomaly that matters for coastal flooding [3]. So the premise that sea level lacks seasonality is really a statement about the global average, not about any particular coastline.
The seasonal peak can move, which hides the pattern
Even where a seasonal cycle exists, its timing is not fixed. A global analysis found that the regular, predictable seasonal cycles of mean sea level and storm surge have shown non-stationarity in some locations—meaning the expected peak can arrive earlier or later than it used to [5]. This matters because if the high mean sea level shifts to a time of year when storms or high tides also peak, coastal flooding risk rises [5]. A cycle whose phase wanders is harder to spot as a stable seasonal signature.
Large-scale climate patterns also reshape the cycle. The same analysis found that dominant modes of large-scale climate variability influence changes in the seasonal cycles of both sea level and storm surge, and that mean sea level and storm surge peak within 30 days of each other at a major portion of tide gauges studied [5]. Meanwhile, ocean wave seasonality was more stable over time, with peaks mostly in the same month across the study period [5]—a useful contrast showing that not every coastal variable behaves the same way.
Seasonal sea level can be predicted months ahead
The seasonal and year-to-year swings in coastal water level are not just noise—they raise the baseline on which storm surges and tides act, and they are potentially predictable months in advance [4]. On the Australian north and west coasts, sea level anomalies tied to tropical climate drivers can reach ±20 cm, and a seasonal forecast system showed skill there at lead times of one to eight months [4]. Adding the inverse barometer effect as a post-processing step improved skill around most of the coast, especially in the east [4]. The forecast skill came largely from the system's ability to predict El Niño–Southern Oscillation and its connection to Australian sea level through oceanic teleconnections, with other modes of variability playing a minor role [4].
This predictability is the practical flip side of seasonality: the cycle is real enough, and linked closely enough to large-scale climate patterns, that it can be forecast. That would be impossible if sea level had no seasonal structure at all.
Why flood-risk estimates now build seasonality in
The seasonal structure also matters for estimating extreme sea levels, which is what coastal flood defences are designed around. A newer statistical method treats sea level as two components: peak tide, which is predictable, and skew surge—the difference between the observed high tide and the predicted one—which is random [1]. The method captures within-year seasonality by adding a daily term to the model of the surge tail and by letting the distribution of peak tides change across months and years [1]. The authors report that this improves on previous return-level estimates, which typically underestimated them [1].
In other words, the seasonality is real enough that ignoring it biases the numbers used for flood planning. The reason sea level can seem seasonless is not that the cycle is absent, but that it is region-specific, sometimes shifting, and often smaller than the tides and surges that dominate day-to-day water levels.
Sources used in this answer
Accounting for seasonality in extreme sea-level estimation
A new statistical method for estimating extreme sea levels captures within-year seasonality, interannual variation and long-term change, and improves on earlier methods that typically underestimated return levels.
Accouting for seasonality in extreme sea level estimation
The same seasonal extreme-sea-level method models skew surge with a non-stationary generalized Pareto tail and lets peak-tide distribution vary by month and year, illustrated at four UK tide gauges.
Drivers for seasonal variability in sea level around the China seas
In the China Seas the seasonal cycle explains 37%–94% of monthly sea level variability at 81 tide gauges, with amplitudes of 18–59 cm and four distinct cycle shapes driven by different processes.
Seasonal sea level forecasts for the Australian coast
Australian coastal sea level anomalies linked to ENSO can reach ±20 cm and are forecast with skill one to eight months ahead, improved by including the inverse barometer effect.
Analyzing the Global Changes in the Seasonal Cycle of Sea Level, Storm Surge, and Ocean Waves
Seasonal cycles of mean sea level and storm surge have shifted in timing at some locations, and they peak within 30 days of each other at many tide gauges, while ocean wave seasonality was more stable.