Finding No. 19 · Empirical Swell History · Revised 27 September 2026

Was the June 2026 swell the biggest in a decade at Saladita?

Every June from 2014 to 2026 in ERA5-Wave reanalysis at the nearest ERA5-Wave grid point (17.5°N, 102.0°W, about 45 km offshore to the south-west) — offshore forcing, not conditions at the break — each peak matched against the hurricane best-track record. The honest answer still breaks from the headlines, but for a different reason than this page first gave.

Revised 27 September 2026 The June 2026 version of this page read NOAA WaveWatch3 at 17.5°N, 101.5°W, a cell about 40 km south-east of the break, and called the 2015 and 2018 peaks pure south-hemisphere groundswell. Rerun on ERA5-Wave at the corrected offshore node, with every peak day checked against the NHC HURDAT2 best track, the numbers are smaller and the story changes: two of the three bigger Junes were hurricane or tropical-storm swell, and the 2015 day had Hurricane Blanca 623 km offshore. The finding that June 2026 was not the biggest June since 2014 at Saladita survives. So does the finding that Marie 2014 was not the high-water mark here. Old figures are listed at the foot of the page.
Finding At Saladita, the June 2026 event (51.9 kW/m on 8 June) was not the biggest June since 2014. Three Junes had bigger peaks: 2018 (93.8), 2025 (79.0) and 2015 (54.7). Two of those three were tropical-cyclone swell, not south-hemisphere groundswell.
Why it matters The "biggest in a decade / since Hurricane Marie 2014" headlines were about California. At this point on Mexico's south coast, the biggest days of the year are usually a storm passing within a few hundred kilometres, not a distant groundswell. Marie's biggest day here was 53.6 kW/m, and it was not even the biggest day of 2014: Tropical Storm Polo gave 63.2 kW/m on 17 September.
How we know Hourly ERA5-Wave reanalysis (Open-Meteo, models=era5_ocean) requested at the break, 17.8375°N, 101.7606°W, resolved by Open-Meteo to the nearest 0.5° ERA5-Wave node, 17.5°N, 102.0°W, for every June 2014–2026 and the Marie window. Peak wave power per local day from 0.49·Hs²·T. Each peak day matched to the nearest NHC HURDAT2 storm fix within 1,000 km and ±1 day.

In early June 2026, surf media called it "the biggest south swell of the decade". At The Wedge and Malibu, that may well be true. But La Saladita sits at 17.8°N on Mexico's south Pacific coast, where the hardest days usually come from storms passing close by. Three Junes in the last twelve delivered more peak power to this point than 8–9 June 2026 did, and two of them were cyclone swell.

01 The empirical record

Top June day, each year 2014–2026

Peak hourly wave power in each local day (kW per metre of wave crest, from significant height and wave period via the deep-water energy-flux formula), the ERA5-Wave offshore node. The storm column is the nearest best-track fix within 1,000 km and one day of the peak. A day is counted as cyclone swell when that storm is within 400 km or the swell direction points within 60° of it.

YearPeak dayHsTDirectionkW/mNearest storm
2018Jun 104.26 m10.6 s169°93.8 ← recordHurricane Bud, 374 km, 65 kt · cyclone swell
2025Jun 144.22 m9.1 s157°79.0TS Dalila, 283 km, 50 kt · cyclone swell
2015Jun 5–63.20 m10.9 s207°54.7Hurricane Blanca, 623 km, 85 kt · storm-influenced
2026Jun 82.78 m13.7 s220°51.9 ← the headline eventnone · groundswell
2016Jun 242.70 m14.5 s205°51.8none · groundswell
2021Jun 273.30 m9.2 s210°48.8TS Enrique, 283 km, 55 kt
2022Jun 153.34 m7.9 s144°43.2Hurricane Blas, 301 km, 75 kt
2014Jun 303.14 m7.8 s145°37.4TS Elida, 161 km, 45 kt
2023Jun 62.12 m13.0 s202°28.6none
2020Jun 42.20 m12.1 s213°28.6none
2017Jun 192.18 m12.1 s200°28.2none
2019Jun 252.36 m8.9 s165°24.3TD Alvin, 421 km, 25 kt
2024Jun 231.98 m11.1 s194°21.3none

The 2026 event ranks fourth of thirteen Junes, a tenth of a kW/m ahead of 2016, and the biggest of the thirteen June peaks with no storm in range. That is the honest version of the headline: at Saladita it may have been the biggest groundswell June peak since at least 2014, and it was not the biggest June day.

2015 is not storm-free. The peak ran from 23:00 on 5 June to 00:00 on 6 June local time (49.7 kW/m on the 5th, 54.7 on the 6th) from 206–207° at about 10.8 s. Hurricane Blanca (85 kt) was 623 km to the south-west, and the swell direction sits within 20° of the bearing to it. The first version of this page called this "pure SH groundswell, no hurricane involved". Some of the energy may well be southern-hemisphere groundswell, but one wave period and one direction cannot split the two, and a hurricane 600 km away cannot be ruled out.

02 Hurricane Marie 2014, the original "biggest since"

What Marie actually did at Saladita.

Hurricane Marie 2014 is the canonical "biggest south swell" reference in California surf media: a Category 5 in late August whose swell flooded Malibu and lit up The Wedge. At the ERA5-Wave offshore node, Marie's biggest day came before it became the famous storm: 23 August, 53.6 kW/m (Hs 3.64 m, T 8.2 s, from 141°), as it strengthened from tropical storm to Category 1 hurricane 530–565 km to the south and south-west. On 24–25 August, as Marie intensified to Category 5 some 800–1,200 km to the west, Saladita read 32–34 kW/m from 207–209°. That puts Marie:

The headline that doesn't survive The "biggest at Saladita since Marie 2014" framing assumes Marie was the high-water mark. At this point on the coast it was not even the biggest storm of its own season. Marie's legacy belongs to California; at Saladita the big days come from storms that pass close.
03 Counter-insights from the same data

The swell pocket is about where the engine puts it

Of 1,450 June hours at or above 20 kW/m in 2014–2025, 52% came from 200–220° and the median direction was 202°. The first version of this page read a 207° centre from the old cell and called it a calibration win. At the corrected offshore node the forecast engine's 195–205° centreline holds.

Withdrawn: the 14–18 s period claim

The first version said 74% of big June hours had 14–18 s peak periods. That came from WaveWatch3 peak period at the old cell. ERA5's period here is a mean period (median 11.8 s in the same hours), which is not comparable, so the claim is neither confirmed nor refuted and is withdrawn.

Counter: storms own Saladita's biggest single days

In 47 years (1979–2025) the biggest day of the year matched a tropical cyclone in 34 years, a distant storm with the direction off in 3, and no storm at all in only 10. The record is Hurricane Calvin, 6–7 July 1993, 129 km offshore. Hurricane John on 26 September 2024 gave 101.3 kW/m. See the 47-year wave power record.

But groundswell Junes last longer

June 2015 had four straight days above 42 kW/m (5–8 June), and June 2021 five days above 40. Cyclone spikes are brief: 2018's Bud swell stayed above 56 kW/m for three days, 2025's Dalila above 50 kW/m for two. The first version's "five consecutive days above 60 kW/m" in 2015 does not hold at this node.

Counter: half the Junes were duds

Six of the twelve Junes from 2014 to 2025 (2014, 2017, 2019, 2020, 2023, 2024) had no day at or above 40 kW/m. A big June is not the default. The first version's comparison of 2015's median with 2023's maximum does not hold at this node (13.6 against 28.6 kW/m).

Withdrawn: the 21.7-second 2016 swell

The 21.7 s period for 23–25 June 2016 was a WaveWatch3 peak period at the old cell. ERA5's mean period at the 2016 peak was 14.5 s. The swell was long and from the south-west (205°, no storm within 1,000 km), consistent with a distant southern-hemisphere source, but the 21.7 s figure and the distance inferred from it are withdrawn.

04 What this means for the forecast engine

The first version of this page proposed three calibration changes for the live forecast at lasaladita.com/forecast. At the corrected offshore node:

  1. Do not shift the pocket centreline. The median big-June direction is 202°, inside the engine's 195–205° band. The 207° figure was an old-cell artifact.
  2. The period band is untested. Testing the "ideal 12–14 s" band needs peak period at the offshore node. ERA5's mean period cannot settle it.
  3. Surface multi-day pulses. This one survives. Groundswell Junes deliver energy over several days, while cyclone swell spikes for one or two, so ranking pulses by integrated energy across their window still makes sense.
Honest limit on this analysis ERA5-Wave is a reanalysis at a coarse ocean cell, not a buoy at the break. The storm match is a nearest-fix rule inside 1,000 km and ±1 day, so it can tell storm-influenced from storm-free but cannot split a mixed sea into its sources. The longer question ("biggest in 20 years", or whether the biggest days are getting bigger) is taken up on the wave power trend page, which runs the same series back to 1979.
05 Methodology
Data sources. Wave power formula. Deep-water energy flux: P = (ρ·g² / 64π) · Hs² · T ≈ 0.49 · Hs² · T, in kW per metre of wave crest. The wave power trend page rounds the coefficient to 0.50, so its figures run 2% higher (its 1993 record hour reads 232.4 kW/m against 227.7 here).

Storm matching. For each peak day, the nearest HURDAT2 fix within 1,000 km of the break and within one day of the peak. Cyclone swell: storm within 400 km, or the swell direction within 60° of the bearing to the storm. Groundswell: no storm fix within 1,000 km and ±1 day. The rule was set after looking at the data, so treat it as a description, not a test.

What this isn't. Peak wave power says nothing about ride quality. A 50 kW/m wall-up day is not the same as a 50 kW/m peeling day. Wave shape depends on direction relative to the reef, wind, tide, and how the swell stacks with secondary trains. This page is about energy, not experience.

Superseded figures (June 2026 version, WaveWatch3 at 17.5°N, 101.5°W). 2015-06-05 110 kW/m “pure SH groundswell, no hurricane” → 49.7–54.7 kW/m on 5–6 June, Blanca at 623 km. 2018-06-12 98 kW/m “SH groundswell” → 93.8 kW/m on 10 June, Bud at 374 km, cyclone swell. 2025-06-15 99 kW/m → 79.0 kW/m on 14 June, Dalila at 283 km, cyclone swell. Marie 24 August 76.6 kW/m → 23 August 53.6 kW/m. June 2026 “about 80–85 kW/m equivalent, rank 5–7” → 51.9 kW/m, rank 4 of 13. The 2022–2026 Open-Meteo cross-check table used the same old cell and has been removed.

Published 2026-06-13. Revised 2026-09-27: rerun on ERA5-Wave at the corrected offshore node with best-track storm matching.