Summary

  • The source article documents the Hōlei sea arch’s collapse between Sept. 23 and Sept. 27, 2026, and frames the event through two parallel causal accounts — a long-anticipated geological-erosion account and a storm-timing account — without specifying how the accounts combine.
  • The article introduces a climate attribution chain (strong El Niño, ocean warming, fossil-fuel combustion) inside one paragraph in which the only hedging qualifier (“high probability”) attaches to a forecaster-class attribution rather than to the chain’s contested steps.
  • The article attributes the climate framing to a generic class of “forecasters” and “experts” and does not identify a forecasting office, a named meteorologist, or a published outlook that supplies the qualifier “supercharged” or the El Niño-to-fossil-fuel claim.
  • The article labels Nolo as both “Tropical Storm Nolo” and a storm that “rapidly intensified into a hurricane” on Sept. 24, and groups Nolo with hurricanes Lala and Lowell as “the third major storm,” without specifying the threshold “major” refers to or whether Nolo counted as a hurricane during its impact on the archipelago.
  • The article brackets the collapse window between Sept. 23 and Sept. 27, identifies Nolo as the relevant storm, and does not cite a park observation, a surf log, a seismic record, or a visual documentation that ties the specific moment of failure to a particular wave, swell, or earthquake.
  • The article places damage from the 2018 Kīlauea seismic events in the causal chain as having “destabilized the surrounding cliffs,” without specifying whether that cliff-side destabilization affected the arch itself or whether the arch would have failed on a similar timescale absent the 2018 events.

The National Park Service confirmed on Friday, Oct. 2, 2026, that the Hōlei sea arch — a roughly 90-foot formation on Hawaii’s Big Island southern coast — collapsed into the Pacific Ocean between Sept. 23 and Sept. 27, 2026, after the park had closed during Tropical Storm Nolo on Sept. 23 and the formation was documented as gone by the evening of Sept. 27. The article documents the collapse through two causal accounts that run in parallel rather than as a unified explanation: a long-anticipated geological-erosion account, sourced to a 2014 U.S. Geological Survey statement and to NPS and USGS characterizations of the collapse as “long expected,” and a storm-timing account that connects the collapse to a “supercharged” El Niño season in which Nolo arrives as “the third major storm to hit Hawaii this year.” The join between the long-cycle endpoint and the proximate storm trigger is left for the reader to construct, and the climate attribution chain the article introduces — strong El Niño amplified by ocean warming attributed to fossil-fuel combustion — is reported inside one paragraph in which the only hedging qualifier (“high probability”) attaches to a forecaster-class attribution rather than to the chain’s contested steps.

Two parallel causal accounts the article does not reconcile

The article carries two causal accounts of the arch’s collapse and does not specify how the two relate. The first account is geological-inevitability, reported through the U.S. Geological Survey’s 2014 statement that “The arch formed in ancient lava flows by removing rock along cracks and weaker rocks. It will not be here for ever. Every day the ocean crashes against the pillar of the arch, eroding it a little more. One day, it will be gone,” and through the National Park Service and USGS characterizations of the collapse as having been “long expected” once the cliffs showed signs of instability. The second account is climate-attributed-storm-timing, reported through the article’s framing of Nolo as the storm that closed the park on Sept. 23, through the article’s identification of Nolo as “the third major storm to hit the Hawaiian islands in recent months,” and through the surrounding El Niño season the article describes as “supercharged.” The article does not specify whether Nolo provided the trigger that arrived at the inevitable moment, whether the arch would have fallen on a similar timescale absent Nolo, or whether the storm season is the operative cause at all. The join between the long-cycle endpoint and the proximate storm trigger is left for the reader to construct.

The article’s surface presentation foregrounds the storm-timing account by placing Nolo in the lead summary and the El Niño framing in the second paragraph, and embeds the long-anticipated framing further into the piece, where it serves as one corroborating detail among others rather than as the dominant account. The geological-inevitability account is the more fully sourced — it is backed by the 2014 USGS statement on the record and by NPS and USGS characterizations — while the storm-timing account’s connective tissue between storm activity and the specific collapse remains unspecified.

The climate paragraph: hedge asymmetry and attribution-to-class gap

Inside the article’s climate paragraph, a hedge asymmetry runs from one step of the chain to the next. The article reports that “Forecasters warn a ‘supercharged’ El Niño is driving above-average Pacific storm activity this season” — an institutional speaker named — and that “Forecasters have warned there is a high probability that this ‘supercharged’ El Niño will reach a historic intensity.” The article then states that the El Niño’s effects are “compounded by the broader overheating of the planet caused by the burning of fossil fuels” — no forecaster, no expert, no institutional attribution; the claim reads as one the piece has supplied to a paragraph otherwise populated by attributed voices.

The chain the article constructs runs: strong El Niño → El Niño amplified by ocean warming → warming attributed to fossil-fuel combustion. Each step is a claim that depends on attribution science the article does not cite. The chain is presented as a settled narrative rather than as a contested attribution, while the article’s own framing is probabilistic — the only hedge in the paragraph (“high probability”) attaches to the El Niño step, not to the ocean-warming or fossil-fuel steps. A reader is left to weight a contested causal claim alongside named-source forecasts without the article having marked the join.

The “supercharged” qualifier appears in quotation marks, suggesting the article is reproducing a forecast term rather than asserting the qualifier. The article does not, however, identify the forecasting office, the named meteorologist, or the published outlook that supplies the qualifier, and the attribution runs to a generic class of “forecasters.” A separate “Experts say the pattern is consistent with El Niño years, when the Pacific often sees more tropical storms and hurricanes” carries the same attribution-to-class pattern: no discipline, institution, or individual is named, while neighboring sentences are attributed to “Forecasters” and “the U.S. Geological Survey.” The article treats the storm-activity claim as established context, but that context is itself an attributed claim with less specific attribution than the institutional voices elsewhere in the same paragraph.

Storm labels: Nolo’s classification and the “third major storm” framing

The article’s framing of Nolo contains a definitional inconsistency the article does not resolve. The lead summary and the source’s headline subtype label Nolo as a “tropical storm”; the body reports that Nolo “rapidly intensified into a hurricane on September 24.” The article does not specify whether the documented collapse window — Sept. 23 to Sept. 27 — aligns with Nolo’s tropical-storm phase, its hurricane phase, or the period after Nolo cleared the islands. The label travels through the piece without a stable referent.

A related label-travel observation concerns the three named storms — Lala, Lowell, and Nolo. The article introduces Nolo as “Tropical Storm Nolo” and describes Nolo as having “rapidly intensified into a hurricane” on Sept. 24. Lala and Lowell are described simply as “hurricanes” without the tropical-storm-then-hurricane label travel. The article presents the three storms as commensurable “major storms” in the summary, but the body uses different storm-class labels for each. What threshold “major” refers to, and whether Nolo counted as a hurricane during its impact on the archipelago, are not stated. The underlying NHC classifications of Lala, Lowell, and Nolo at the time of their impact on the islands — and whether the article’s stylistic simplification preserves those classifications — is not visible from the substrate.

What the article does not specify

The article brackets the collapse between Sept. 23 and Sept. 27, identifies Nolo as the relevant storm, and does not cite a park observation, a surf log, a seismic record, or a visual documentation that ties the specific moment of failure to a particular wave, swell, or earthquake. The window is presented as the relevant period and Nolo’s intensification is presented as the relevant storm, but the article does not specify which of the two is the operative cause. The premise the article does not substantiate is the specific contribution of Nolo to this collapse, as distinct from the longer erosional history the 2014 USGS statement describes.

The article places damage from the 2018 Kīlauea eruption and summit collapse in the causal chain, stating that the seismic events “destabilized the surrounding cliffs.” The connection between that cliff-side destabilization and the arch’s later collapse is left implicit. The article does not say whether the cliff-side instability affected the arch itself, or whether the arch would have failed on a similar timescale absent the 2018 seismic events. The article surfaces the destabilization without measuring its weight. A USGS or NPS statement linking the arch’s 2018-present instability to the 2018 seismic event, as distinct from a general cliff-instability finding, would close the join.

The article also moves between four scales of analysis without flagging the transitions: a 550-year geological inevitability; a multi-year documented instability (2018 Kīlauea, January 2020 cracks, 2020 Darby); a four-day storm window (Sept. 23–27); and a multi-decadal climate attribution (El Niño pattern, fossil-fuel burning). Each scale is legitimate on its own terms. The article’s structure treats them as additive rather than as different units of analysis, and the reader is left with a single causal account in which geological inevitability, named storms, an El Niño pattern, and fossil-fuel combustion all act on the same arch in the same September week.

What the substrate supports and what it does not

The article documents what happened between Sept. 23 and Sept. 27 and supports the following: the park closed on Sept. 23 during Tropical Storm Nolo; Nolo rapidly intensified into a hurricane on Sept. 24; the formation was gone by the evening of Sept. 27; the park and USGS described the collapse as “long expected” given prior cliff instability and damage from the 2018 Kīlauea seismic events; January 2020 cracks prompted the park to close the original viewing area and open a safer one three months later; 2020 surf from Tropical Storm Darby broke off a chunk of rock from the arch’s lower portion; and a “supercharged” El Niño was forecast to drive above-average Pacific storm activity.

The joins the article does not specify are: the join between the long-cycle erosion account and the proximate storm trigger; the join between the named 2018 seismic destabilization and the arch’s specific collapse; the join between the asserted compounding effect of planetary overheating on this season’s storms and the collapse timing; and the join between the documented collapse window and a specific moment-of-failure observation tying the loss to Nolo. The substrate supports a more modest claim — the arch fell, the fall had been predicted, and the storm season provided the proximate occasion. It does not support a confident attribution of the timing to the specific climate mechanism the article names.

Whether the climate attribution chain is presented as more or less settled in the underlying source material than the source language suggests cannot be determined from the substrate. The source uses “high probability” for the El Niño framing; the observation that the chain is presented as a settled narrative is sourced. Whether the source reporter was drawing on a specific attribution-science document or on a generic framing is not visible from the substrate. The source article does not name a forecaster, an attribution scientist, or a forecasting office for any of the climate-attribution claims, and the attribution-to-class pattern is the only thing the substrate supports.

Analytical techniques used in this piece

This analysis applies the methods below. Each links to a short, plain-English explainer you can read and reuse.

Coherence Audit
Tests whether an argument hangs together — spotting contradictions, gaps, and circular reasoning.
Superforecasting (Tetlock)
The habits — calibration, updating, track records — that make some forecasters reliably better.