2016年熊本地震: 地震ハザード規則を書き直した二重本震

April 14 and 16: Two Destructive Earthquakes in 28 Hours

Japan is one of the most seismically active countries on Earth, and its residents and scientists are accustomed to earthquake sequences. The standard model — a large 本震一連の地震活動の中で最大規模の地震で、その地震活動全体の規模を決定づける。前震（発生する場合もある）に先行され、余震（必ず発生する）が続く。 followed by progressively smaller 余震本震と同じ断層領域で本震の後に続く、より規模の小さい地震。余震活動は数週間から数年続くことがあり、最大の余震は通常本震よりマグニチュードが1.0〜1.2小さい。s — is so well established that it forms the basis of Japan's earthquake communication system, which after any significant event issues probability estimates for future strong shaking based on this decay model.

The Kumamoto earthquake sequence of April 2016 violated the standard model in a way that surprised even veteran Japanese seismologists.

On April 14, 2016, at 9:26 PM local time, a magnitude 6.2 earthquake struck the Kumamoto region of Kyushu, Japan's southernmost major island. Japan Meteorological Agency classified it as a large earthquake, warned of potential 余震本震と同じ断層領域で本震の後に続く、より規模の小さい地震。余震活動は数週間から数年続くことがあり、最大の余震は通常本震よりマグニチュードが1.0〜1.2小さい。s, and the standard machinery of emergency response activated. Nine people died, more than 1,000 were injured, and tens of thousands left their homes either voluntarily or under evacuation order. Emergency shelters filled with frightened residents who had been told the worst was probably over.

Twenty-eight hours later, at 1:25 AM on April 16, an earthquake of magnitude 7.3 struck almost the same area. This was significantly larger than the April 14 event — roughly eight times more energetic. The ground shaking exceeded that of the first event, causing fresh building collapses in structures that had been damaged but not destroyed, and devastating areas that had survived the first earthquake largely intact. The total death toll from the sequence reached 50 direct deaths, with more than 3,000 injuries. Over 44,000 buildings were damaged or destroyed.

The seismological puzzle was immediately apparent: had the April 14 event been a 前震同じ地域で本震に先立って発生する地震。前震は事後にしか特定できず、事前に通常の地震と区別する確実な方法は存在しない。 or a 本震一連の地震活動の中で最大規模の地震で、その地震活動全体の規模を決定づける。前震（発生する場合もある）に先行され、余震（必ず発生する）が続く。? And what did the reclassification tell us about the limits of earthquake prediction and the design of 地震警報システム1991年から運用されている、世界初の公共向け緊急地震速報システムの一つであるメキシコのSASMEX。沿岸部の地震から、メキシコシティに最大60秒の警報時間を提供する。 communications?

Use Earthquake Energy Calculator to compare the energy release of the April 14 M6.2 versus the April 16 M7.3. Use Distance from Epicenter to understand how the close 震央地下で地震が発生した震源の真上にあたる地表の地点。ニュース報道では地震の発生場所としてよく報じられる。 proximity of both events contributed to the cumulative damage.

The Futagawa-Hinagu Fault System: Kyushu's Hidden Hazard

Kyushu island sits in a geologically complex position. The subduction of the Philippine Sea Plate beneath the southwestern tip of Japan has loaded the region with compressional stress, but the interior of Kyushu is simultaneously being pulled apart along a series of northeast-trending grabens — rift-like valleys where the crust is extending. This combination of regional compression and local extension creates a system of active 横ずれ断層岩盤のブロックが水平方向に互いにすれ違う断層。サンアンドレアス断層と北アナトリア断層は、破壊的な地震を引き起こす主要な横ずれ断層である。s and normal faults across the island.

The 2016 earthquakes occurred on the Futagawa-Hinagu fault system, a northeast-trending right-lateral 横ずれ断層岩盤のブロックが水平方向に互いにすれ違う断層。サンアンドレアス断層と北アナトリア断層は、破壊的な地震を引き起こす主要な横ずれ断層である。 that passes directly beneath the city of Kumamoto and extends into the Aso volcanic caldera to the northeast. The system consists of multiple 断層セグメント特有のすべり挙動を持つ、より大きな断層系の中の明確な区間。異なるセグメントが独立して、あるいは連鎖的に破壊することがあり、地震の規模に影響する。s, of which the Futagawa segment (approximately 64 km) and the Hinagu segment (approximately 81 km) are the primary active branches.

Before 2016, the Futagawa-Hinagu system was recognized as active on Japan's national active fault maps, and the relevant 地震ハザードマップ特定の期間内に地震の揺れが指定の水準を超える確率を示した地図。技術者・都市計画者・保険会社が地震リスクを評価するために用いる。 had assessed it as potentially capable of generating an M7.1–7.4 earthquake. In this sense, the hazard was identified. What was not adequately anticipated was the multi-segment, multi-event nature of the rupture — specifically, that the system would produce two major earthquakes in rapid succession from different 断層セグメント特有のすべり挙動を持つ、より大きな断層系の中の明確な区間。異なるセグメントが独立して、あるいは連鎖的に破壊することがあり、地震の規模に影響する。s.

The April 14 earthquake primarily ruptured the Hinagu segment, while the April 16 mainshock broke the Futagawa segment to the northeast. This spatial progression, combined with the timing, suggests that the first rupture transferred stress — a process known as クーロン応力伝達地震が周辺の断層の応力状態を変化させ、将来の地震を誘発または遅延させうる過程。どの断層が破壊に近づいたかを予測するために用いられる。 loading — onto the adjacent Futagawa segment, ultimately triggering the larger second event. The mechanics are well understood in principle, but applying this understanding in real time to issue warnings about impending larger earthquakes remains a challenge at the frontier of operational seismology.

Detailed geodetic analysis using satellite interferometry (InSAR) mapped the ground deformation pattern of both events with high precision in the weeks following the sequence. The combined deformation extended over an area of several thousand square kilometres, with maximum surface displacements exceeding one metre near the fault traces. This detailed deformation mapping constrained the fault geometry and slip distributions in ways that would inform 地震ハザードマップ特定の期間内に地震の揺れが指定の水準を超える確率を示した地図。技術者・都市計画者・保険会社が地震リスクを評価するために用いる。 updates for the Futagawa-Hinagu system and analogous fault systems throughout Japan.

Foreshock or Mainshock? Why Classification Matters

The question of whether the April 14 M6.2 was a 前震同じ地域で本震に先立って発生する地震。前震は事後にしか特定できず、事前に通常の地震と区別する確実な方法は存在しない。 to the April 16 M7.3 本震一連の地震活動の中で最大規模の地震で、その地震活動全体の規模を決定づける。前震（発生する場合もある）に先行され、余震（必ず発生する）が続く。, rather than a 本震一連の地震活動の中で最大規模の地震で、その地震活動全体の規模を決定づける。前震（発生する場合もある）に先行され、余震（必ず発生する）が続く。 in its own right, is more than academic. It has profound practical implications for how authorities communicate risk and how the public responds.

When Japan Meteorological Agency classified the April 14 event as a 本震一連の地震活動の中で最大規模の地震で、その地震活動全体の規模を決定づける。前震（発生する場合もある）に先行され、余震（必ず発生する）が続く。 and began issuing standard 余震本震と同じ断層領域で本震の後に続く、より規模の小さい地震。余震活動は数週間から数年続くことがあり、最大の余震は通常本震よりマグニチュードが1.0〜1.2小さい。 forecasts, they were following established protocols built on the well-known statistical pattern that aftershocks are always smaller than the mainshock. This pattern — encapsulated in Bath's Law, which states that the largest aftershock is typically about 1.2 magnitude units smaller than the mainshock — applies to the vast majority of earthquake sequences. A M6.2 mainshock would be expected to generate aftershocks no larger than about M5.0.

When the M7.3 earthquake struck 28 hours later, it could not be called an 余震本震と同じ断層領域で本震の後に続く、より規模の小さい地震。余震活動は数週間から数年続くことがあり、最大の余震は通常本震よりマグニチュードが1.0〜1.2小さい。 under any conventional definition — it was larger than the event it supposedly followed. Seismologists quickly reclassified the April 14 event as a 前震同じ地域で本震に先立って発生する地震。前震は事後にしか特定できず、事前に通常の地震と区別する確実な方法は存在しない。 and the April 16 event as the true 本震一連の地震活動の中で最大規模の地震で、その地震活動全体の規模を決定づける。前震（発生する場合もある）に先行され、余震（必ず発生する）が続く。. But for the thousands of people who had returned to their damaged homes after being reassured that the worst was over, this reclassification was cold comfort. Some of the fatalities from the April 16 earthquake involved people who had returned to buildings that should not have been reoccupied.

The Kumamoto sequence ignited an intense debate about how to communicate earthquake hazard in the immediate aftermath of a significant event. The probability that any given earthquake will be followed within a few days by a larger event is statistically small — typically one to five percent — but not negligible. The challenge is communicating this small but non-trivial probability to a public and to officials who must make real-time decisions about evacuation, building reoccupation, and resource deployment.

Japan Meteorological Agency revised its aftershock forecast communication protocols as a direct result of the Kumamoto experience. Rather than implying that the probability of a larger event was negligible, the revised communications acknowledge that while unlikely, a larger earthquake cannot be ruled out in the days following a significant event. This shift — toward expressing genuine uncertainty rather than conveying false confidence — represents an important evolution in earthquake risk communication with global implications for how similar agencies communicate after major events.

Kumamoto Castle: 400 Years of Heritage Destroyed

Kumamoto Castle is one of Japan's three premier feudal castles, originally constructed between 1601 and 1607 under the warlord Kato Kiyomasa. Its elaborate stone foundations, massive earthen ramparts, and multi-tiered main tower have made it one of the most architecturally sophisticated defensive structures from the Edo period. The castle had survived numerous historical conflicts, including the 1877 Satsuma Rebellion, and had been partially rebuilt in concrete in 1960.

The 2016 earthquakes damaged Kumamoto Castle severely. Stone walls that had stood for four centuries collapsed in sections. The roofs of multiple subsidiary structures were damaged or destroyed. The main tower itself, constructed of reinforced concrete, survived but exhibited structural damage requiring years of repair. In total, more than 3,500 of the castle's stone wall sections required repair or reconstruction.

The castle's resilience — and its partial failure — provides a case study in historical construction and seismic performance. The massive stone retaining walls, built using a technique called "nozurazumi" that uses uncut natural stones fitted together without mortar, are inherently flexible and in many cases performed reasonably well. The sections that collapsed were often those that had been compromised by vegetation growth in the joints, subsurface drainage issues, or previous interventions that had altered the original construction. The 耐震基準建物の最低限の耐震安全性を確保するための、設計・建設に関する法的要件の体系。大地震で新たな脆弱性が明らかになるたびに更新される。s governing restoration mandated modern seismic performance standards while attempting to preserve authentic construction techniques.

The castle became a symbol of Kumamoto's determination to rebuild, and its phased restoration — estimated to take 20 years and cost approximately 63 billion yen — has been the subject of intensive documentation and public education about both historical construction methods and modern earthquake engineering. The process of restoring the walls while maintaining their traditional appearance required collaboration between archaeologists, historians, structural engineers, and master stonemasons — an interdisciplinary effort that itself generated new knowledge about traditional Japanese castle construction.

Landslides in the Aso Caldera: Volcanic Terrain Amplification

The M7.3 mainshock triggered extensive 地震誘発地すべり地震の揺れによって引き起こされる、土砂や岩石の斜面下方への移動。地すべりは地域全体を埋没させることがあり、揺れそのものより多くの犠牲者を出すこともある。s across the Kumamoto region, with the most dramatic occurring in and around the Aso volcanic caldera — the collapse structure of the ancient Aso Volcano, approximately 25 kilometres east of Kumamoto city. The caldera rim and its flanks consist of volcanic ash deposits, pyroclastic flows, and weathered volcanic rock — materials that are inherently weak when saturated and prone to slope failure when shaken.

The Aso Ohashi Bridge, a major highway bridge over the Kurokawa River on National Route 57, was destroyed when a massive slope failure on the caldera rim overtopped and crushed the eastern abutment. The 地震誘発地すべり地震の揺れによって引き起こされる、土砂や岩石の斜面下方への移動。地すべりは地域全体を埋没させることがあり、揺れそのものより多くの犠牲者を出すこともある。 volume was estimated at several million cubic metres. The failure blocked the main road through the caldera, disrupting transportation links for months and effectively isolating communities in the caldera interior. A new bridge — the Shin-Aso Ohashi — was eventually constructed as part of the reconstruction effort.

The amplification of ground shaking in the volcanic terrain around Aso was significant. Thick sequences of soft volcanic deposits can dramatically amplify shaking in the same way that alluvial sediments do, concentrating energy in particular frequency bands. This 地盤増幅（サイト効果）軟弱な土壌や堆積層が地震波を増幅させることによって生じる、揺れの強さの増大。軟弱地盤上の構造物は、基盤岩上の構造物に比べて2〜10倍強い揺れを経験することがある。 in volcanic terrain is systematically less well characterised than amplification in alluvial settings, partly because the deposits are more heterogeneous and partly because there are fewer strong-motion recordings in these environments. The Kumamoto dataset provided an opportunity to improve seismic amplification models for volcanic terrain.

The Aso region also experienced increased volcanic unrest in the days following the 本震一連の地震活動の中で最大規模の地震で、その地震活動全体の規模を決定づける。前震（発生する場合もある）に先行され、余震（必ず発生する）が続く。. The Nakadake crater of Aso Volcano, which had been episodically active, exhibited elevated activity — a reminder that the region sits at the intersection of tectonic and volcanic hazards. The relationship between seismicity and volcanic system pressurization in caldera environments is an active research area, and the 2016 sequence contributed new observational data on this interaction.

Rethinking Earthquake Sequences: Lessons for Early Warning

The 2016 Kumamoto earthquakes produced lasting changes in how Japan's seismological agencies communicate with the public about earthquake sequences. The experience of the multi-本震一連の地震活動の中で最大規模の地震で、その地震活動全体の規模を決定づける。前震（発生する場合もある）に先行され、余震（必ず発生する）が続く。 sequence — two major earthquakes from adjacent 断層セグメント特有のすべり挙動を持つ、より大きな断層系の中の明確な区間。異なるセグメントが独立して、あるいは連鎖的に破壊することがあり、地震の規模に影響する。s within 28 hours — catalyzed a review of the statistical frameworks used to forecast 余震本震と同じ断層領域で本震の後に続く、より規模の小さい地震。余震活動は数週間から数年続くことがあり、最大の余震は通常本震よりマグニチュードが1.0〜1.2小さい。 sequences and the language used to describe earthquake hazard immediately after major events.

The sequence provided high-quality data on クーロン応力伝達地震が周辺の断層の応力状態を変化させ、将来の地震を誘発または遅延させうる過程。どの断層が破壊に近づいたかを予測するために用いられる。 transfer between adjacent fault segments, contributing to the growing body of evidence that stress changes caused by one earthquake can advance or delay rupture on nearby faults. This physical mechanism is well understood theoretically but difficult to use in operational forecasting because the relevant fault geometry and stress state are rarely known with sufficient precision. The Kumamoto data, with its detailed instrumental records of both events and the subsequent 余震本震と同じ断層領域で本震の後に続く、より規模の小さい地震。余震活動は数週間から数年続くことがあり、最大の余震は通常本震よりマグニチュードが1.0〜1.2小さい。 sequence, constrained models of stress transfer in ways that will inform future research.

The behaviour of Japan's 緊急地震速報（EEW）地震を検知し、強い揺れが到達する前に人々やシステムに警報を送るシステム。数秒から数十秒の猶予を提供し、身を守る行動をとるのに十分な時間となる。 system during the sequence was also scrutinised. The system issued P-wave based warnings for both events, but the close 震央地下で地震が発生した震源の真上にあたる地表の地点。ニュース報道では地震の発生場所としてよく報じられる。 proximity to Kumamoto city meant that warning times were very short — a few seconds at most. For populations within 20 to 30 kilometres of a fault capable of producing M7+ earthquakes, the physics of wave propagation simply do not allow enough time for effective warning-based action. This "blind zone" around active faults is a fundamental limitation of seismic early warning that the Kumamoto sequence illustrated with exceptional clarity.

The Kumamoto earthquakes of 2016 added a significant and enduring chapter to the global literature on complex earthquake sequences, fault system behaviour, and the profound challenges of seismic communication in the immediate aftermath of damaging events. Their legacy is embedded in revised warning protocols, improved 地震ハザードマップ特定の期間内に地震の揺れが指定の水準を超える確率を示した地図。技術者・都市計画者・保険会社が地震リスクを評価するために用いる。s for the Futagawa-Hinagu system, and a more nuanced public and professional understanding of the limits of earthquake science in the hours immediately following a damaging event. In particular, the sequence made it impossible to maintain the fiction that the largest earthquake in any sequence is always the first one — a lesson with direct operational implications for every seismically active country in the world.

The Aso Region: Volcanic and Seismic Hazard Intersection

The 2016 Kumamoto sequence illuminated a category of compound hazard that is specific to volcanic island arc settings: the intersection of seismic and volcanic hazard systems that share geographic space and, in some cases, share physical coupling mechanisms. Kyushu is one of the most volcanically active regions of Japan, with Aso, Sakurajima, Unzen, and Kirishima all having erupted within living memory. The 横ずれ断層岩盤のブロックが水平方向に互いにすれ違う断層。サンアンドレアス断層と北アナトリア断層は、破壊的な地震を引き起こす主要な横ずれ断層である。 systems that generated the 2016 earthquakes exist within a broader regional stress field that also drives magma migration and volcanic activity.

The question of whether the 2016 earthquakes had any effect on the volcanic system beneath Aso was actively investigated by volcanologists in the weeks following the sequence. Elevated SO₂ emissions from the Nakadake crater and a small ash emission event in the days after the mainshock suggested some response of the volcanic system, though scientists were careful to distinguish these observations from evidence of an imminent eruption. The Aso Volcano Observatory, which monitors the caldera continuously, noted that the seismic sequence had temporarily perturbed the hydrothermal system feeding the crater lake.

Research on the coupling between seismic and volcanic systems in volcanic island arc settings has been energized by the Kumamoto sequence, contributing to a growing body of evidence that large earthquakes can sometimes trigger or modulate volcanic activity at distances of tens of kilometres. The physical mechanisms — including changes in pore fluid pressure, dynamic stress changes from passing 地震波地震や爆発によって発生し、地球内部を伝播する弾性波。地震波は、震源で放出されたエネルギーを遠方の地点まで運ぶ。s, and static stress changes from fault displacement — are well understood in principle but difficult to quantify for specific volcanic systems. The Kumamoto dataset has provided one of the most precisely characterized examples of near-field seismic-volcanic interaction, contributing to improved monitoring protocols for volcanic systems in seismically active environments.

Reconstruction and Economic Impact

The total economic loss from the 2016 Kumamoto earthquake sequence was estimated at approximately 4.6 trillion yen (roughly 44 billion US dollars at 2016 exchange rates), making it one of the costliest earthquake sequences in Japanese history. The damage extended well beyond the immediately affected areas: the temporary closure of major Toyota and Honda manufacturing plants — located in the Kumamoto region because of its position in Japan's automotive supply chain — disrupted vehicle production nationally and internationally, as just-in-time supply chains proved vulnerable to the geographic concentration of component manufacturing.

The disruption of the Aso Caldera highway — one of only two major road connections between Kumamoto and the Oita region on Kyushu's eastern coast — had cascading economic effects on tourism, freight, and emergency access for months. The reconstruction of the Aso Ohashi Bridge and the bypasses around the 地震誘発地すべり地震の揺れによって引き起こされる、土砂や岩石の斜面下方への移動。地すべりは地域全体を埋没させることがあり、揺れそのものより多くの犠牲者を出すこともある。 zones occupied civil engineers and construction workers for nearly two years, at costs substantially exceeding initial estimates because of the complexity of working in active volcanic and seismically unstable terrain.

Kumamoto's Contribution to Building Performance Research

The 2016 earthquakes provided an exceptionally rich dataset for studying the seismic performance of buildings designed to Japan's modern earthquake-resistant 耐震基準建物の最低限の耐震安全性を確保するための、設計・建設に関する法的要件の体系。大地震で新たな脆弱性が明らかになるたびに更新される。 — one of the strictest in the world. Japan had experienced multiple major earthquakes in the preceding decades, each one providing data that had been incorporated into successive code revisions. The Kumamoto performance data allowed researchers to assess the cumulative effect of these incremental improvements.

The findings were broadly positive but not uniformly so. Buildings designed to the current version of the code (revised most recently after the 2011 Tohoku earthquake) generally performed well, with few structural collapses. Buildings designed to codes from the 1970s and early 1980s — before major revisions following the 1978 Miyagi-ken Oki earthquake — showed significantly higher damage rates. Buildings in the intermediate generation — designed to the revised 1981 code but not yet to the most current standards — showed moderate damage rates, confirming the value of each successive code revision while also demonstrating that some residual vulnerability remained in the older modern building stock.

This "generation gap" in building performance — where each successive code revision produces measurably safer buildings but leaves a stock of older buildings with higher vulnerability — is a universal challenge in earthquake engineering, and the Kumamoto data has provided one of the clearest empirical demonstrations of its magnitude. It has strengthened arguments in Japan and internationally for accelerating the retrofitting of buildings designed to older codes, even when those buildings appear superficially sound.

The data from Kumamoto also contributed to a revision of Japan's probabilistic 地震リスク評価特定の地域や構造物について、地震ハザード・建物の脆弱性・想定される損失を評価する過程。ハザードマップ、建物台帳、被害モデルを組み合わせて行われる。 for Kyushu, which had previously assigned relatively lower probabilities to large events on the Futagawa-Hinagu system compared with other major fault systems in Japan. The multi-event sequence and its クーロン応力伝達地震が周辺の断層の応力状態を変化させ、将来の地震を誘発または遅延させうる過程。どの断層が破壊に近づいたかを予測するために用いられる。 transfer mechanism demonstrated that the system was capable of more complex behavior than single-segment rupture scenarios had anticipated, and this complexity was incorporated into updated national hazard assessments.

Japan's national programme for seismic retrofitting of existing buildings has been in place since 1995, when it was established following the 1995 Kobe earthquake. The programme provides subsidies and technical assistance for voluntary building upgrades, and has been progressively strengthened through incremental legislation. The 2016 Kumamoto data on building performance provided fresh evidence for the economic as well as humanitarian justification for the programme: buildings that performed poorly in 2016 had to be demolished or extensively repaired, at costs that substantially exceeded the cost of proactive retrofitting. The case for investment in seismic strengthening of the existing building stock is both a life safety argument and an economic one, and the Kumamoto experience articulated both with unusual clarity.

The 2016 sequence also generated detailed data on non-structural damage — the failure of ceilings, partitions, contents, and equipment that does not affect structural safety but causes injury, economic loss, and disruption of function in buildings that remain structurally intact. Non-structural damage in hospitals, schools, and emergency facilities can prevent these critical buildings from functioning when they are most needed, even when the structure itself has not been compromised. Post-Kumamoto research on non-structural vulnerability has highlighted this often-neglected dimension of earthquake risk and contributed to revised guidelines for securing non-structural elements in critical facilities throughout Japan.