Scan: Will there be an at least 8.0 magnitude earthquake in Japan before 2030?

← Back to scans

Will there be an at least 8.0 magnitude earthquake in Japan before 2030?

KXEARTHQUAKEJAPAN-30 · Climate and Weather · 2026-05-01

42%

Agent

66%

Market Price

-24.0%

Edge

40%

Confidence

Volume: 21,591

Spread: 10.0c

Days to resolution: 1340

Markets in event: 1

Final Rationale

My independent estimate of 33% was anchored on a Poisson base rate (~25% over 3.67 years from ~10 M8+ events since 1900) with modest upward adjustment for elevated indicators. I PARTIALLY ACCEPT challenges 1-3: the base rate likely underestimates Japan-relevant events (post-2000 rate is higher), the confluence of independent elevated signals (Nankai caution, Dec 2025 advisory, Shimazaki ~2030, Kamchatka stress transfer) deserves more weight, and multiple independent M8+ pathways (Nankai, Japan Trench outer-rise, Kuril, Hokkaido) compound rather than substitute. I REJECT full convergence to the market's 66%: the orderbook shows a $0.97 spread with inconsistent bid data, suggesting the posted price is not a true equilibrium and reflects post-Dec-2025 recency bias rather than informed consensus. Settling at 42% reflects the critic's +6pp adjustment plus modest additional credit for pathway aggregation, while maintaining strong skepticism of a market price that implies a hazard rate ~3x the historical Poisson baseline.

Price vs. prediction

How the market moved after the agent's call. Solid blue line is the actual market price; dashed line is the agent's static forecast at $t{=}0$.

$n = 58$ follow-ups

Re-scan Context

This market has been scanned before. Previous predictions:

Date	Predicted	Market Price	Confidence
2026-04-05	43%	54%	45%

1. Planner Opus · Fermi decomposition + tool plan

Tools chosen: kalshi_data kalshi_orderbook wikipedia_lookup article_search web_search code_execution

Sub-questions (Fermi decomposition)

Based on historical frequency, will Japan experience at least one M8.0+ earthquake in any given 3.7-year window? (w=0.50)
Are current seismic indicators for Japan elevated, suggesting higher-than-baseline probability of M8.0+ in 2026-2029? (w=0.30)
Will the Nankai Trough megaquake (anticipated M8-9 event) occur before January 1, 2030? (w=0.20)

Planner reasoning

This question requires combining historical base rates with known fault-specific risk assessments. Japan sits at the confluence of 4 tectonic plates making it among the world's most seismically active regions. The analytical framework: (1) Model M8.0+ occurrence as a Poisson process using historical frequency since 1900 - approximately 8-10 qualifying events in 126 years yields an annual rate of ~0.063-0.079/year, giving roughly 21-25% probability over the 3.67-year window. (2) Layer on the Nankai Trough specific risk - the Japanese government publishes 70-80% probability of an M8-9 Nankai Trough event within 30 years from around 2013, implying elevated near-term risk. (3) Consider whether the 2011 Tohoku M9.0 relieved or redistributed stress to adjacent zones. The temporal dynamics matter: for YES to resolve, a single M8.0+ event with epicenter in Japan or territorial waters must occur in the next 3.67 years. Japan experiences M7+ events routinely so the threshold question is energy accumulation to M8. The Nankai Trough is the primary named threat. The prediction market price will serve as a strong prior given this is a quantifiable physical process with active expert monitoring.

2. Research Parallel tool calls · legacy v2

Tool	Status	Time	Items
kalshi_data	OK	2.1s	-
kalshi_orderbook	OK	0.2s	-
wikipedia_lookup	OK	2.2s	-
article_search	OK	0.8s	-
web_search	OK	53.0s	-
code_execution	OK	0.2s	-

3. Evidence Synthesis Sonnet · 14 items · legacy v2

#	Strength	Credibility	Direction	Source	Claim	Priced In
1	STRONG	75	↓ DOWN	code_execution	Using a Poisson process with historical rate of ~0.0635 M8.0+ earthquakes/year in Japan, the probability of at least one such event in a 3.67-year window is approximately 20.8% (low estimate) to 25.3% (high estimate).	Yes
2	STRONG	65	↑ UP	kalshi_data	The Kalshi prediction market prices this question at 66%, a significant premium over the Poisson base rate of ~21-25%, with a +18% move over the past 30 days suggesting recent information has shifted trader sentiment upward.	No
3	MODERATE	55	NEUTRAL	kalshi_orderbook	The orderbook shows extremely wide spread ($0.01 yes bid vs $0.02 no bid on a $0.97 spread) with high depth (29,739), suggesting the displayed 66% price may reflect thin liquidity and should be interpreted cautiously.	Yes
4	WEAK	45	NEUTRAL	wikipedia_lookup	Wikipedia's list of Japan earthquakes confirms Japan is among the world's most seismically active regions, though the raw article data provided no specific count of M8.0+ events since 1900 to validate the ~8-10 event historical frequency used in the Poisson calculation.	Yes
5	STRONG	82	↑ UP	article_search	In December 2025, Japan issued a megaquake advisory after a M7.5 earthquake struck northeastern Japan, with officials noting only ~1% probability of M8+ in the following week but warning of elevated general risk and calling it a 'wake-up call.'	Yes
6	STRONG	88	↑ UP	web_search	Japan's Earthquake Research Committee (January 2025) raised the Nankai Trough megaquake probability to 'around 80%' within the next 30 years, up from 70-80%, with risk rising approximately 1% per year.	Yes
7	STRONG	90	↑ UP	web_search	Following a Hyūga-nada earthquake in August 2024, the JMA issued an unprecedented 'caution' warning about elevated Nankai Trough risk — the first such warning in Japanese history — though it stressed a quake was not imminent.	Yes
8	MODERATE	72	↑ UP	article_search	The Tokara Islands experienced over 1,800 earthquakes in three weeks (July 2025), and the Russian Kamchatka M8.8 quake in July 2025 sent tsunamis to Japan — indicating elevated regional seismic activity in the broader Pacific Ring of Fire context.	Yes
9	STRONG	85	↑ UP	article_search	A BBC article from December 2025 states Japan's earthquake investigation panel assigned a 60-90% probability of a Nankai Trough megaquake within the next 30 years, confirming elevated but long-horizon risk framing from official sources.	Yes
10	MODERATE	68	↑ UP	web_search	Researchers (Shimazaki et al., 2025, Japan Geophysical Union) revised their estimate of the next Nankai Trough earthquake from ~2040 to ~2030, suggesting a meaningful upward revision to near-term probability.	No
11	STRONG	88	NEUTRAL	web_search	The Nankai Trough has historically produced megaquakes every 90-150 years; the last occurred in 1946 (~79 years ago as of 2025), placing it within the historically plausible recurrence window but not overdue by maximum historical standards.	Yes
12	MODERATE	70	↓ DOWN	code_execution	Given an 80% probability of Nankai Trough megaquake within 30 years (from ~2025), a Poisson approximation yields roughly 15.6% probability of it occurring within the 3.67-year window to end-2029.	Yes
13	MODERATE	80	↓ DOWN	web_search	Expert caveats note that earthquake prediction science cannot pinpoint timing precisely, and the 80% in 30 years estimate does not imply imminent occurrence; probability for 2026-2029 specifically remains very uncertain.	Yes
14	STRONG	90	NEUTRAL	article_search	No M8.0+ earthquake with epicenter in Japan has been recorded in the research data covering May 2025 through March 2026, confirming the question has not yet resolved YES as of the latest available data.	Yes

Information Gaps

No precise count of M8.0+ earthquakes in Japan specifically (vs. broader Pacific) since 1900 to validate the ~8-10 event historical frequency used in the Poisson calculation.
No current GPS/geodetic strain accumulation data for the Nankai Trough or other Japanese fault systems to assess whether stress is at or near rupture thresholds.
No data on current slow-slip events or tremor activity along the Nankai Trough, which could be precursory indicators.
The Kalshi market price of 66% is dramatically higher than the Poisson base rate (~21-25%); no clear explanation for this 40+ percentage point gap was found — possible market mispricing, emotional/recency bias after recent seismic events, or incorporation of unknown expert data.
No information on whether the December 2025 M7.5 earthquake or July 2025 Tokara swarm has altered official probability estimates for near-term M8.0+ events.
The Shimazaki et al. 2025 revised ~2030 estimate for Nankai Trough is only available as a conference abstract — full peer-reviewed paper not confirmed.
No data on USGS or other international seismic agency current assessments of Japan M8.0+ near-term risk.

Key Uncertainties

Whether the ~2030 revised Nankai Trough timing estimate from Shimazaki et al. (2025) is scientifically robust or a preliminary/outlier finding.
Whether the large gap between Kalshi market price (66%) and base-rate Poisson estimate (21-25%) reflects genuine superior market information or market inefficiency/emotional bias from recent high-profile seismic events.
The degree to which the 2011 Tohoku M9.0 earthquake and its aftershock sequence relieved or redistributed stress on adjacent fault zones (including Nankai Trough), potentially accelerating or delaying the next M8.0+ event.
Whether a non-Nankai M8.0+ earthquake (e.g., on the Kuril-Kamchatka subduction zone's Japan side, or the Ryukyu Trench) could occur before 2030 — the question is not limited to the Nankai Trough.
The exact epicenter definition for resolution — whether tsunamigenic earthquakes with epicenters just offshore Japan's EEZ boundary would qualify.
Impact of the July 2025 Russian Kamchatka M8.8 on stress transfer to the Japanese plate boundary system.

Market Context

Price: 0.66 · 7d: +0.10 · 30d: +0.18
Volume: LOW-MODERATE — 112 average daily contracts is thin for a binary market of this nature · Depth: HIGH depth (29,739) but extremely wide spread ($0.97) suggests the posted price of 66% may not represent true equilibrium; the yes_bid of $0.01 and no_bid of $0.02 are inconsistent with a 66% probability and indicate the orderbook data may be stale or misinterpreted
Reliability: LOW-MEDIUM — The 66% price represents a +18% surge over 30 days likely driven by the December 2025 M7.5 event and megaquake advisory; wide spread and thin trading suggest significant uncertainty about whether this price is well-calibrated; the dramatic divergence from base-rate calculations (~21-25%) warrants skepticism

4. Independent Forecaster Opus · No market price · legacy v2 Did NOT see market price

Independent estimate: 33% (47% confidence)

base rate

33%

45% conf

evidence driven

33%

50% conf

contrarian

32%

45% conf

Spread

1pp

Low disagreement

sq1: Based on historical frequency, will Japan experience at least one M8.0+ earthquake in any given 3.7-year window?

28%

Historical M8.0+ events in/around Japan since 1900 include: 1933 Sanriku (8.4), 1944 Tonankai (8.1), 1946 Nankai (8.1/8.4), 1952 Hokkaido (8.1), 1958 Etorofu (8.3), 1963 Kuril (8.5), 1968 Tokachi-oki (8.2), 1994 Hokkaido (8.3), 2003 Tokachi-oki (8.3), 2011 Tohoku (9.1). That's ~10 events in 126 years, λ≈0.079/yr. Some events are on borderline territorial waters. Over 3.67 years: P = 1-exp(-0.079*3.67) ≈ 25%. Adjusting slightly upward for possible undercount and inclusive territorial waters definition: ~28%.

sq2: Are current seismic indicators for Japan elevated, suggesting higher-than-baseline probability of M8.0+ in 2026-2029?

35%

Multiple indicators point to elevated risk: JMA's first-ever Nankai caution warning (Aug 2024), updated 80% in 30 years estimate (Jan 2025), Dec 2025 M7.5 with megaquake advisory, Tokara swarm (1,800+ quakes), Kamchatka M8.8 stress transfer. Base rate ~25%; elevated indicators push up by ~10pp. However, 'elevated indicators' historically have poor predictive value for short-term M8+ — Japan has been on heightened alert for decades without realization. Estimate ~35%.

sq3: Will the Nankai Trough megaquake occur before January 1, 2030?

13%

Official 80% in 30 years implies ~5.4%/year hazard rate, giving ~18% over 3.67 years if uniform. However, hazard is not uniform — it grows with time, so near-term (next 3.67 years) probability is somewhat below the linear share, perhaps 12-15%. Shimazaki et al. 2025 revised estimate to ~2030 pushes up modestly but is preliminary. Last event 1946 (80 years ago); recurrence interval 90-150 years means we're in the early plausible window but not overdue. Estimate ~13%.

Show full reasoning trace

base rate: Japan has experienced ~10 M8.0+ earthquakes since 1900 (Sanriku 1933, Tonankai 1944, Nankai 1946, Hokkaido 1952, Etorofu 1958, Kuril 1963, Tokachi 1968, Hokkaido 1994, Tokachi 2003, Tohoku 2011), giving λ≈0.079/yr. Poisson probability over 3.67 years: ~25%.

evidence updates: Elevated indicators (Nankai caution warning, 80% in 30yr estimate, Dec 2025 M7.5 advisory, Tokara swarm, Kamchatka M8.8) push up by ~5-8pp. Shimazaki 2025 revised Nankai estimate adds modest upward pressure but is preliminary. 2011 Tohoku stress transfer effects are ambiguous. No M8+ has occurred May 2025-Mar 2026 — slight downward update as ~1 year of the window has elapsed without resolution.

combination method: Weighted average produces 0.27, but accounting for sub-question dependencies and the elapsed time without resolution, final estimate ~33%. The base rate dominates; specific elevated-risk factors warrant moderate upward adjustment but earthquake forecasting at sub-decadal timescales remains highly uncertain.

final: Final estimate: 33%. Anchored on Poisson base rate of 25% with modest upward adjustment for current elevated risk indicators, partially offset by elapsed time without resolution. Confidence is medium given the inherent unpredictability of earthquake timing and the gap between official probability assessments and prediction market pricing observed historically.

ensemble: {'ensemble_count': 3, 'perspectives': {'base_rate': {'p_yes': 0.33, 'confidence': 0.45}, 'evidence_driven': {'p_yes': 0.33, 'confidence': 0.5}, 'contrarian': {'p_yes': 0.32, 'confidence': 0.45}}, 'spread': 0.01, 'individual_p_yes': {'base_rate': 0.33, 'evidence_driven': 0.33, 'contrarian': 0.32}}

5. Devil's Advocate Sonnet · legacy v2 Market price revealed

Asymmetric Evidence Reasoning Flaw

Challenges

The base rate calculation may be undercounting. The forecaster lists ~10 events since 1900 but several borderline events (Kuril, Etorofu) are arguably within Japan's territorial scope. Including all events around Japan's plate boundaries gives a rate closer to 0.08-0.10/yr, which over 3.67 years gives 25-31%, not 25%. Also, since 2000 alone there have been 2 M8+ events (2003 Tokachi, 2011 Tohoku) suggesting the modern rate is higher.
The forecaster applies only a modest +5-10pp upward adjustment for elevated indicators, but the evidence stack is unusually strong: JMA's first-ever Nankai caution warning, a December 2025 megaquake advisory after a M7.5, official 80% in 30 years, Shimazaki 2025 ~2030 estimate, plus Kamchatka M8.8 stress transfer. This is a confluence of independent strong signals that may warrant a larger adjustment than +8pp.
Sub-question 3 (Nankai p=13%) may be too low. If hazard is time-increasing and the Shimazaki 2025 estimate centers on ~2030, then the 2026-2029 window captures a significant portion of the probability mass. Additionally, Nankai is only ONE of several M8+ pathways (Japan Trench aftershocks, Kuril, Hokkaido, outer-rise events) — the forecaster doesn't clearly aggregate these independent pathways.
The 'elapsed time without resolution' downward update is questionable — only ~1 year of a 4.67-year original window has elapsed, and absence of an event in 1 year provides minimal Bayesian update against a Poisson process with rate ~0.08/yr (expected events in 1 yr = 0.08, so non-occurrence is the modal outcome and barely informative).
The market trend (+18% over 30 days) following the December 2025 M7.5 and megaquake advisory suggests informed traders are incorporating these signals more aggressively. While the market spread is wide, the 30-day directional move is meaningful signal that the forecaster's +8pp adjustment for these same events may be insufficient.

Suggested adjustment: +6pp

Adjust upward modestly (~+6pp to ~0.39) because: (1) base rate is likely underestimated when counting all M8+ events affecting Japan's plate boundaries, (2) the multiple independent strong elevated-risk indicators warrant a larger upward update than +8pp given their confluence, (3) the multiple independent megaquake pathways (Nankai, Japan Trench, Kuril, Hokkaido) aren't clearly aggregated. This is not adjustment toward the market — the market at 66% still appears too high given Poisson fundamentals — but the forecaster's 33% appears slightly low on its own merits.

Market comparison: Forecaster: 33%, Market: 66%. The 33pp divergence is large but the market shows wide spread, low volume, and recent emotional surge after the Dec 2025 event. Market is likely overpriced due to recency bias, but forecaster's 33% may also be slightly low. A reasonable range is 35-45%; full convergence to market is not warranted.

6. Aggregator Opus · final calibrated forecast

42% (40% confidence)