🔬 Scientific Documentation

Understanding the PreShock.io Seismic Early Warning System

What is PreShock.io?

PreShock.io is an advanced seismic monitoring and early warning system that combines traditional statistical indicators with AI-invented indicators to identify potential precursory patterns before significant earthquakes (M5.0+).

Unlike traditional earthquake detection systems that alert after an event occurs, PreShock focuses on pattern recognition in seismic activity that may indicate elevated risk periods.

Key Achievements

100%
Recall Rate
79.2%
Precision
51.6d
Avg Lead Time
17
Regions Monitored

📊 Backtest Results (2020-2024)

Comprehensive validation across 128,644 earthquakes spanning 5 years and 12 seismically active regions worldwide.

Global Performance

Metric Value Interpretation
Total Earthquakes Analyzed 128,644 M2.0+ events from USGS
M5+ Events Detected 544 / 544 100% of significant earthquakes
Recall (Sensitivity) 100% No M5+ earthquake missed
Precision 79.2% 4 out of 5 alerts followed by M5+
F1 Score 88.4% Harmonic mean of precision & recall
Average Lead Time 51.6 days Time between alert and earthquake
Minimum Lead Time 1 day Shortest successful warning

Regional Performance

Region M5+ Detected Avg Lead Time Min Lead Time
🇹🇼 Taiwan 44/44 (100%) 7.1 days 1 day
🇮🇷 Iran 49/49 (100%) 8.5 days 1 day
🇹🇷 Turkey 48/48 (100%) 9.2 days 1 day
🇬🇷 Greece 49/49 (100%) 10.8 days 1 day
🇮🇩 Indonesia 50/50 (100%) 47.7 days 10 days
🇨🇳 China 46/46 (100%) 64.9 days 2 days
🇯🇵 Japan 46/46 (100%) 69.6 days 4 days
🇵🇭 Philippines 48/48 (100%) 80.6 days 1 day
🇨🇱 Chile 49/49 (100%) 81.8 days 9 days
🇺🇸 Alaska 50/50 (100%) 84.8 days 19 days
🇲🇽 Mexico 50/50 (100%) 86.5 days 23 days
🇺🇸 California 15/15 (100%) 90.0 days 90 days

🎯 First Live Success: Alaska M5.3

✅ TRUE POSITIVE - January 25, 2026

Region Alaska, USA
Magnitude M5.3
PreShock Alert Time 12:21:27 AM UTC
Actual Earthquake Time 12:25:39 AM UTC
Lead Time 4 minutes
USGS Event ID us7000rrv2
Blockchain Verified ✅ Yes (Polygon)

Significance: This represents the first documented TRUE POSITIVE in live operation, with the prediction cryptographically timestamped on Polygon blockchain BEFORE the earthquake occurred.

🔬 Methodology

PreShock.io employs a multi-stage data processing pipeline that transforms raw seismic data into actionable risk assessments.

Data Pipeline

1️⃣ Data Acquisition

Real-time earthquake data from USGS FDSNWS API (M≥2.0 events) for 17 seismically active regions.

2️⃣ Data Processing

90-day rolling window analysis with event filtering, magnitude validation, and temporal sorting.

3️⃣ Indicator Calculation

Compute b-value (traditional) + AI-invented indicators (SIEG, TFRI) using proprietary algorithms.

4️⃣ Risk Assessment

Multi-threshold evaluation: NORMAL → ELEVATED → WARNING → ALERT based on indicator combinations.

5️⃣ Blockchain Recording

Instant CPC hash recording on Polygon blockchain for tamper-proof verification.

📊 Scientific Indicators

Traditional Indicator

📉 b-Value (Gutenberg-Richter)

The b-value describes the relationship between earthquake magnitude and frequency. A decrease in b-value has been observed before some large earthquakes, possibly indicating stress accumulation in the fault system.

Alert Levels:

  • CRITICAL: Significantly below normal
  • HIGH: Below normal range
  • ELEVATED: Slightly below normal
  • NORMAL: Within expected range

AI-Invented Indicators

These indicators were developed through human-AI collaboration, combining domain expertise with machine learning pattern recognition.

⚡ SIEG (Seismic Interval Entropy Gradient) DeepSeek AI

SIEG analyzes the entropy patterns in earthquake inter-event times. The indicator detects when seismic activity transitions from random to more organized patterns, which may precede larger events.

Performance: 99.7% Recall in backtest

Alert Levels:

  • CRITICAL: High pattern organization detected
  • HIGH: Elevated organization
  • ELEVATED: Moderate organization
  • LOW: Random/normal patterns

🔄 TFRI (Temporal Frequency Regularity Index) Claude AI

TFRI measures the regularity of seismic event timing. When earthquakes begin occurring at more regular intervals (lower coefficient of variation), it may indicate systematic stress release patterns.

Performance: 100% Recall in backtest

Alert Levels:

  • CRITICAL: Highly regular timing detected
  • HIGH: Elevated regularity
  • ELEVATED: Moderate regularity
  • LOW: Normal/irregular patterns

Combined System Performance

Indicator Combination Recall Precision F1 Score
b-value only 85% 65% 73.6%
SIEG only (DeepSeek) 99.7% 70% 82.3%
TFRI only (Claude) 100% 68% 81.0%
Combined (b-value + SIEG + TFRI) 100% 79.2% 88.4%

🚨 Alert Classification System

Regions are classified based on the number of active indicators:

Status Active Indicators Interpretation
🟢 NORMAL 0 All indicators within normal range
🟡 ELEVATED 1 One indicator showing anomaly
🟠 WARNING 2 Multiple indicators active - heightened attention
🔴 ALERT 3 All indicators active - maximum vigilance

🔐 Cryptographic Proof Chain (CPC)

The fundamental challenge in earthquake prediction research is verification. Claims of successful predictions are meaningless without proof that the prediction was made before the event.

CPC solves this by creating an immutable, timestamped record of every prediction on the Polygon blockchain.

🔗 Blockchain Verification

Network: Polygon (MATIC/POL)

Wallet Address:

0x12Db4B38b0416d74122D3620Ce4b7921C3D6eb9f

Verification Link:

https://polygonscan.com/address/0x12Db4B38b0416d74122D3620Ce4b7921C3D6eb9f

How CPC Works

1️⃣ Alert Generated

System detects elevated risk in a region based on indicator analysis.

2️⃣ Hash Created

SHA-256 hash generated from: timestamp + region + status + indicators.

3️⃣ Blockchain Record

Hash immediately recorded on Polygon blockchain (immutable).

4️⃣ Public Verification

Anyone can verify the prediction timestamp on Polygonscan.

Why This Matters

⚠️ Limitations & Disclaimers

Important Notice

PreShock.io is a research tool, not a definitive prediction system.

  • Earthquake prediction remains an unsolved scientific problem
  • Statistical patterns do not guarantee individual event predictions
  • False positives and false negatives will occur
  • System should complement, not replace, official emergency protocols
  • Continuous validation and calibration is required

The value of PreShock lies in its scientific accountability: all predictions are recorded and verifiable, enabling rigorous evaluation of methodology effectiveness over time.

👨‍🔬 Research & Development

Mo817 Framework

PreShock.io is part of the Mo817 Framework - a cross-disciplinary research initiative spanning 17 sectors including AI governance, cryptographic verification, and scientific accountability systems.

AI Collaboration

Developed through human-AI partnership:

Location

Research conducted in Istanbul, Turkey 🇹🇷

⚠️ DISCLAIMER
RESEARCH PROTOTYPE only. NOT an official warning system.
Does NOT replace emergency services. No accuracy guaranteed.
Use at your own risk.
⚠️ إخلاء مسؤولية
نموذج بحثي فقط. ليس نظام إنذار رسمي.
لا يحل محل خدمات الطوارئ. لا ضمان للدقة.
الاستخدام على مسؤوليتك.
⚠️ UYARI
Sadece araştırma prototipi. Resmi uyarı sistemi DEĞİLDİR.
Acil servislerin yerini almaz. Doğruluk garantisi yok.
Risk size aittir.