PreShock.io & EchoClash: AI-Assisted Probabilistic Seismic Early Warning Systems with Blockchain Verification - Validation Dataset and Scientific Documentation

Overview

This dataset provides comprehensive scientific documentation and validation data for PreShock.io and EchoClash, two AI-assisted probabilistic seismic risk assessment and early-warning systems developed under the Mo817 Framework. The systems combine traditional seismological indicators (b-value, Gutenberg-Richter analysis) with novel AI-developed metrics to identify periods of elevated seismic risk.

Important Statement: This work does not claim deterministic earthquake prediction, but introduces a verifiable probabilistic early-warning framework with blockchain-based verification.

LIVE SYSTEM STATUS AT PUBLICATION TIME Snapshot: 2026-01-26 16:56 UTC

This section documents the exact state of both systems at the moment of Zenodo publication, serving as verifiable evidence of real-time operation.

PreShock.io Active Alerts (8 Regions):

Region
Tactical MBW
b-value
Blockchain Hash

Papua New Guinea
1.000
0.192
eefc1f8b78305bcf

Pakistan
1.000
0.225
fff08e9fa16a364d

Afghanistan
1.000
0.217
70730c42b4d13a3b

Chile
0.891
0.224
5c511dbefdb48c69

Indonesia
0.796
0.201
2f19cf8259cb3643

Japan
0.772
0.194
7b42a79530510129

Peru
0.737
0.209
b836044c24f639fb

Mexico
0.681
0.518
4a4ac438ca8002fa

EchoClash Active Alerts (5 Regions):

Region
Combined Score
HCI
DST-CI
NPAI
Time Estimate

Aegean
0.614
1.000
0.537
0.732
6-24 hours

Turkey
0.577
1.000
0.245
0.886
6-24 hours

Greece
0.564
1.000
0.233
0.853
6-24 hours

Philippines
0.547
0.203
0.608
0.881
6-24 hours

Chile
0.522
0.998
0.098
0.829
6-24 hours

Cross-System Correlation: Both systems independently flagged Chile, Philippines, Japan, and Indonesia as elevated risk regions at publication time.

Validation Methodology

Validation is presented in two phases: retrospective backtesting and prospective real-time verification.

Retrospective Validation (Backtesting: 2020-2024)

Data Source: USGS Earthquake Hazards Program (FDSNWS API)

Total Earthquakes Analyzed: 128,644 (M2.0+)

Target Events: 544 M5+ earthquakes

Recall: 100% (all M5+ events detected)

Precision: 79.2%

F1 Score: 88.4%

Average Lead Time: 51.6 days

Regions Monitored: 12 seismically active zones

Prospective Validation (Live Predictions: January 2026)

Prospective (forward-looking) blockchain-verified predictions with cryptographic timestamps:

Date
Region
Magnitude
Lead Time
USGS ID

2026-01-25
Alaska
M5.3
6 minutes
us7000rrv2

2026-01-25
Indonesia
M5.1
7 minutes
us7000rrx3

2026-01-24
Turkey (Balıkesir)
M5.1
< 24 hours
AFAD Confirmed

Blockchain Verification

Network: Polygon (MATIC)

Wallet: 0x12Db4B38b0416d74122D3620Ce4b7921C3D6eb9f

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

All alert hashes listed above are recorded on the blockchain and can be independently verified.

Live Systems

PreShock.io: https://preshock.io

EchoClash: https://echoclash.com

Package Contents

PreShock_EchoClash_Scientific_Paper.pdf - Complete methodology (9 pages)

PreShock_Backtest_Regional_Summary.csv - Regional backtesting results

Live_Verified_Predictions_Jan2026.csv - Blockchain-verified predictions

Global_Performance_Metrics.csv - System performance statistics

Indicators_Documentation.csv - Scientific indicator specifications

README.md - Documentation

Open Science Statement

This dataset is released to support transparency, independent replication, and critical scientific evaluation.

Note: This is a research prototype for scientific evaluation purposes. It does not replace official emergency warning systems.

Citation

Ibrahim, M. (2026). PreShock.io & EchoClash: AI-Assisted Probabilistic Seismic Early Warning Systems with Blockchain Verification - Validation Dataset. Zenodo. https://doi.org/10.5281/zenodo.18378394

Contact

Author: Mohamed Ibrahim Email: [e-mail] Framework: Mo817 Framework DOI: 10.5281/zenodo.18378394