DTS-Model

DTS Model (short for Deprem Tipi Sınıflandırma Modeli or Earthquake Type Classification System) is an AI/ML based system designed to classify earthquakes. Earthquake classification is done after the earthquake occurs.

The main goal is to predict the “Foreshock” with Machine Learning (ML) Models Classification. After determining the characteristics of being a “Potential Foreshock”, a preparation can be made for the “Mainshock”.

• Foreshock (preliminary event before a major quake)
• Mainshock (the primary and largest event in a sequence)
• Aftershock (subsequent smaller quakes following the main event)

WITH MACHINE LEARNING (ML) MODELS CLASSIFICATION OF EARTHQUAKE TYPES (EARTHQUAKE TYPES) VE MODEL ESTIMATES OF THE PROBABILITY OF AN OCCURRING EARTHQUAKE BECOMING A POTENTIAL PREDICTIVE EARTHQUAKE ~ AROUND TURKEY

Core Components:

1. Data Input: Earthquake metadata (e.g., magnitude, location, depth)
2. AI Classifier: Uses machine learning or deep learning algorithms (like Random Forest, LSTM, CNN) to determine the type of earthquake.
3. Alert and Reporting: Provides a classification and probability score, useful for early warning systems and emergency planning.

AI-Based Earthquake Type Classification System (DTS-Model) – Global Comparison

🌍 Similar Systems Worldwide vs. DTS-Model

1. ShakeAlert – USA (USGS)

• 🔧 Focus: Rapid estimation of magnitude and location to issue alerts
• 🚫 Classification: No distinction between foreshocks, mainshocks, or aftershocks
• ✅ Usage: Early alerts, mobile notifications, infrastructure response

2. JMA EEW – Japan

• 🔧 Focus: Early warnings based on detection of P-waves within seconds
• 🚫 Classification: Does not classify quake types
• ✅ Advantage: Very fast and reliable alerting system

3. MyShake – UC Berkeley (California)

• 🔧 Focus: Uses smartphone sensors to detect ground movement and generate alerts
• 🚫 Classification: Focused only on tremor detection; no type classification
• ✅ Strength: Broad user network, low-cost sensor system

4. Earthquake Network App

• 🔧 Focus: Leverages accelerometers in smartphones to detect seismic activity
• 🚫 Classification: No classification of earthquake type
• ✅ Benefit: Decentralized system with global reach

---

✅ How DTS-Model Differs and Innovates

Feature	Existing Systems	DTS-Model
Early warning	✅	✅
Earthquake type prediction	❌	✅
AI-based classification	❌ (partial)	✅
Seismic signal analysis	✅	✅
Sequential event analysis	❌	✅

---

📌 Conclusion

The DTS-Model is one of the first and most innovative systems that focuses not only on early detection, but also on classifying an earthquake in as a foreshock, mainshock, or aftershock. This capability has the potential to revolutionize disaster response, automatic alert systems, and emergency preparedness strategies.

📊 Comparison of Earthquake Type Classification Systems

System Name	Country / Organization	Type Classification	Types (Foreshock / Mainshock / Aftershock)	AI / ML Usage	Notes
DTS-Model (Prototype)	Turkey (Proposed)	✅ Yes	✅ All	✅ Yes (LSTM, CNN, RF)	One of the first examples
ShakeAlert	USA (USGS)	❌ No	❌ None	⚠️ Partial (for prediction)	Only magnitude & location estimation
JMA EEW	Japan	❌ No	❌ None	❌ No	Only early warning (within seconds)
MyShake App	UC Berkeley	❌ No	❌ None	⚠️ Yes (for data processing)	Mobile-based earthquake alert
GFZ Seismic Tools	Germany (GFZ)	❌ No	❌ None	❌ No	Research-oriented cataloging
IRIS QDDS	International (IRIS)	❌ No	❌ None	❌ No	Signal visualization, no classification

---

📌 Key Highlight:

The DTS-Model aims to provide a functionality that other systems currently lack:

🔹 Classification of earthquake types (foreshock / mainshock / aftershock)
🔹 Analyzing seismic patterns using AI/ML techniques

1900 – 2025 Mainshock in Türkiye (Rule Based)

1900 – 2025 Mainshock – Aftershock in Türkiye (Rule Based)

1900 – 2025 Mainshock – Aftershock – Foreshock in Türkiye (Rule Based)

1900 – 2025 Mainshock – Aftershock – Foreshock in Türkiye (ML Based)