← Back to Sessions

WITHDRAWN Seismic Hazard Assessment of Northern Pakistan, Nw Himalayas Using ‎Machine Learning

Description: 

WITHDRAWN Over the last decade, Artificial Intelligence (AI) has experienced substantial advancement, notably in the domain of Machine Learning (ML), which has emerged as an important and extensively adopted pathway within AI. In the present study an earthquake analysis was conducted along the region of Northern Pakistan, Northwestern Himalaya. Seismicity in this region occurred mainly along major faults such as the Main boundary thrust (MBT), and Panjal thrust (PT). A complete catalog of the North Western Himalayas Pakistan (Lat 33°-35° and Long 71°-74°) for the period of 1900-2023 has been compiled with earthquake magnitude computed in Moment Magnitude (Mw). In this study, apply the Classification and Regression Trees (CART), which is a supervised machine learning algorithm mostly used to perform predictive analysis. The CART algorithm is a versatile and powerful non-parametric machine learning algorithm. It is a decision tree algorithm where each fork is split into a predictor variable and each node has a prediction for the target variable at the end scale. On the basis of the CART algorithm, time series prediction is made from 2023-2191, predicted high magnitude (~ 7.7 Mw) earthquake in ~ 2104. Additionally, exploratory data analysis, hypothesis testing and model evaluation were performed for the forecasting of the earthquake. The higher frequency of earthquakes in this region was observed during the period of 1998 - 2010 (3.5 – 5 Mw). The analysis indicates that earthquake producing faults are high at shallow depth. One of the recent major destructive earthquakes that have occurred in this area is “2005 Muzafrabad earthquake”. The seismotectonic characteristics of this zone clearly indicate that it is capable of generating greater Earthquakes (Mw ≥7.5) at shallower depths. The seismic activity trends imply the likelihood of an earthquake akin to the Kashmir Earthquake occurring around the year 2105 (±5 years). The model's correlation with real-time geological data reinforces this forecast, presenting a coherent prediction of seismic activity over the coming century.

Session: Regional-Scale Hazard, Risk and Loss Assessments [Poster Session]

Type: Poster

Date: 5/3/2024

Presentation Time: 08:00 AM (local time)

Presenting Author: Muteeb

Student Presenter: Yes

Invited Presentation: 

Authors