Dr. Mo Geophysical Hub

21/04/2026

Honored to share a moment from my presentation at the 4th Miskolc Campus Forum – Green and Smart 🌍
at the University of Miskolc.

I presented on:

AI in Earth & Environmental Sciences: Machine Learning for Subsurface Characterization with Physics, Data, and Uncertainty

The main idea was simple but important:
we cannot see the subsurface directly, so better decisions depend on combining data, physics, AI, and uncertainty.

It was a great opportunity to discuss how these ideas can support more intelligent and responsible exploration in Earth and environmental sciences.

Many thanks to the organizers and everyone who attended and supported the session.

18/04/2026

في الفيديو ده هنفهم واحدة من أهم الأفكار في الـ Inversion 👇

لو كنت بتسأل نفسك:

يعني إيه Mixed Inversion؟
ليه بنضيف model constraint مع الـ data؟
إيه الفرق بين Underdetermined و Overdetermined problems؟
وليه بنستخدم Lagrange multipliers أحيانًا، وأحيانًا نستخدم E + λ²M؟

📌 الفيديو ده هيخليك تفهم فلسفة الـ Inversion مش مجرد معادلات

🧠 هنبدأ من الأساس:
إيه معنى إن المشكلة تكون ill-posed؟
وليه الحل مش بيكون unique؟

📐 وبعدها ندخل في أهم المفاهيم:

Data Misfit vs Model Norm
فكرة Trade-off parameter (λ)
الفرق بين Constrained vs Unconstrained optimization
ليه Mixed Inversion بيشتغل في كل الحالات

⚡ هتركز في الفيديو على:

الفهم الفيزيائي قبل الرياضيات
إزاي تختار model منطقي مش بس fit حلو
تبسيط فكرة الـ Regularization بشكل واضح

🎯 الفيديو مناسب لـ:
طلاب الجيوفيزياء – الهندسة – الفيزياء – Data Science

🚀 لو عايز تفهم Inversion بجد وتفكر زي باحث…
الفيديو ده مهم جدًا ليك

📌 متنساش:

تعمل Subscribe ❤️
تسيب سؤالك في الكومنتات
وتتابع باقي السلسلة

شرح Mixed Inversion ببساطة | Lagrange و E + λ²M من الصفر للاحتراف

15/04/2026

في الفيديو ده هنفهم مع بعض أهم طرق حل مشاكل الـ Inversion بشكل عملي وسهل 👇لو كنت محتار في:* يعني إيه Linearization؟* إزاي نحول مشكلة Nonlinear لمشكلة Linear...

25/01/2026

Stress Sensitivity and Water Content: The study confirms that increasing stress reduces permeability by closing "pore throats" in the rock. However, it uniquely identifies that water content acts as a resistance factor against applied stress; consequently, permeability in wet samples is reduced to a lesser extent than in dry samples

21/01/2026

Happy to share our latest open-access paper in Acta Geophysica:

“Cuckoo Search Algorithm-assisted inversion for estimating petrophysical characteristics using well-logging data”
🔗 https://link.springer.com/article/10.1007/s11600-025-01765-5

What’s the scientific value?
We propose a hybrid inversion workflow (CSA–DLSQ) that combines the global search power of the Cuckoo Search Algorithm with the fast convergence of damped least squares—aiming for better accuracy with much lower computation time in well-log petrophysical inversion.

Key highlights:

Invert depth-local petrophysical parameters while also estimating zone-dependent global variables to improve geological consistency in layered formations.

A tuning/diagnostic study of CSA hyperparameters + a layered strategy to reduce boundary oscillations.

Field test on a Hungarian borehole: 56.6% reduction in total misfit and about 87× speedup per depth point compared to stand-alone CSA.

If you work on well logging, inversion, or optimization, I’d love your thoughts.

This paper presents a CSA-DLSQ hybrid inversion that uses Cuckoo Search Algorithm (CSA) which provides better estimation for petrophysical properties. This hybrid technique significantly reduces computational costs while preserving the global convergence properties of CSA and the fast convergence of...

21/01/2026

I’m happy to share my PhD thesis:

“Geostatistics assisted well-logging inversion method developments”
📄 Read it here: https://real-phd.mtak.hu/2203/

Scientific value (in brief):
My work focuses on building a more robust and automated inversion workflow for wireline well-logging interpretation, combining geostatistics, machine learning–style clustering, and advanced inversion strategies.

Key contributions include:

Automated zonation & layer-boundary detection using MFV-based robust clustering, plus additional boundary-detection ideas (Hurst exponent and factor analysis).

Series expansion–based inversion to improve both forward modeling and inversion stability.

Interval inversion (beyond depth-by-depth inversion) to better estimate parameters over zones.

A more stable inversion workflow using golden-section + SVD-based techniques, and treating zone parameters as unknowns to reduce modeling error and uncertainty.

These developments support more reliable reservoir characterization (e.g., porosity, clay content, water saturation, matrix volumes) with validation on synthetic and real datasets.

Abdelrahman, Moataz Mohamed Gomaa (2025) Geostatistics assisted well-logging inversion method developments. PhD thesis, Miskolci Egyetem.

18/01/2026

18/01/2026

This video examine the theoretical and practical frameworks of geophysical inversion, a process used to estimate the physical properties of underground structures from measured data. The text distinguishes between linearised methods, which use iterative optimization to refine initial models, and global optimization techniques like Simulated Annealing and Genetic Algorithms that seek to avoid local minima. A significant focus is placed on joint inversion, which integrates multiple datasets to reduce ambiguity and improve the reliability of geological interpretations. Furthermore, the material introduces series expansion-based discretization, a strategy designed to transform underdetermined problems into stable, overdetermined ones. By utilizing various mathematical norms and weighting matrices, these methodologies aim to produce accurate petrophysical estimates while suppressing the influence of data noise. Overall, the literature provides a comprehensive guide to modern computational tools used to resolve complex geophysical inverse problems.

10/12/2025

🎄 An Early Christmas Gift! 🎄
I’m delighted to share that our manuscript, “Cuckoo Search Algorithm assisted inversion for estimating petrophysical characteristics using well logging data,” has been accepted for publication in Acta Geophysica.

This acceptance truly feels like a special Christmas gift to close the year with gratitude and pride. In this work, we introduced several innovative contributions, including:
🔹 A Cuckoo Search–driven inversion framework that enhances the estimation of key petrophysical parameters from well-logging data.
🔹 Improved convergence stability compared to traditional optimization techniques, especially in highly non-linear parameter spaces.
🔹 A flexible inversion architecture capable of integrating multiple logging curves to reduce ambiguity and improve subsurface characterization.
🔹 Robust performance on both synthetic and field datasets, demonstrating the practical value of the algorithm in real geophysical applications.

I’m grateful to my supervisor and co-author (Hadeer Hassan) for the teamwork and scientific discussions that shaped this work. My sincere thanks also go to the editors and reviewers for their constructive feedback throughout the process.
Looking forward to continuing research at the intersection of optimization, inversion, and petrophysics in the coming year.