🔬 physics

Assessment of natural radionuclide distribution and radiological hazards in Egyptian soil using gamma-ray spectroscopy and evaluating with Monte Carlo simulation and Artificial Neural Network

This study evaluates the distribution and radiological hazards of natural radionuclides in Egyptian soil from Zagazig and 10th of Ramadan using gamma-ray spectroscopy, finding that Zagazig samples exhibit approximately double the activity levels of the other city while both remain within international safety limits, with Monte Carlo simulations and Artificial Neural Networks demonstrating excellent agreement with experimental data.

Original authors: A. F. Saad, M. Al-Abyad, Amany T. Gendya, E. M. Sedqy, Mahmmoud A. Ibraheim, Mohammed Elywa

Published 2026-08-06

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: A. F. Saad, M. Al-Abyad, Amany T. Gendya, E. M. Sedqy, Mahmmoud A. Ibraheim, Mohammed Elywa

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Invisible Rain and the Soil's Secret Recipe

Imagine the ground beneath your feet isn't just dirt, but a giant, slow-burning campfire. It's not hot enough to burn your shoes, but it is constantly emitting a tiny, invisible "rain" of energy called radiation. This isn't the scary kind from nuclear bombs; it's natural background radiation, a quiet hum of energy that has been there since the Earth was formed. It comes from tiny, unstable atoms inside rocks and soil, like Uranium, Thorium, and Potassium, which are slowly breaking apart over billions of years.

Why should we care about this invisible rain? Because we spend most of our lives indoors, and the soil is the foundation of our homes. If the soil has too much of this "radioactive rain," it can make the air inside our houses a bit more charged than it needs to be. Scientists act like weather forecasters for this invisible energy. They want to know: Is the soil in our neighborhood safe to build a house on? Is it safe to grow vegetables in? To answer this, they use special tools to count the atoms, and increasingly, they use super-smart computer programs that can predict the future risks based on patterns, almost like a crystal ball made of math and data.

The Great Soil Detective Story: Egypt's Eastern Delta

In this study, a team of scientists from Egypt decided to play detective in two cities located in the eastern part of the Nile Delta: 10th of Ramadan and Zagazig. Think of these two cities as neighbors with very different personalities. 10th of Ramadan is a newer, industrial city built on sandy soil, surrounded by desert. Zagazig, on the other hand, is an older, historic city with soil rich in clay, sitting right on a branch of the Nile River.

The researchers wanted to see if these different "soil personalities" meant different levels of radioactive rain. They collected 40 samples of dirt—25 from the sandy 10th of Ramadan and 15 from the clay-heavy Zagazig. To measure the invisible energy, they used a high-tech device called an HPGe gamma-ray spectrometer. You can imagine this machine as a super-sensitive microphone that doesn't listen to sound, but to the specific "notes" (gamma rays) that radioactive atoms sing when they decay.

The Double-Check: Computers as Time Travelers

But the scientists didn't stop at just listening to the soil. They wanted to see if they could predict the danger using two powerful computer tricks:

Monte Carlo Simulation: Imagine rolling a pair of dice thousands of times to see every possible outcome of a game. This computer method does the same thing with radiation data, running 10,000 different scenarios to see how much the results might wiggle or change. It helps them understand the "uncertainty" in their measurements.
Artificial Neural Network (ANN): This is like teaching a computer to be a genius student. Instead of giving the computer a strict formula, they fed it all their soil data and let it learn the patterns on its own. It's like showing a child a thousand pictures of cats until they can spot a cat in a new picture without being told exactly what a cat looks like.

What They Found: The Tale of Two Cities

When the scientists compared the soil from the two cities, they found a clear difference. The soil in Zagazig was about two times more radioactive than the soil in 10th of Ramadan.

The Numbers: In Zagazig, some samples had very high levels of Uranium-238 (up to 52.4 ± 18.9 Bq kg⁻¹) and Potassium-40 (up to 498.0 ± 27.3 Bq kg⁻¹). In fact, some Zagazig samples had double the amount of Uranium-238 compared to the world's average limit.
The Comparison: The 10th of Ramadan samples were much lower. Their radioactive levels were generally about half of what was found in Zagazig.

However, here is the most important part: Both cities are safe.

Even though Zagazig had higher numbers, the scientists calculated the "Radium Equivalent" (a way to mix all the different radioactive elements into one safety score) and the "External Hazard Index." They found that the radiation levels in both cities were below the international safety limits set by experts. The "radiation rain" in these areas is not heavy enough to be dangerous to the people living there.

The Computer vs. The Real World

The coolest part of the paper is how well the computers matched the real world.

The Artificial Neural Network learned the patterns so well that its predictions were almost a perfect match for the real measurements. The computer's "guess" was so accurate that the error was tiny (a regression coefficient of R = 1, which is basically a perfect score).
The Monte Carlo simulation confirmed that the uncertainty in their data was low. The "dice rolls" showed that the results were stable and reliable.

The paper explicitly states that the theoretical calculations from the computer models are in "excellent agreement" with the actual measurements taken from the soil. They didn't find any hidden dangers that the computers missed, nor did they find that the computers were wildly wrong.

The Final Verdict

So, what's the bottom line? The soil in both Egyptian cities is safe for building homes and farming. While Zagazig has a slightly "spicier" radioactive recipe than 10th of Ramadan, neither is hot enough to be a problem. The study proves that we can trust our high-tech computers (Monte Carlo and Neural Networks) to help us understand these invisible risks, saving us time and money while keeping our communities safe. The invisible rain is falling, but it's just a gentle drizzle, not a storm.

Technical Summary: Assessment of Natural Radionuclide Distribution and Radiological Hazards in Egyptian Soil

Problem Statement
Natural radioactivity in soil, originating from primordial radionuclides such as $^{238}$ U, $^{232}$ Th, and $^{40}$ K, constitutes a primary source of ionizing radiation exposure for the human population. This exposure occurs via external gamma radiation and internal inhalation of radon ( $^{222}$ Rn) and its progeny. Given that humans spend a significant portion of their time indoors, understanding the distribution of these radionuclides in soil is critical for assessing radiation risks, particularly in regions where soil is used for construction or agriculture. While experimental measurement is standard, there is a need for robust theoretical frameworks to predict radiological hazards and model uncertainties efficiently. This study addresses the assessment of natural radioactivity and associated radiological hazards in two specific cities within the Sharqia Governorate, Egypt: 10th of Ramadan and Zagazig.

Methodology
The research employed a hybrid approach combining experimental measurement, Monte Carlo Simulation (MCS), and Artificial Neural Networks (ANN).

Experimental Data Collection:
- Sampling: Forty soil samples were collected (25 from 10th of Ramadan and 15 from Zagazig).
- Preparation: Samples were dried at 105°C, sieved through a 1-mm mesh, and sealed in 1-liter containers for one month to establish secular equilibrium.
- Measurement: A high-purity germanium (HPGe) gamma-ray spectrometer (70% relative efficiency) coupled with a GENIE-2000 multichannel analyzer was used to measure activity concentrations of $^{226}$ Ra, $^{238}$ U, $^{232}$ Th, and $^{40}$ K.
- Calculations: Experimental data were used to calculate radiological parameters including Radium Equivalent Activity ( $R_{aeq}$ ), Absorbed Gamma Dose Rate ( $D$ ), Annual Effective Dose Rate, and External Hazard Index ( $H_{ex}$ ).
Theoretical Modeling:
- Monte Carlo Simulation (MCS): The Oracle Crystal Ball program was utilized to perform 10,000 iterations for sensitivity analysis and uncertainty propagation. This stochastic approach generated random realizations of input variables to estimate exposure risks and confidence intervals.
- Artificial Neural Network (ANN): A multi-layered ANN model was developed to model the nonlinear relationships between radionuclide concentrations and radiological hazard indices. The network architecture included input, hidden, and output layers. Training was conducted to minimize the Mean Squared Error (MSE), with validation checks performed to prevent overfitting.

Key Contributions

Integrated Framework: The study demonstrates a methodology that integrates experimental gamma-ray spectroscopy with MCS and ANN to evaluate radiological hazards.
Regional Assessment: It provides specific radiological data for the 10th of Ramadan and Zagazig cities, areas characterized by distinct geological profiles (sandy soil with clay lenses vs. high clay content derived from ancient Nile branches).
Predictive Validation: The paper validates the use of ANN as a low-computational-cost tool for predicting radiological parameters, showing that it can effectively replicate experimental results derived from complex physical measurements, though with noted variations in specific isotopic predictions.

Results

Radionuclide Distribution:
- Zagazig: Samples exhibited higher activity concentrations compared to 10th of Ramadan. Specifically, Zagazig samples showed $^{226}$ Ra, $^{238}$ U, $^{232}$ Th, and $^{40}$ K concentrations roughly double those of 10th of Ramadan. Several Zagazig samples exceeded the world average for $^{238}$ U (25.5 Bq kg $^{-1}$ ) and $^{40}$ K (400 Bq kg $^{-1}$ ), with some $^{40}$ K values reaching up to 498.0 Bq kg $^{-1}$ .
- 10th of Ramadan: While generally lower, six samples from this city showed $^{238}$ U concentrations significantly higher than the world average.
- Comparison: The mean values for $R_{aeq}$ , absorbed dose, annual effective dose, and $H_{ex}$ in Zagazig were approximately two times higher than those in 10th of Ramadan.
Radiological Hazards:
- Despite the higher concentrations in Zagazig, the mean absorbed dose rate for both cities remained below the world average of 51 nGy h $^{-1}$ (with the exception of two Zagazig samples).
- All samples from both cities were found to be below the international recommended limits for radiological safety (e.g., $R_{aeq} < 370$ Bq kg $^{-1}$ ).
Model Performance:
- ANN: The trained network achieved a minimum MSE of approximately $1.1 \times 10^{-7}$ at epoch 1000. The regression coefficient ( $R$ ) between predicted and measured values was 1, indicating an almost perfect correlation.
- MCS: The simulation produced near-normal distributions for $R_{aeq}$ and absorbed dose rates, confirming low uncertainty in the estimates.
- Agreement: The theoretical predictions from the combined MCS/ANN model showed excellent agreement with experimental data for most parameters. However, the study noted significant discrepancies in the prediction of $^{238}$ U concentrations: the model underestimated the measured values by more than a factor of two for 10th of Ramadan (average squared error 0.476%) and by approximately a factor of six for Zagazig (average squared error 0.159%). Conversely, the mean squared error for the other three isotopes ( $^{226}$ Ra, $^{232}$ Th, and $^{40}$ K) and for radiological hazard indices (ranging from 0.012% to 0.42%) remained extremely low, indicating high accuracy for those specific outputs.

Significance and Claims
The authors claim that the study successfully establishes a concise comprehension of the correlation between natural radioactivity in soil and associated radiological hazards in the eastern Nile Delta. The primary significance lies in the demonstration that the integrated approach of Monte Carlo simulation and Artificial Neural Networks provides a powerful, reliable, and computationally efficient framework for environmental radiation assessment.

The paper concludes that while Zagazig presents higher background radiation levels than 10th of Ramadan, both cities are generally safe regarding health hazards from radon and external gamma radiation according to international standards. The study asserts that the theoretical calculations based on MCS and ANN agree well with experimental data for most radiological parameters and isotopes, validating these computational tools as effective alternatives or supplements to extensive experimental campaigns for predicting radionuclide distributions and dosimetric parameters, despite observed deviations in specific $^{238}$ U concentration predictions.