© Achraf Chakir Baraka, Kaoutar Baraka, Mehdi Rahmaoui, Nada Yamoul, Yassine Bahi, Hamid Khalifi. Article available under the CC BY-SA 4.0 licence
ARTICLE
ABSTRACT
The rise in seismic waves in Morocco within the last five years prompted an accurate multivariate analysis based on such statistical methods as the classification by the K-Means algorithm and principal component analysis (PCA) of seismic wave quantitative variables for Morocco. The adopted results of statistics and analyses can be processed to computer systems for the purpose of optimization and simplification in managing risks of seismic activity in Morocco. A method of statistical treatment that would evaluate diverse seismic threats associated with technological challenges . It also studies the limits of integration and machine learning algorithms inside infrastructural monitoring.
The principal output of the component analysis indicated that the PC1 and PC2 components explained 34.82% and 27.85% of the total variation, respectively. The first component was mainly associated with the “magnitude” and “significance” variables. The second component had a strong relationship with “latitude” and “time,” which could describe seismic occurrences in temporal and geographical dimensions. Four clusters were identified and classified by the K-Means algorithm as “Low”, “Medium”, “High” and “Very High”, based on the magnitude of earthquakes.
The application of multivariate analyses, namely the principal component analysis and the K-Means algorithm are useful not only for reducing dimensionality and classification but also for facilitating risk modeling and disaster prevention. However, both approaches have limitations: the PCA assumes linear relationships between variables, while the K-Means algorithm is influenced by the initial positioning of the switchboard. The study shows the importance of integrating multivariate analyses to develop advanced statistical solutions in order to optimize disaster risk management and real-time seismic monitoring. All graphical results are from Python.
KEYWORDS
seismic event statistics, multivariate analysis, principal component analysis, k-means algorithm, risk management, inferential statistics, prediction of natural disasters
REFERENCES
Aschheim, M. A., Black, E. F. and Cuesta, I., (2002). Theory of principal components analysis and applications to multistory frame buildings responding to seismic excitation. Engineering Structures, 24, pp. 1091–1103.
Bloemheuvel, S., van den Hoogen, J., Jozinović, D., Michelini, A. and Atzmueller, M., (2023). Graph neural networks for multivariate time series regression with application to seismic data. International Journal of Data Science and Analytics, 16, pp. 317–332.
Chakir, B. A., Mentagui, D., Bourakadi, A. and Nada, Y., (2021). Principal component analysis and application to public expenditure efficiency indicators. Pakistan Journal of Statistics, 37.
Di Giuseppe, M. G., Troiano, A., Troise, C. and De Natale, G., (2014). K-means clustering as a tool for multivariate geophysical data analysis: Application to shallow fault zone imaging. Journal of Applied Geophysics, 101, pp. 108–115.
Dumay, J., Fournier, F., (1988). Multivariate statistical analyses applied to seismic facies recognition. Geophysics, 53, pp. 1151–1159.
Jain, A. K., (2010). Data clustering: 50 years beyond K-means. Pattern Recognition Letters, 31, pp. 651–666.
Jufriansah, A., Pramudya, Y., Khusnani, A. and Saputra, S., (2021). Analysis of earthquake activity in Indonesia by clustering method. Journal of Physics: Theories and Applications, 5, p. 92.
Kertanah, K., Rahadi, I., Novianti, B. A., Syahidi, K., Sapiruddin, S., Putra, H. M. and Sabar, S., (2022). Applying K-means algorithm for clustering analysis of earthquakes data in West Nusa Tenggara province. Indonesian Physical Review, 5, pp. 197–207.
Lubo-Robles, D., Bedle, H., Marfurt, K. J. and Pranter, M. J., (2023). Evaluation of principal component analysis for seismic attribute selection and self-organizing maps for seismic facies discrimination in the presence of gas hydrates. Marine and Petroleum Geology, 150, p. 106097.
Orozco-Del-Castillo, M. G., Ortiz-Aleman, C., Martin, R., Avila-Carrera, R. and Rodriguez-Castellanos, A., (2011). Seismic data interpretation using the Hough transform and principal component analysis. Journal of Geophysics and Engineering, 8, pp. 61–73.
Paolucci, E., Lunedei, E. and Albarello, D., (2017). Application of principal component analysis to HVSR data aimed at seismic characterization of earthquake-prone areas. Geophysical Journal International, 211, pp. 650–662.
Russell, B. H., (2004). The application of multivariate statistics and neural networks to the prediction of reservoir parameters using seismic attributes. PhD Thesis, Department of Geology and Geophysics, Calgary, Alberta.
Scheevel, J. R., Payrazyan, K., (2001). Principal component analysis applied to 3D seismic data for reservoir property estimation. SPE Reservoir Evaluation & Engineering, 4, pp. 64–72.
Weatherill, G., Burton, P. W., (2009). Delineation of shallow seismic source zones using K-means cluster analysis: Application to the Aegean region. Geophysical Journal International, 176, pp. 565–588.
Wilkin, G. A., Huang, X., (2007). K-means clustering algorithms: Implementation and comparison. IMSCCS 2007, IEEE, pp. 133–136.
Žalik, K. R., (2008). An efficient k'-means clustering algorithm. Pattern Recognition Letters, 29, pp. 1385–1391.