Predicting soil parameters in Maroua 1st, Cameroon, using correlations of geophysical and geotechnical data

This study aims to predict specific soil parameters based on quantitative and qualitative correlations between electrical and geotechnical data. A total of 21 geotechnical boreholes followed by sampling, 21 light dynamic penetrometer tests, and 76 vertical electrical sounding surveys were carried out in Maroua 1st. The electrical resistivity data of the soil layers were obtained through 1D and 2D inversions of the ERT surveys. The geotechnical data were obtained from field tests and laboratory experiments. The statistical analysis of the entire dataset indicates that most of the samples are clay formations, as demonstrated by the means of variables and low variance coefficients, enabling qualitative identification. The distribution analysis of the parameters confirms the complexity of the physical characteristics of soils and the measurement procedures for each parameter. The Spearman's rank correlation matrix indicates that these observations can be used to propose predictive models for free swelling, allowable stress, natural water content, and electrical resistivity. Correlations between free swelling (εg) and other variables are not significant, as the degree of significance is lower than the fixed limit, except for models εg - W (R2 = 0.94), εg - ρ (Ώ.m) (R2 = 0.803), and εg - σ adm (R2 = 0.757). Grouping variables in multiple linear regression enables the development of a mathematical model. The predicted values of εg from this model demonstrate good agreement with the measured values for a validity domain of 0 to 16%. The soil's natural water content has a strong correlation (R2 = 0.80 and rs = – 0.87) with electrical resistivity, meaning that an increase in water content results in a decrease in electrical resistivity. The permissible stress σadm has a strong correlation (R2 = 0.89 and rs = 0.85) with electrical resistivity. This indicates that changes in soil electrical resistivity are associated with soil stiffness. This study aims to predict specific soil parameters based on quantitative and qualitative correlations between electrical and geotechnical data. A total of 21 geotechnical boreholes followed by sampling, 21 light dynamic penetrometer tests, and 76 vertical electrical sounding surveys were carried out in Maroua 1st. The electrical resistivity data of the soil layers were obtained through 1D and 2D inversions of the ERT surveys. The geotechnical data were obtained from field tests and laboratory experiments. The statistical analysis of the entire dataset indicates that most of the samples are clay formations, as demonstrated by the means of variables and low variance coefficients, enabling qualitative identification. The distribution analysis of the parameters confirms the complexity of the physical characteristics of soils and the measurement procedures for each parameter. The Spearman's rank correlation matrix indicates that these observations can be used to propose predictive models for free swelling, allowable stress, natural water content, and electrical resistivity. Correlations between free swelling (εg) and other variables are not significant, as the degree of significance is lower than the fixed limit, except for models εg - W (R2 = 0.94), εg - ρ (Ώ.m) (R2 = 0.803), and εg - σ adm (R2 = 0.757). Grouping variables in multiple linear regression enables the development of a mathematical model. The predicted values of εg from this model demonstrate good agreement with the measured values for a validity domain of 0 to 16%. The soil's natural water content has a strong correlation (R2 = 0.80 and rs = – 0.87) with electrical resistivity, meaning that an increase in water content results in a decrease in electrical resistivity. The permissible stress σadm has a strong correlation (R2 = 0.89 and rs = 0.85) with electrical resistivity. This indicates that changes in soil electrical resistivity are associated with soil stiffness.