APPLICATION OF ANN AND MATLAB IN PHYSICOCHEMICAL EVALUATION OF GROUNDWATER IN SELECTED COMMUNITIES IN ABA, ABIA STATE NIGERIA

This study focused on groundwater quality in Aba Metropolis. Analysis of physicochemical characteristics of groundwater such as pH, conductivity, total dissolved solids (TDS), alkalinity, anions and cations and some heavy metals were determined using titrimetric, spectrophotometric and atomic absorption spectroscopic methods to obtain concentrations of the physicochemical parameters results in mg/L. The results compared favourably with WHO standard limit except in most of the parameters under consideration. The results of WHO standard error of limit of these parameters were below WHO permissible limit, however the percent deviation for Lead is 2 for groundwater from Umuchichi and Umueme while the percent deviation for Uratta was far much low. The waste disposal of the companies around Umueme region includes antiknock agents such as Tetraethyllead (TEL) used in their heavy-duty machines. When these waste are not properly treated before disposal it seeps down into aquifers and affects groundwater quality leading to high pollution levels. The value for Iron concentration in groundwater in Umuchichi is much within WHO permissible limit while for Umueme and Uratta the values indicated 1.33% increase. The results varied from 6.3 ± 1.44, 7.97 ± 0.47 to 7.27 ± 0.94 in the case of pH for Umuchichi, Umueme and Uratta respectively, whereas the conductivity level in µS/cm, varied from 0.5 ± 0.23, 0.75 ± 0.12 to 0.51 ± 0.22. The pH values for the three communities are all within the tolerable range according to WHO (2006). It is highly recommended that the lead level of groundwater in Umuchichi and Umueme is closely monitored. Since surface lead seeps down to aquifers determining the level of lead on the surrounding surface can help determine safe level of lead in the groundwater. Furthermore, Artificial Neural Network (ANN) and MATLAB were employed to support the data analysis and predict pollution patterns. The integration of ANN technique demonstrated high predictive accuracy, highlighting their effectiveness in water quality assessment and environmental monitoring of groundwater. In conclusion data from this study can serve as a guide to current groundwater quality and areas of improvement when making feasibility studies when it comes to well-drilling on these sites.