Tidal effects on groundwater levels in Maputo, Mozambique

Abstract

<p>Groundwater is an important source for water supply in Maputo City, Mozambique. A groundwater monitoring network has been established for systematically monitoring of groundwater level and salinity changes. Automatic data loggers were installed to register groundwater levels and EC values every hour. Three observation wells were installed in the coast to detect seawater intrusion. The measured groundwater levels show clearly tidal effects. In this study, time series analysis methods were used to identify dominant periodic changes in groundwater levels (GWLs), effects of tide, and estimation of aquifer diffusivity using tidal effect parameters. Autocorrelation and cross-correlation analysis were used to estimate the periodic components and lag time between the tide and GWLs, respectively. Spectral analysis was used to ascertain the dominant periodic components in the tide and GWLs by means of estimating amplitude spectrum and power spectrum density. Cross-spectral analysis was used to determine the lag time between the tide and GWLs by means of estimating cross-power spectrum density. Furthermore, wavelet analysis was used to investigate changes of periodic components over the measured period. The estimated amplitudes and lag times were used to estimate aquifer diffusivity. The results identified dominant periodic component with a 12hour period both in the tide and GWLs. However, groundwater level is lag behind the tide with 2-4 hours depending on the distance of the observation wells to the costal line. The wavelet analysis results show no changes of dominant periodic components over the time. Therefore, the estimated amplitude and lag time were used to estimate aquifer diffusivity. The estimated parameter values are 2.72595E-05 h/m<sup>2</sup>, 6.97843E-05 h/m<sup>2</sup>, and 6.14906E-05 h/m<sup>2</sup> from observation wells. These correspond values of transmissivity as 440 m<sup>2</sup>/d, 172 m<sup>2</sup>/d, and 195 m<sup>2</sup>/d, respectively. The estimated transmissivity values are useful for constructing saltwater intrusion models.</p>