The study of coastal groundwater depth and salinity variation using time-series analysis

https://doi.org/10.1016/j.eiar.2006.06.003Get rights and content

Abstract

A time-series approach is applied to study and model tidal intrusion into coastal aquifers. The authors examine the effect of tidal behaviour on groundwater level and salinity intrusion for the coastal Brisbane region using auto-correlation and spectral analyses. The results show a close relationship between tidal behaviour, groundwater depth and salinity levels for the Brisbane coast. The known effect can be quantified and incorporated into new models in order to more accurately map salinity intrusion into coastal groundwater table.

Introduction

Coastal intrusion of saltwater has caused major disasters around the world. Tidal dynamics is an integral component of coastal hydrology and the periodic effect of tidal pressure on water level coastal aquifers has been studied for a number of years (Carr and Van Der Kamp, 1969). The authors noted that a propagating tidal-wave dampens the amplitude of the groundwater head and these tidal fluctuations have an effect on the energy or pressure being exerted on the water within the coastal aquifer. The effect of this may be important to solute transport and in this regard analytical models have been developed to study tidal fluctuations and salinity levels of coastal aquifers. Both Philip (1973) and Smiles and Stokes (1976) demonstrated that tidal fluctuations on vertical shorelines have a nonlinear influence on the groundwater level, even with a zero net discharge to the sea. Recently, Ataie-Ashitiani et al. (1999) developed a variable-density groundwater model in order to analyse the effects of tidal dynamics on seawater intrusion in an unconfined aquifer. Their results demonstrated that effects from tidal pressure resulted in seawater travelling further inland and that tidal behaviour had an increased effect when the shore was a sloped beach opposed to a vertical shore face.

Time-series analysis is becoming an increasingly important method of studying temporal variations in groundwater quality. The method has been used to determine which hydrological processes are the major influences (Duffy and Gelhar, 1986). Mangin (1984), Padilla and Pulido-Bosch (1995), and Laroque et al. (1998) studied groundwater dynamics in karst-type aquifers. Lee and Lee (2000) used time-series to study the effect of rainfall on water in coastal aquifers. More recently, Kim et al. (2005) applied time-series analysis to study the effects of tidal pressure on groundwater quality. Despite the broad use of time-series analysis in studying temporal variations in groundwater quality, this method has been rarely employed to quantify the causal relationship between tidal fluctuations and salinity levels in coastal aquifers. In this paper, in the main, the time-series analytical approach of Kim et al. (2005) is used to study the temporal variations of coastal groundwater quality, in particular the salinity level, and how they vary over time. The main aim is to study the effects of tidal dynamics on groundwater quality using time-series method applied to the Brisbane coastal lowlands.

Section snippets

Problem description

The permeation of seawater through coastal shores and eventual intrusion into unconfined coastal aquifers affects the salinity levels of the groundwater. The pressure exerted by the seawater, and correspondingly, the degree of seawater intrusion varies with the tidal fluctuations (Fig. 1).

Level of rainfall is an additional factor to the system. An increase in rainfall eventually dilutes the groundwater, reducing the salinity levels to a certain degree. The model presented here incorporates

Method

In a manner similar to the scheme developed by Mangin (1984) to study karst-type aquifers, the coastal aquifer is treated as a “black-box”, which transforms, retains, and eliminates input signals to ultimately produce a single output signal. For the coastal aquifer model presented here, the input signals are the tidal fluctuations and local rainfall levels, while the output signal is salinity level of the groundwater. Time-series analysis is applied here to gain insight into the input–output

Correlation and spectral analyses

Performing an auto-correlation analysis yields a quantitative measurement of the linear dependency of certain process values over a given time range (Box et al., 1994). The auto-covariance is a measure of variation a factor over itself over time and is given in Eq. (1):C(k)=1nt=1nk(xtx¯)(xt+kx¯)where k is the time-lag (k = 0 to m), n is the number of time-series measurements, x is the value of the process, is the mean of the process and m is the cutting point. The cutting point determines

Site, hydrological and tidal data

The study area is the Brisbane River estuary system. The measurements for the tidal heights were taken at the Brisbane Bar region (27°21′S, 153°10′E). The salinity data was collected at the Indooroopilly Bridge, which is located several kilometres upstream from the mouth of the Brisbane River. Other relevant data sets were obtained from Cartwright (2004) who conducted an in-depth analysis of the coastal dynamics of the Brisbane estuary.

The hydrological and tidal data studied was for a 20-day

Results and discussion

The hourly tidal height for a 20-day period is shown in Fig. 2. The tidal and salinity measurements both produced comparably smooth results, eliminating the need for smoothing techniques. The graphed tidal heights show the expected periodic behaviour with a period of roughly 12 h. These results show that the second longer periodic behaviour of spring times begins to appear.

The auto correlations of tidal and salinity in Fig. 3, Fig. 4 reflect that the likelihood of saltwater intrusion being

Conclusion

The time-series approach of analysing the effect of tidal behaviour on coastal groundwater dynamics was shown to be another effective way of analysing groundwater depth and salinity variation. The results showed a strong relationship between tidal behaviour, groundwater level and salinity concentration. This also suggests that water table depth and salinity will be influenced by changes in coastal weather conditions. The seaward boundary in any new 3D flow and transport models of coastal

Dr G A Tularam has worked as a senior postdoctoral fellow at the University of Wollongong working on acid sulphate soils before joining the Griffith University in 2001, where he initiated the research on groundwater contaminant transport and concentrated research salinity intrusion in coastal lowlands. Dr Tularam is presently developing new models of salinity but also furthering his work on time-series and its applications to new areas. Dr Tularam lectures in mathematics, statistics and

Cited by (19)

  • Effects of land use and irrigation practices on Ca, Mg, K, Na loads in rice-based agricultural systems

    2014, Agricultural Water Management
    Citation Excerpt :

    Important factors that affect salt balance in agricultural lands are irrigation practices, land use and crop distribution, drainage system characteristics, weathering effects and climate conditions (Tedeschi et al., 2001; Haag and Kaupenjohann, 2001; Causape et al., 2004; Isidoro et al., 2006; Bhat and McDowell, 2007). The soils in lowland coastal regions are particularly vulnerable to salinity due to high water table and sea water intrusion (Tularam and Keeler, 2006). Moreover, the need for continuous monitoring of water and soil quality is important for the maintenance of agricultural ecosystem health as well as for food production and security when the main cultivation in such areas is rice (Wallace, 2000; Antonopoulos, 2010).

  • A lattice Boltzmann model for groundwater flows

    2007, International Journal of Modern Physics C
View all citing articles on Scopus

Dr G A Tularam has worked as a senior postdoctoral fellow at the University of Wollongong working on acid sulphate soils before joining the Griffith University in 2001, where he initiated the research on groundwater contaminant transport and concentrated research salinity intrusion in coastal lowlands. Dr Tularam is presently developing new models of salinity but also furthering his work on time-series and its applications to new areas. Dr Tularam lectures in mathematics, statistics and time-series.

Paul Keeler completed his honours from Griffith University and is presently taking up PhD in Melbourne University. During his honours, Paul did courses and projects such as specialized course in time-series under the supervision of Dr Tularam. Paul's interests also lies in applied mathematics.

View full text