Spectral responses of gravel beaches to tidal signals.

Tides have been recognized as a major driving forcing affecting coastal aquifer system, and deterministic modeling has been very effective in elucidating mechanisms caused by tides. However, such modeling does not lend itself to capture embedded information in the signal, and rather focuses on the primary processes. Here, using yearlong data sets measured at beaches in Alaska Prince William Sound, we performed spectral and correlation analyses to identify temporal behavior of pore-water pressure, temperature and salinity. We found that the response of the beach system was characterized by fluctuations of embedded diurnal, semidiurnal, terdiurnal and quarterdiurnal tidal components. Hydrodynamic dispersion of salinity and temperature, and the thermal conductivity greatly affected pore water signals. Spectral analyses revealed a faster dissipation of the semi-diurnal component with respect to the diurnal components. Correlation functions showed that salinity had a relatively short memory of the tidal signal when inland freshwater recharge was large. In contrast, the signature of the tidal signal on pore-water temperature persisted for longer times, up to a week. We also found that heterogeneity greatly affected beach response. The response varied from a simple linear mapping in the frequency domain to complete modulation and masking of the input frequencies.