Diagnostic analysis of estuarine residence times

Abstract

Residence times are a conceptually important indicator of the ecological behavior and self-cleaning capacity of estuaries. However, the quantification of residence times is ambiguous and can be often misleading. Most traditional approaches are integral, providing a "single" characteristic number for the whole system, thus ignoring the strong spatial and temporal variability of residence times in estuaries. In this paper, we propose a new methodology that overcomes these limitations through the diagnostic use of numerical particle tracking for the analysis of residence times in estuaries. Unlike the traditional methods, this approach can also be applied to a wide range of situations that include residence times of pollutants that do not enter the system by the main riverine source (e.g. run-off), or systems without riverine input (lagoons).

In spite of their attractive properties, numerical methods have been scarcely used for evaluation of residence times in estuaries. Two different approaches have been applied using particle models and concentration (transport) models. However, past studies using numerical models were focused on localized problems, and detailed, system-wide studies of the spatial and time variability of residence times have not been done yet.

The methodology proposed herein is based on very accurate numerical tracking of large number of particles, carefully distributed in the domain and released at several instants in time, which address both the variability within the tidal cycle and with neap/spring tides. The results are then integrated through histogram analysis into detailed, yet condensed descriptions of the space- and time variability of residence times.

The first part of this paper describes our particle tracking algorithms, and studies their numerical accuracy as a functions of controlling parameters: machine accuracy and flow characteristics. Numerically very accurate tracking is then used in the second part to diagnostically assess the spatial and temporal characterization of residence times, using two-dimensional, depth-averaged flow simulations in the Tagus estuary (Portugal). Finally, we integrate the results from the spatial and temporal analysis into a single histogram, and compare it with a traditional integral approach to freshwater residence time.

At a conceptual level, our results support the use of particle-tracking methods as a basis for the evaluation of residence times in estuaries, and suggest that histograms are a good tool for residence time characterization. However, caution should be used in the application of the concept for practical purposes, since the accuracy of the method is significantly dependent on the quality of the driving flow field.

Key Words: Residence Time, Particle methods, Numerical methods, Estuaries