A methodology to estimate the residence time in estuaries Frank Braunschweig (Instituto Superior Técnico, Sala 363, Nucleo Central, Taguspark, P , Oeiras, Portugal, ) Paulo Chambel (Instituto Superior Técnico, Sala 363, Nucleo Central, Taguspark, P , Oeiras, Portugal, ) Flávio Martins (Escola Superior de Tecnologia - University of Algarve, Campus da Penha, P Faro, Portugal, ) Ramiro Neves (Instituto Superior Técnico, Av. Rovisco Pais Lisboa, Portugal, ) Overview Introduction The residence time has long been used as a classification parameter for estuaries and other semi-enclosed water bodies. It aims to quantify the time water remains inside the estuary, being used as an indicator both for pollution assessment and for ecological processes. Estuaries with a short residence time will export nutrients from upstream sources more rapidly then estuaries with longer residence time. On the other hand, the residence time determines if micro-algae can stay long enough to generate a bloom. As a consequence, estuaries with very short residence times are expected to have much less algae blooms then estuaries with longer residence time. In addition, estuaries with residence times shorter than the doubling time of algae cells will inhibit the formation of algae blooms (EPA, 2001). The residence time is also an important issue for processes taking place in the sediment. The fluxes of particulate matter and associated adsorbed species from the water column to the sediment depend on the particle’s vertical velocity, water depth and residence time. This is particularly important for the fine fractions with lower sinking velocities. The question is how to compute the residence time and how does it depend on the computation method adopted. The traditional methods are usually based on integral methods. They account for advection and mixing inside the estuary but, because they consider the domain as a whole, they lack to describe accurately the dynamics of the water masses. Using high-resolution hydrodynamic models it is possible to overcome that problem simulating explicitly the transport processes. In this paper, a methodology to quantify residence time in estuaries is proposed. The methodology is based upon the MOHID primitive equation hydrodynamic model, coupled to its lagrangian transport module. Besides the estimation of overall estuary residence time, the subdivision of the estuary into monitoring boxes enables the quantification of residence times inside each box and the water exchange among the boxes. These boxes have two functions: on one hand, they are used to release lagrangian tracers at the beginning of the simulation and, on the other hand, they are used to monitor the tracers passing through them during the simulation. With this approach two basic questions, related to the physical dynamics of the estuaries can be answered:  In which boxes are located the water masses initially released in box i? The answer to this question can show how a specific area can influence other areas of the estuary. In this way, the user looks from the origin of the water masses (one box) to their destination (all boxes).  From which boxes came the water masses that occupy box i at a certain instant? The answer to this question can show the influence over a given area from other areas in the estuary. In this way, the user looks to the destination of water masses (one box), from given origins (all boxes). The Tool - MOHID Methodology Hydrodynamic Module The model was applied to several Portuguese estuaries. Here the results for Tagus Estuary are presented. The hydrodynamic module is used with a high resolution mesh to obtain the hydrodynamic solution. The grid step is variable with a minimum grid step of 300m Lagrangian Module The estuaries are divided into conceptual boxes. These boxes have two functionalities: (i) to release lagrangian tracers and (ii) to monitor the lagrangian tracers passing through them. At the beginning of the simulations the estuaries are fulfilled with lagrangian traces. The total volume of the tracers matches the volume of the estuary. The figures below show the evolution during the first day. Characteristics MOHID is a modular water modelling systems which uses finite volumes. It is written in ANSI Fortran 95 in an object oriented way. The main modules and their characteristics are: Hydrodynamic – solves the three-dimensional incompressible primitive equations. Hydrostatic equilibrium is assumed as well as Boussinesq and Reynolds approximations; Turbulence – implementation of the GOTM system into MOHID; Geometry – cell geometry which permits the implementation of any vertical coordinate. Actually Sigma, Cartesian, Lagrangian and Fixed Spacing are supported; Water Properties – manages the evolution of water properties using an eulerian approach. Responsible for the transport of the properties (advection and diffusion), manages the exchanges with the surface (ex. heat fluxes), bottom (ex. sedimentation), discharges (ex. rivers), sink and sources (ex. nitrogen cycle), etc. Lagrangian – manages the evolution of water properties using an lagrangian approach. Transports lagrangian tracers and manages their evolution (coliform decay, water quality, oil spills…) Water Quality - a zero dimensional water quality model which can be either used by the water properties module or by the lagrangian module. The flow chart below shows the information flux among MOHID’s modules. Water Properties Surface Wind Rain / Evap. Air Temp. Radiation Hydrodynamic Velocity Fluxes Levels Viscosity Diffusivity Mix. Length Properties Shear Properties Density Turbulence Bottom Water Quality Lagrangian Geometry Areas Volumes More information and downloads at : Results Residence Time Two methods were developed: the tracer’s fraction and the integrated contribution. In the first method one can measure the fraction of tracers which are located inside their original box and the fraction of tracers inside any other box/estuary. Defining the residence time equal to the time when just 20% of the tracers remain, either in the original box or the estuary, one obtain the results shown below. The integrated contribution of different origins j for the water in each box i is calculated by the equation below. As the water is renewed in the box, the contribution of the initial water for the actual water inside the box tends to zero and the integral tends to an asymptotic value. This method can be applied to a box or to the estuary as a whole. Tracer’s fraction inside a given box. Tracer’s fraction inside the estuary.