In any coastal lagoon, a thorough characterization of hydraulics must recognize the existence of both temporal and spatial variability. The protocol outlined in this chapter addresses three distinct time scales: (1) seasonal variations which define the annual periodicity; (2) subseasonal frequency variations over time scales of several days to two weeks which occur in response to synoptic scale metrological forcing; and (3) higher frequency fluctuations including the diurnal and semi-diurnal variations associated with the astronomical tides, the daily solar heating and cooling cycle, and supertidal small scale meteorologically forced events. The minimum suggested protocol involves seasonal sampling periods of two months each. While four seasons are necessary in temperate zones, in tropical or sub-tropical zones, two may suffice. But where pronounced rainy and dry seasons exist, some effort should be made to document the transition periods. And special attention should be directed toward periods of colonization by target species. Half-hourly measurements made during each season will sufficiently resolve the shortest time scales without compromising important local events. Thirty- to forty-hour low-pass filtering of data collected within any given season quantifies variability over time scales on the order of days or weeks; monthly to seasonal means are created using fortnightly filters and annual cycles are defined by combining seasonal means.
Establishing the existence of spatial variability, and quantifying it in the horizontal and vertical, are important components of a field sampling program in any coastal lagoon. Vertical and horizontal gradients in temperature and salinity, and hence density, separately affect circulation and local mixing; horizontal gradients, especially in salinity, can often partially explain the observed distributions of plant and animal communities. Significant differences in freshwater inflow and in lagoon-shelf exchanges from one lagoon to the next, combined with significant differences in lagoon size, make it necessary to design a sampling strategy that will be appropriate to characterize temporal and spatial variability in every case.
The sampling and analytical methodology suggested in this chapter is necessarily generalized to be applicable to a broad range of coastal lagoons. The protocol described is intended to provide a fundamental understanding of the magnitude and relative importance of physical processes relevant to fisheries and aquaculture. Within this restricted scope, several common physical variables have not been included. For the purpose of characterizing coastal lagoons in a hydraulic sense, one must know basin geomorphology (including perimeter, depth, surface area and volume), and one must measure water level, current speed and direction, temperature and salinity (including vertical and horizontal gradients). In some specific settings, it may be apparent at the outset that modifications to this protocol are required. When specific questions are asked, modifications (usually additions) to the sampling and/or analysis will almost certainly be required.
The sampling and analytical methodologies described in the following four sub-sections include characterizing the regional climate, establishing local bathymetry and freshwater run-off, and quantifying physical processes over three distinct and dominant time scales using measurements made in the lagoon itself. Taken together, the sampling protocol described here will provide an adequate characterization of the hydraulics. When the relationships between hydraulics and productivity have been quantified, it should then be possible to estimate increases (or decreases) in productivity which can be expected when the hydraulic regime is modified from a pre-existing state.
In many cases, climatological statistics will be available for the study area, and they should be utilized fully. An understanding of both regional climatology and local weather will be useful in setting up the necessary and sufficient sampling program for collecting in situ data. Climatology will help align the seasonal sampling to coincide, for example, with wet and dry seasons or with midwinter and midsummer seasons.
As a general rule, spatial gradients in climatological statistics are considerably greater across the coastal zone than along the coast. An existing weather station at the coast may provide usable data even if it is several tens of kilometers from the study site. On the other hand, if the weather station is more than a few tens of kilometers inland, it may not be satisfactory owing to topographic boundary frictional effects, and instrumentation may have to be brought to the lagoon to assemble a local data base.
Relevant weather and climatological variables include air temperature, any measure of atmospheric moisture, air pressure, wind speed and direction and precipitation. The annual cycles of all of these variables can be defined from multi-annual monthly mean values calculated from daily accumulations or 3-hourly observations recorded at the weather station. Seasonal variations in climatological variables may be related, directly or indirectly, with similar long-period variations in lagoon hydraulics. For example, seasonal variations in wind speed, combined with monthly rainfall amounts, may be important in explaining the presence or absence of vertical stratification. Similarly, month-to-month variations in wind direction may explain seasonal differences in the circulation of the lagoon. If an investigation is intended to reveal cause-and-effect relationships, the climatological setting will provide an invaluable backdrop for interpreting in situ measurements.
An important preliminary step in addressing many questions related to the hydraulics of a coastal lagoon involves a careful bathymetric survey of the lagoon. The three primary objectives are (1) to determine the volume of the lagoon, (2) to calculate the surface area, and (3) to determine the shape of the perimeter, paying attention to the existence, location and extent of primary and secondary channels. Several basic measures of flushing, such as the freshwater residence time or the hydraulic replacement time, require an estimate of the volume of the body of water under investigation. The surface area of a lagoon must be known if freshwater gains by direct precipitation, or evaporative losses are to be quantified. Knowledge of primary and secondary channels is necessary for understanding both the response of these channels to lagoonal influences and the internal circulation of the lagoon, as well as for planning field studies.
The amount and source of surface fresh water entering the lagoon should be quantified at gauging stations, or estimated from periodic visits. If rivers and channels carrying water to the lagoon cannot be gauged, discrete measurements must be made over time intervals on the order of a few days to record effects of individual and transient storm events. The volume of fresh water discharged into the lagoon, corrected for surface evaporation and direct precipitation, is an important variable for estimating flushing. The temporal variability of freshwater inflow, over time scales ranging from days to seasons, is of primary importance to aquaculture and fisheries.
In some cases significant groundwater input exists. Quantifying this has proved extremely difficult where it has been attempted. A groundwater hydrologist should be consulted where necessary, or this source of freshwater input can be roughly estimated by difference.
In most cases, half-hourly to hourly observations will be sufficient for quantifying super semi-diurnal up to diurnal variations in in situ variables. Hourly observations of water level, made over a complete synodic month (e.g., full moon to full moon-29 days), can be used in a harmonic analysis to quantify the harmonic constants (amplitudes and phase angles) of the principal semi-diurnal and diurnal constituents, plus those of the compound tides and over-tides characteristic of shallow-water settings. Harmonic constants are not entirely constant as a result of low-frequency variations in water depth, and local nontidal forcing in a variety of forms. In practice, then, it is desirable to collect time series over longer time intervals, calculate harmonic constants from overlapping 29-day segments, and then vector-average the harmonic constants to obtain more representative values. A total record length of two months is recommended. This will provide four 29-day segments with a 10-day offset between starting times. Longer records are necessary to investigate harmonic constants of the fortnightly, monthly, semi-annual and annual constituents, and these cannot be quantified with data from the four seasonal field studies suggested here.
Hourly current measurements at a given study site can be analyzed in a similar manner to quantify the ebb and flood of the tide, and to estimate the tidal excursion associated with a given constituent. As a general rule, current measurements from a particular study site represent the surrounding area to a much lesser extent than do water level measurements. Currents are influenced by local bathymetry. The number and spacing of study sites necessary to characterize the tidal or wind-driven circulation of a lagoon will depend upon the size and geomorphology of the lagoon. Initially however, current meters should be deployed near inlets and in the primary channels.
An important component of the in situ collection of time series data involves direct readout measurements of hydrographic variables (temperature and salinity) and current speed and direction. In addition, turbidity should be recorded at each station. Such measurements complement time series data by improving spatial resolution. It is not possible to specify the horizontal station spacing necessary and sufficient to resolve spatial gradients adequately. Owing to the great diversity in coastal lagoons, adequate spacing in one case may be excessive in another; and insufficient in a third. One can only state the objectives that the data are intended to attain: Vertical profiles are intended to reveal the causes and effects of layering of the water column; hydrographic data from a series of stations separated horizontally and collected repeatedly over one or more tidal cycles can be interpreted in terms of water movement within the lagoon. In situ salinity data are especially helpful in describing the transition from fresh to marine conditions, as well as documenting any preferred paths followed by fresh and salt water in leaving and entering the lagoon, respectively. In areas of well-defined gradients, data collected over any half tidal cycle (e.g., slack water to slack water) will be useful for quantifying the tidal excursion, as well as for estimating mixing. Measurements of water clarity are useful for interpreting and predicting density stratification. Turbid water stratifies more readily, because insolation is absorbed in near-surface layers. Stratification of low turbidity water generally requires vertical gradients in salinity. Vertical profiles of current speed and direction made from a boat anchored at a long-term mooring will be helpful for interpreting time series data from a single level in terms of vertically averaged net transport. Finally, if direct read-out measurements can be obtained in profile form at a series of stations spaced laterally along a line perpendicular to the predominant flood and ebb directions, one can estimate water and salt budgets, as well as volume transport, and thus the tidal prism. Such measurements are fundamental to an understanding of tidal and nontidal mixing and flushing.
When designing field studies which involve both water level and current measurements, an important point regarding the interrelationship of these two variables should be kept in mind. Specifically, water level measurements can be used in some cases to calibrate current measurements. Variations in water level, combined with the surface area of the lagoon, provide an indirect measure of the volume of water arriving and leaving on the flood and ebb tide, respectively. Current data may be combined to quantify volume transport in this way, but as noted above, local perturbations may influence calculations significantly. Whenever possible, one should take advantage of this relationship, and collect water level data to calibrate current measurements, as well as to answer questions related to water level itself. In some cases water level data may significantly reduce the need for current data, and water level data is considerably easier to obtain than is current data.
In situ measurements of temperature, conductivity and other relevant hydraulic variables must also be closely spaced in time if one is to quantify tidal period variations, or to average out tidal effects to investigate longer-period variations. Sampling must be carried out in such a way that measurements intended to define the longer time scales are not contaminated by shorter period variability. When making direct readout measurements of currents, tidal effects can be minimized if one samples consistently relative to the phase of the tide. For example, if one starts at an inlet at slack water, it may be possible to follow slack water into the lagoon, sampling at each location as slack water occurs locally. One can minimize effects of diurnal heating by sampling consistently within the diurnal cycle. Specifically, the diurnal temperature curve passes through the 24-hour mean temperature at about noon and midnight. Sampling at those times would provide values which approximate closely the daily mean; sampling at a given station consistently at any other time would minimize the effects of diurnal heating and cooling and emphasize longer-period variations, but temperatures would be systematically high or low.
Longer-period variability in hydraulic variables are considerably more difficult to quantify, but they are especially significant in coastal lagoons. Wind-driven circulation takes on greater importance where tidal amplitudes are low; warming and cooling cycles associated with frontal passages or changes in cloud cover are exaggerated in the shallow waters of a lagoon. Nontidal variability is not only quasi-periodic at best in a temporal sense, but the magnitude of the response to meteorological forcing can exhibit considerable variability as well. In practice, a long time series will be needed to characterize these features adequately - both time scales and magnitudes may have to be described as a range, rather than a value. Nontidal variability will change with the seasons in many settings. Often, summer months are relatively quiescent, whereas in winter months variables are influenced by vigorous frontal passages at weekly or bi-weekly intervals.
Long-period variability stands out clearly when semi-diurnal and diurnal period fluctuations have been removed. If data have been recorded at hourly intervals, many numerical filters are available for this purpose once data have been digitized and put in a computer. A plot of the low-pass filtered output will indicate the magnitude of the reponse to meteorological forcing, as well as the characteristics time scales. In practice, a long time series will be needed to characterize these features adequately - both time scales and magnitudes may have to be described as a range, rather than a value. Each study site, and each study may be different, but in most cases a two-month record will contain enough events to provide a representative characterization of local conditions.
The longest time scale that can be quantified in most studies is the annual cycle. The annual cycle can be represented adequately by four seasonal means, once the seasons have been defined and aligned using local climatological data. Variables which show significant seasonal variability include temperature, salinity (if climatological data include wet and dry seasons), resultant flow (if seasonal variations in wind speed and/or directions are indicated) and vertical stratification (depending upon seasonal variations in wind speed and freshwater outflow). Climatological data from a nearby weather station will be useful for deciding which in situ variables vary significantly over this time scale. Ideally, tidal harmonic constants should be quantified seasonally. Both amplitudes and phase angles can change over the course of a year, as a result of variations in water depth and wind speed, among other things. Harmonic constants should be vector-averaged within each season, and seasonal vector averages should be listed along with seasonal means of the other hydrographic and hydraulic variables. Variables which will show significant seasonal variability include temperature, salinity (if climatological data indicate wet and dry seasons), resultant flow (if seasonal variations in wind speed and/or directions are indicated) and vertical stratification (depending upon seasonal variations in wind speed and freshwater outflow).
Year-to-year variations exist in many variables, however, in many cases they are relatively subtle. Midwinter temperatures in many settings will vary from one year to the next according to the intensity and frequency of frontal passages. Salinity may vary from one year to the next according to rainfall amounts on the lagoon and its watershed. In any case, once the lagoon is characterized hydraulically in terms of its response to external forcing, such year-to-year variations can be modelled.
Once the principal physical forcing mechanisms in the lagoon, and the lagoon's hydraulic responses to them, are understood, the hydraulic model can then be used to predict the effects of manipulations. We must re-emphasize, however, that only a rudimentary knowledge exists of the relationships between the physics and biology. Therefore, the primary function of developing models for lagoons is to be able to characterize the biological responses to the physics. In most cases, synoptic biological investigations should be conducted-especially those aimed at determining the spatial and temporal variability of production within the lagoon.
