Most tide stations require a stilling well that is connected to the sea through a hole located below the low tide level, sometimes equipped with a pipe whose length depends on the topography of the site.
Stilling wells are necessary for several reasons. The main purpose is to filter out the fluctuations in the sea level due to waves, swell or seiches in order to obtain a horizontal surface whose height is identical to that of the external sea level, averaged over a period of time equal to the measurement sampling time. Moreover, the shelter mounted over the stilling well protects the measuring equipment from the weather.
However, the hydraulic system consisting of the stilling well has the disadvantage of not being a linear filter. It is therefore important to examine the errors associated with the internal response of the system based on variations of the external sea level. This issue is discussed in Appendix C of the book La Marée Océanique Côtière by B. Simon (2007).
Other errors related to wells
In addition to the intrinsic error (independent of the measurement system) related to the nature of the response of the stilling well hydraulic system, other errors (also independent of the measurement system) of different origins may occur and bias the response. The system must be monitored carefully for blockages in the hole and the pipe due to siltation and concretions (related to the development of marine life).
This initially results in a phase difference without a substantial impact on the amplitude. It is often undetectable, but seriously compromises the quality of the measurements over time. The pipe must be inspected and cleaned yearly to avoid this problem.
- Errors due to density differences
The density difference between the outside and inside of the well is another source of error which must be taken into account for accurate measurements. During a tidal cycle, the temperature and salinity of coastal waters can vary significantly, particularly in summer and near the mouths of rivers. The well is connected to the ocean through a small hole. The water in the well is never completely renewed. Because the hole is located at the bottom of the well, the water that enters as the tide is rising is often denser than the water outside at equal depth. As a result, the level in the well is lower than the level outside. Conversely, if lower density water has entered the well at low tide and is not renewed, the opposite effect occurs: the water in the well is lighter and the level is higher than outside.
With a deep well of some ten meters (large tidal ranges), a relative difference of one-thousandth in the density induces a difference in height of around one centimetre.
Another often unrecognised cause of error related to differences in density comes from the condensation of atmospheric water vapour on the walls of deep large diameter wells. This creates a layer of fresh surface water in the well, which tends to increase over time if the evaporation is not sufficient. Knowing the relative difference in the densities between seawater and fresh water, the accumulation of a 1m layer of fresh water in a deep well induces a higher measurement of 3 centimetres compared to the external sea level.
It is not a common practice to measure the vertical distribution of densities in the wells of a tide gauge and on the outside of the well. However, there are modern techniques that make it possible to do so without difficulty. In systems designed to provide accurate measurements for scientific applications, this practice could be useful, at least as a check, in addition to tide staff and water level indicator tape readings.
In estuaries as well, differences related to density variations almost always occur. In a tidal river, for example, the density of the water in the well is lower than the external density and level differences can reach over 6 centimetres for 2 meters in amplitude.
Errors due to currents near the hole
Water movements near the hole (continuous or tidal currents, wave or swell orbital velocities) cause undesirable pressure fluctuations that affect the level inside the well, due to the Venturi effect.
This negative pressure at the entrance to the hole is at a maximum when the plane of the hole is parallel to the flow upstream of the well, causing the lower level to drop nearly 20 cm with a current upstream of 1.5 m/s ≈ 3 knots.
The curves in this figure show that improvements can reduce measurement errors. By changing the shape and location of the hole whose characteristics are visible on the figures (see diagram above) it is possible to reduce the error by more than one order of magnitude. The optimal solution for a well with a hole (no pipe) is shown by curve g.
Note also that there may be differences in atmospheric pressure between the outside and inside of the tide station (where the stilling well is located): low pressure caused by strong winds (Venturi effect around the shelter) or high pressure due to warmer air inside if the shelter is too tightly sealed. These differences generally induce errors of around one centimetre. To our knowledge, this type of error has not been routinely studied.
Fortunately, all the problems mentioned above correspond to extreme cases. They are generally small or negligible for very high tidal ranges in port sites where the tide gauges are protected from waves, currents and siltation. Though there may be frequent differences between the outside and inside of the well, they are generally low. It would be inappropriate to modify an installation that has worked for decades just to correct this error, as this would interfere with observations of events with long periods.
The height differences are similar in origin to the differences that may exist between the inside and outside of a port, a roadstead or bay.
To find out more:
- Simon B. (2007). La Marée - La marée océanique et côtière. Edition Institut océanographique, 434pp.
Last updated: 12/12/2012