Tide Datum: Chart Zero Is Not Prediction Zero

The chart said 2.1 metres over the bank. The tide table said low water would be 0.7 metres. The skipper added the numbers, expected 2.8 metres, and watched the sounder settle at 3.2. The extra 40 centimetres was not a malfunction. It was the gap between the chart's zero and the tide table's zero, and on a falling tide that gap is the difference between a comfortable pass and a bent prop.

tides
Cover illustration: Tide Datum: Chart Zero Is Not Prediction Zero

Key facts

DatumWhat it isWhere it is used
Chart datumReference surface for charted soundingsAll nautical charts
Mean lower low waterAverage of the lower daily low over a datum epochUS chart datum
Lowest astronomical tideLowest level predictable from astronomyUK and many other chart datums
Mean high waterAverage of all high watersBoundaries, some predictions
Mean sea levelAverage of the water level over an epochVertical reference, trends
Station datumFixed mark at a gaugeRaw gauge records
Prediction datumWhatever the table's header statesTide tables and apps

Two numbers, two zeros

Datum arithmetic is simple once the two references are identified, and dangerous when they are assumed. The chart and the tide table are produced by different processes for different purposes, and there is no rule that they share a zero. A nautical chart does not print the depth of the water. It prints the depth below a chosen reference surface, so that most of the time the real depth is greater than the charted number. That reference is the chart datum. The same logic applies to a tide table. The predicted height is measured from a reference surface, and that surface is stated in the table's header or metadata rather than in the numbers themselves. Two tables can show the same coast with different zeros and neither is wrong. The error appears when the two are combined. Adding a tide height to a charted depth assumes both are measured from the same surface. When they are not, the answer is wrong by the offset, and the error can be in the unsafe direction.

The datums that matter

In the United States the chart datum is generally mean lower low water, the average of the lower of the two daily low waters over a long datum epoch. The National Ocean Service computes it from tide observations at each station, and publishes it with the station's tide predictions. In the United Kingdom and many other countries the chart datum is usually lowest astronomical tide, the lowest level that can be predicted from astronomical forces alone. Because it is the lowest predicted level rather than an average of lows, it sits lower than mean lower low water in most places. Mean sea level sits between high and low water and is used for vertical reference rather than for charting. Station datum is a fixed physical mark at a gauge, chosen for practical measurement, and it can sit at any level the operator found convenient. The names matter less than the geometry. What you need to know is which surface is lower, because that decides whether the arithmetic adds or subtracts.

Where the gap comes from

Mean lower low water is an average, so about half the lower lows fall below it in any given epoch. Lowest astronomical tide is the lowest predictable level, so almost nothing falls below it. The second surface therefore sits lower than the first, and the difference between them is the gap. The size of the gap scales with the local range. In a region with a large spring range, the two zeros can be tens of centimetres apart. In a region with a small range they may be nearly identical, which is why the error is easy to miss until you travel or until you use a chart from another country. The gap also changes when a national agency recomputes its datums. A new epoch averages a different set of years, and sea level has moved in the meantime. A chart and a tide table published decades apart can be internally consistent and mutually incompatible.

Reading the header of a tide table

Every official tide table states its reference level, and every table you use should be checked for it before any depth calculation. The statement is usually in the first pages or in the metadata panel of an app. The check is worth doing once per source, not once per trip. A table that states "heights are above mean lower low water" behaves one way; a table that states "heights are above chart datum" behaves another. An app that states nothing is not usable for depth work. Official predictions account for the astronomical positions of the Sun and Moon, including the nodal and perigean modulations. That means the predicted height is already the full astronomical height above its stated datum. No further astronomical correction is needed; only the datum conversion is.

The arithmetic with a worked example

Suppose the chart is drawn to lowest astronomical tide and shows 2.1 m over a bank. The tide table predicts 0.7 m above mean lower low water, and the published offset between the two datums is 0.4 m, with mean lower low water above lowest astronomical tide. The tide height above the chart datum is 0.7 plus 0.4, which is 1.1 m. The depth over the bank at that moment is 2.1 plus 1.1, which is 3.2 m. Subtract the vessel's draft and a margin from 3.2 to get the clearance under the keel. If the offset ran the other way, with the chart datum lower than the prediction datum, the tide height would be reduced before adding. The rule is to convert both numbers to the same surface first, then add. Doing the sum and then adjusting is how sign errors survive. Unit confusion compounds the problem. A chart in fathoms, a sounder in metres and a tide table in feet can produce a plausible-looking number that is wrong by a factor of three. Check the units before the arithmetic, not after.

Drying heights run the other way

Not every charted number is a depth. Features that uncover at low water are charted as drying heights, measured above chart datum and usually marked as such. A rock with a drying height of 1.2 m is 1.2 m above the chart datum, so it is awash when the tide height reaches 1.2 m. Drying heights are a useful check on the datum. If the tide table says the height is 1.3 m and a drying rock marked 1.2 m is still visible, one of the two numbers is not on the datum you assumed. The rock is a free physical datum. The same logic applies to bridge clearances and overhead cables, which are usually given above a stated water level rather than above chart datum. A clearance figure without a reference level is not a clearance.

The station-datum trap

Some tide stations publish predictions relative to a local station datum rather than to the chart datum. That station datum is a fixed mark at the gauge, chosen for practical measurement, and it may sit at any level the operator found convenient. A prediction referenced to a station datum can look perfectly normal and still be unusable for depth calculations without the conversion. The numbers are correct; the zero is not the one the chart uses. The same trap exists in harbour guides and mobile apps that copy a station's raw numbers without copying its reference level. A number without a datum is not a depth, and a table without a header is not a source.

Old charts and new epochs

A chart's soundings come from a survey made at a particular time. The datum may be stable, but the seabed is not. Sand moves, channels migrate and dredged areas silt. A chart that was accurate when published can be wrong by a metre in a dynamic inlet. The datum itself also changes. National agencies recompute tidal datums when the observation record is long enough and when sea level has moved. A chart and a tide table published in different decades can disagree by more than the survey error. The practical response is to treat any chart older than the local survey cycle as a guide rather than a truth. Compare it with the sounder, with local notices and with the latest edition before relying on a narrow pass.

A field calibration

At a predicted low water, in calm conditions, compare three numbers: the charted depth, the predicted height and the sounder reading corrected for draft. If the result differs from the calculation by more than about 0.3 m, stop and check the datum, the units and the table's reference level. The check is especially useful in a new area. It costs one slack tide and converts an assumption into a measurement. It also reveals whether the chart's zero is lower or higher than the prediction's zero, which is the sign that matters for safety. Repeat the check at a second stage if possible. A datum error is a constant offset, while a survey error varies with position. Two checks at different heights separate the two.

When the datum is unknown

If the datum cannot be established, assume the worst. Treat the charted depth as the depth at the lowest predicted level and do not add the tide height at all. That sacrifices accuracy and buys safety. For a shallow-draft boat the same rule applies with a smaller margin. For anything with a keel, an unknown datum is a reason to take the wider channel or wait for more water, not a reason to do the sum and hope.

Mistakes that ground boats

Adding a tide height directly to a charted depth without checking datums. Assuming every country uses mean lower low water. Reading a station-datum prediction as a chart-datum height. Mixing feet, fathoms and metres. Treating an old chart as a current survey. Using a tide app that does not state its reference level. Checking the arithmetic only after the boat touches bottom. Ignoring drying heights when judging whether a rock is covered.

A one-page method

Write the chart's datum, the tide table's datum and the published offset at the top of the passage plan. Convert the predicted height to the chart datum before adding it to the soundings. Subtract draft and margin last, and write the clearance in the plan. Then compare the result with the sounder during the passage and note any difference. After a few passages the local offset is calibrated, and the plan becomes a measurement rather than an assumption. The chart, the table and the sounder are three different instruments. Datum arithmetic is what allows them to be read as one.

Editorial note

Related reading: tide height versus current, tide lag in bays and estuaries, reading nautical charts.

Sources

  1. Tidal Datums — NOAA Tides and Currents · T1
  2. Tides and Currents Glossary — NOAA Tides and Currents · T1
  3. Tidal Glossary — Permanent Service for Mean Sea Level · T2