Significant height: one statistic, many waves

A sea state contains waves of different heights. One number can summarise that diversity without describing every wave.

Significant height is one such statistic, and its definition belongs alongside its value. After recognising height, period and swell, we compare three summaries of one list and explore why significant height is neither an upper limit nor a promise for the next wave.

For the time-domain definition H1/3, individual heights are ordered and the highest third is averaged. Each wave’s height measures vertical separation between trough and crest under the chosen identification method.

It is not merely crest elevation above mean level. The following list is already ordered and represents twelve invented waves, with no actual place, time or buoy record.

The twelve waves are not all 1.95 m high. Ten are below that value and two above.

Of the four values in the highest third, two remain below its mean: 1.5 m and 1.8 m. Neither “one third of waves equals Hs” nor “one third exceeds Hs” automatically describes the definition.

A group mean is not an exceedance count. This small list reveals a distinction hidden in a bulletin containing only one number.

Ordering heights also removes temporal order. The highest four can occur together or alternate with smaller waves while retaining exactly the same H1/3.

Both sequences can be constructed from the same twelve numbers without changing sum, mean or maximum. The figure’s horizontal axis contains ordered heights, not seconds: spacing between bars therefore cannot yield a period.

Sequence or time between waves would require additional temporal information.

Many products use spectral significant height Hm0. It is calculated as 4√m0, where m0 is surface-elevation variance represented by the spectrum, in m².

The factor 4 belongs to the conventional definition. Hm0 and H1/3 are related, but are not exactly identical operations on a height list.

Method, analysed interval and processing assumptions help explain differences between values.

A longer sample may include a larger wave without invalidating the previous sample’s statistic. It may also span changing conditions, in which case one combined summary loses temporal information.

Twelve values were chosen to make arithmetic legible, not to represent a normal observation duration. Wave count, duration and identification method belong in the record.

Hs alone does not specify wave period, direction or shape. Two sets sharing one summary can have different distributions and components.

A sample maximum is not an absolute ocean maximum; a spectral formula does not determine the greatest impact on a boat. The next lesson opens the spectrum to reveal information a summary omits, keeping application to actual conditions and vessels separate.