Scatter Tables (Wave Height and Period)
Each of the tables shown is a joint probability distribution of wave height and wave period, computed by the methods explained previously from observations of wave and winds in a particular directional class. Thus, each entry in each table is the probability in parts per thousand that waves in the respective directional class will lie in the specified ranges of height and period. It must be noted that the waves concerned will typically include a component due to swell in some other directional sector not defined by the data.
The corresponding row and column totals represent the respective marginal probability distributions of wave height and wave period. The wave heights are labelled in metres, and may be assumed to correspond to significant heights as explained in 'Wave Height Data'. Similarly the scale of wave period is labelled in seconds, and may be assumed to correspond with zero crossing period.
The top line on each scatter table specifies the directional class, in 45 degree sectors, from which the waves concerned were coming. It also cites a figure for the percentage of observations reported for that directional class. This is, in fact, the percentage of normalized 'All Wind' observations and is considered to be the best estimate of the probability that the wind-generated components of the sea state will approach from the specified direction.
In some cases the generation of the scatter table has been inhibited. The affected data sets are indicated by a red cross on the Area Confirmation page, and a red highlight on the Season and Direction selection page. This means simply that the number of joint wave and wind observations, available for computing the relevant modelling parameters in the sector concerned, is below a prescribed acceptable level, which varies with the range of wind speed covered by the modelling. In a very small number of cases (nine records in the whole worldwide database) the data set has been rejected because quality control criteria for the parameters have been violated although the number of observations was acceptable.
In some areas where the total sample of observations is low, omissions are quite frequent. It is important to note however that the percentage of occurrence of waves in the directional sector concerned is still cited, and is still valid since it is derived from the more numerous 'All Wind' data.
The total numbers of observations available for the area concerned in the two categories 'Wind / Wave' and 'All Wind' are given on the Season and Direction selection page. The significance of these two categories is explained in full in Appendix_A3. Briefly, the 'Wind / Wave' observations are joint wind and wave data used for deriving modelling parameters, and the 'All Wind' observations are the input used for computing the long term statistics. The corresponding numbers for individual seasons are not shown on the respective seasonal menus because the numbers would be difficult to interpret due to the normalization procedure explained in Appendix_A3. An approximate indication of the numbers of observations available for each season may be derived, however, by assuming distribution of the 'Annual' totals in proportion to the number of months in the season concerned.