**ABSCISSA **- the horizontal axis (x-axis) of a graph or plot: "The year is shown on the abscissa, while the rainfall is shown on the ordinate (y-axis)".

**AUSTRAL** - pertaining to the Southern Hemisphere: "June through August make up the austral winter".

**BOREAL **- pertaining to the Northern Hemisphere: "While June through August make up the boreal summer, the SST often peaks in early September."

**CORRELATION** - the degree of systematic linear relationship between two variables. A positive correlation implies that when one variable is above its mean,
the other one also tends to be; and likewise for both tending to be below their means. A negative correlation implies that when one variable is above its
mean, the other one tends to be below its mean, and vice versa. A coefficient of linear correlation ranges from -1 for a perfect negative correlation,
through zero for no relationship, to +1 for a perfect positive correlation. "Among adults, height and weight are positively correlated to a moderately high
degree."

**ENSO** - El Niño/Southern Oscillation: "The 1997-98 ENSO episode (not covered in this atlas) will go down in history in similar fashion to 1982-83".

**MEDIAN **- the middle-ranking member of a set of numbers: "Of the numbers 2, 3, 5, 9 and 17, the median is 5 even though the mean is over 7."

**SST** - sea surface temperature: "In early 1983, the SST anomaly along parts of the Ecuador and Peruvian coasts exceeded +8øC."

**STANDARD DEVIATION **- the amount of variability from case to case, as computed by summing the squares of the deviations of each case from the overall
mean, dividing by the number of cases, and then taking the square root: "Even though the amount of rainfall at Atuona, French Polynesia is not much
higher in austral summer than in austral spring, the year-to-year standard deviation is markedly higher in summer than spring."

**STANDARDIZED**- scaled by subtracting the mean and then dividing by the standard deviation: "The data were standardized so that the rainfall anomalies
at one station and season could be readily compared with those of other stations and seasons."

**STATISTICAL SIGNIFICANCE**- the state whereby the probability that a relationship (as shown by a correlation coefficient or a difference between two means)
could have come about by chance alone, is below a given low level--such as below 5 percent (0.05). Given an underlying relationship of a certain strength
(e.g. the number of tropical Pacific cyclones occurring east of the date line as a function of the ENSO situation), statistical significance becomes easier
to obtain as the sample size increases, since it becomes clearer with more cases that the suspected relationship is not accidental. Alternatively, for a given
sample size, statistical significance increases as the strength of the underlying relationship increases. "During boreal winter at Christmas Island, the
relationship between El Niño and rainfall amount is so strong that only about two ENSO cases would be needed in order to attain statistical significance".