Sometimes, if you listen to the sea area forecast, you may hear the visibility at Rosslare or maybe Malin Head described as being "unlimited". But as we noted yesterday in this column, you should not take this information literally; it just means that visibility is greater than say 50 miles, the highest numerical value which meteorologists concern themselves with.
It says that visibility will not be a limiting factor for any normal activity that might be undertaken.
But sometimes, too, you will hear figures quoted for the visibility that may seem to suggest a strange devotion to exactitude.
It might be given, for example, as 38 miles - a very precise figure for such a long distance. But these strange numbers are often arrived at, not by ultra-accurate observing, but by a succession of arithmetical approximations arising as the data is recorded and transmitted.
To assess the visibility, the observer peers into the middle distance and makes an estimate based on his experience and skill. As you might expect, he will normally choose a nice round figure; in the instance quoted, the estimated visibility is likely to be 40 miles.
But now what happens to this information? First, the observer converts it into metric measure and arrives at 64 kilometres. Then, to include it with the other elements in his full weather report for the hour in question, he must translate it into the international code which is used for the transmission of such data.
The code book tells him that the appropriate code for 60 kms is 86, that for 65 kms is 87, and that for 70 kms is 88. He chooses, sensibly enough, the figure 87.
Some time later the forecaster - in this case let us assume it is a "she" - decodes the weather report to read it on the radio. Coming to 87 for visibility, she realises that, bearing in mind the structure of the code, the figure could refer to any value from 62.5 to 67.5 kms; being of a conservative turn of mind and not wishing to exaggerate the visibility, she chooses the lowest value, 62.5 kms - which, when translated into imperial measure becomes 38.8 miles. Wishing again to err, if at all, on the conservative side, she rounds it down to 38.
In real life, neither the observer nor the forecaster indulge in all these complicated sums when actually doing their work; they read the appropriate values from a table. But these tables, being constructed according to the same conservative principles, will give 38 miles to correspond to the figure 87. And that, therefore, is what the forecaster reads out.