Armed with a knowledge of the techniques for analysing time series, we must now get down to the business of examining examples of the efforts that have been made by meteorologists to produce evidence of cycles. The best place to start is with instrumental records that enable us to consider the behaviour of a single variable (e.g. temperature, rainfall or atmospheric pressure) as a function of time. The aim will be to sift through the evidence and draw up a balanced assessment of the case for cycles of any given period. We will review a cross-section of the work that has been done to see which periodicities show up most frequently and which seem to be peculiar to a given record. From this inventory we will then be in a position to move on first to the wide range of data, which contain indirect information about the weather (proxy data), and then on to possible physical explanations of what might have caused such regular fluctuations.

As has been made clear in Chapter 1, there has been a huge amount of work done over many years to demonstrate cyclic behaviour in weather records. For a number of reasons we will pass over much of this work and concentrate on more recent work. First, many of the early studies have been the subject of critical review and have been discounted by many meteorologists.