ABSTRACT
Meteorological data are collected in time, and time series analysis can be defined as the analysis of data collected in time. The standard statistical tests, such as a t-test for testing whether two means are significantly different, assume that the observations are statistically independent. When data are collected close together in time, such as daily observations, the data are correlated and the t-test is invalid. However, this serial correlation has advantages; it permits forecasting into the future. In meteorology, the investigator is usually interested in inferring physical principles or forecasting.
