“Many measurement results are so-called time series, especially at the time when measurements are conducted. The term ‘time series’ refers to measurement data where the desired variable, such as temperature, particle concentration or snow depth, is recorded together with the time of the measurement,” the researchers say.
"In practice, this means that the variable can be drawn in a graph as a function of time, allowing an analysis of whether any changes are observed during the measurement period. Combined with other measurement results, one can try to determine what causes the observed changes.”
“Time series can be of any length. Time series of less than one minute are used when studying very rapid chemical or physical reactions, and time series spanning several years are used when studying particle concentrations in Kuopio. Time series of the Earth’s average temperature can span hundreds of thousands of years, and they are comprised of several different time series, most of which were produced by means other than direct measurements.”
“In many disciplines, the benefits of long time series come precisely from the fact that, once the same variable has been continuously measured for several years, it is possible to observe slow changes that would otherwise drown in short-term normal variations.”
“Once again, temperature is a good example: temperature varies greatly during the year, and, for example, Christmas Eve temperatures may vary greatly between years. If measurements are conducted over decades, it is possible to analyse whether the average Christmas Eve temperature has fallen, risen or remained almost the same.”
“Long time series are also useful benchmarks for modelling experiments, which test how well our current understanding of the processes behind the variable corresponds to reality, i.e., how well modelling can reproduce the real measurement results.”