Fluctuation analysis of meteo-marine data

We study long-term behaviour of air temperature, wave heights and wind speed time series recorded for the period 1993-1997 at a meteo-marine station located in the Adriatic Sea. The scaling analysis shows that fluctuations of air temperature display long-range autocorrelations, while those for wave heights show a more complex behaviour, crossing over from a persistent regime at intermediate time scales (up to about 20 days) to an anti-persistence behaviour at longer times. Furthermore, the crosscorrelations of their records are found to be large, with a covariance of about -0.3 (indicating anti-crosscorrelations) within the full 5-years period, giving a quantitative measure of the actual coupling between the two data sets. Windspeed fluctuations are found to be strongly crosscorrelated (about 0.6) with those of wave heights, indicating as expected that wind is the main driving force for waveheight fluctuations.