Parametric Time Series Analysis of Cold and Hot Spells in Daily Temperature: An Application in Southern Italy

Abstract

The statistical analysis of cold air temperatures (cold spells) and hot air temperatures (hot spells) is discussed. Air temperature time series observed at 50 stations in southern Italy am investigated. The deterministic and stochastic components of the time series are identified and described by a dynamic?stochastic model that is periodic in the deterministic part (the annual cycle) and Markovian (first-order autoregressive) in the stochastic part. The annual cycle is described by only a few Fourier coefficients. Based on the model fitted to the data, the theoretical probability of cold (hot) spells is computed and compared to that estimated from the observed data. Spatial patterns are identified that make it possible to extrapolate the probability of cold (hot) spells at locations where no direct observations are available.