Dynamics of Sea-Level Pressure and Wind Velocity: A Study of Volatility at Airports Using Statistical Time-Series Analysis

The constant increase in the number of flights across the world means that aircraft safety today has become an even greater priority. Sudden changes in weather within the earth’s boundary layer, for example wind shear create many risks for take-offs and landings. The key wind variables here are sea-level pressure (SLP), wind speed (ws) and wind direction (wd), together with humidity and air temperature: it is the change in the values of these variables that can cause wind shear and subsequent turbulence. The literature on wind shear deals with the values of these variables: these values are highly correlated and can introduce the so-called level effect. Usually, the logarithm of the ratio of two successive values of a variable, also called the return, is used in time-series analysis. Returns behave like a random variable which is ideal for statistical analysis. The use of changes or returns of ws, and wd, and SLP, for investigating the onset and duration of wind shear is explored in this paper. This use of returns of a variable, rather than the variable itself, is a methodological point and distinguishes our work from that of others. Returns behave like a random variable but in rapidly changing situations they appear to cluster, leading to system-wide instability or volatility. We report on the statistical properties of the returns for wind speed and direction, and sea-level pressure, at three airports using a time series over a seven-year period (2016–2023). As it is the clustering that signifies volatility in weather patterns, we investigate two methods of estimating the clustering: first, model-free estimates of volatility based on squared returns, and second, model-based estimates, like GARCH, that use returns.