Local climatological modeling of ionospheric irregularities detected by GPS in the mid-latitude region

Global Positioning System, or GPS, plays an important role in everyday life. More particularly, precise positioning applications constitute a continuously growing sector whose surveyors, civil engineers and more recently farmers represent the principal users. Ionospheric irregularities are considered as the main threat for those applications as their occurrence and their effects on positioning are generally unknown or unmodeled. This paper aims at setting up a local climatological model of such irregularities which can be used as a forecasting tool. The model is based upon a time series of GPS-derived ionospheric irregularities in Belgium covering 10 years of data (period 2002-2011). Our climatological model is twofold: its first component describes the daily variability and is derived from a principal component analysis (PCA) which allows us to retrieve the main patterns of the time series. With the use of low order polynomial and harmonic functions, the second component describes the influence of solar cycle and seasons on irregularity occurrence. Moreover, a statistical autoregressive formulation adapts the model to current conditions. Model validation covers both low and active solar activity periods (years 2008 and 2011) and shows that model accuracy varies with solar conditions and season: values are lower during winter and active solar activity periods, where modeling error can reach up to 60% of the observed value. During summer, model performance is clearly improved, with relative errors generally smaller than 20% for periods of low but also active solar activity. (C) 2012 Elsevier Ltd. All rights reserved.