Spatio-temporal modelling of lightning climatologies for complex terrain

Abstract

Abstract. This study develops methods for estimating lightning climatologies on the day −1 km −2 scale for regions with complex terrain and applies them to summertime observations (2010)(2011)(2012)(2013)(2014)(2015) of the lightning location system ALDIS in the Austrian state of Carinthia in the Eastern Alps.Generalized additive models (GAMs) are used to model both the probability of occurrence and the intensity of lightning. Additive effects are set up for altitude, day of the year (season) and geographical location (longitude/latitude). The performance of the models is verified by 6-fold crossvalidation.The altitude effect of the occurrence model suggests higher probabilities of lightning for locations on higher elevations. The seasonal effect peaks in mid-July. The spatial effect models several local features, but there is a pronounced minimum in the north-west and a clear maximum in the eastern part of Carinthia. The estimated effects of the intensity model reveal similar features, though they are not equal. The main difference is that the spatial effect varies more strongly than the analogous effect of the occurrence model.A major asset of the introduced method is that the resulting climatological information varies smoothly over space, time and altitude. Thus, the climatology is capable of serving as a useful tool in quantitative applications, i.e. risk assessment and weather prediction.

NWP-based lightning prediction using flexible count data regression

“…A similar synthesis of methods has been applied previously in Simon et al (2018) to predict the occurrence probability of thunderstorms, i.e. using only a single additive predictor.…”

“…the algorithm in Hofner et al, 2015). The number of terms being selected before stopping the boosting algorithm and the cut-off selection frequency are chosen in order to establish an upper bound of unity for falsely selected terms (Hofner et al, 2015;Simon et al, 2018).…”

“…Thus, the predictions resulting from the binary hurdle part can be compared directly to the outcome of previous studies that focus on binary response variables (e.g. Schmeits et al, 2008;Simon et al, 2018). The combined model predicts a full probability distribution of lightning counts, which allows one to derive various quantities such as probabilities for the occurrence of thunderstorms and also quantiles and the exceedance of predefined lightning count thresholds (Fig.…”

“…In order to predict the probability of lightning events (i.e. thunderstorms), numerical weather prediction (NWP) output is often postprocessed by logistic regression (Schmeits et al, 2008;Gijben et al, 2017;Bates et al, 2018;Simon et al, 2018) in which lightning detection data serve as a proxy for the occurrence of thunderstorms. However, these studies present methods to predict only whether a thunderstorm might take place or not.…”

“…This concept has been used successfully in several studies (e.g. Simon et al, 2018;Thomas et al, 2018).…”