Bayesian trend analysis of extreme wind using observed and hindcast series off Catalan coast, NW Mediterranean Sea

Abstract

Abstract. It has been suggested that climate change might modify the occurrence rate and magnitude of large ocean-wave and wind storms. The hypothesised reason is the increase of available energy in the atmosphere-ocean system. Forecasting models are commonly used to assess these effects, given that good quality data series are often too short. However, forecasting systems are often tuned to reproduce the average behavior, and there are concerns on their relevance for extremal regimes. We present a methodology of simultaneous analysis of observed and hindcasted data with the aim of extracting potential time drifts as well as systematic regime discrepancies between the two data sources. The method is based on the Peak-Over-Threshold (POT) approach and the Generalized Pareto Distribution (GPD) within a Bayesian estimation framework. In this context, storm events are considered points in time, and modelled as a Poisson process. Storm magnitude over a reference threshold is modelled with a GPD, a flexible model that captures the tail behaviour of the magnitude distribution. All model parameters, i.e. shape and location of the magnitude GPD and the Poisson occurrence rate, are affected by a trend in time. Moreover, a systematic difference between parameters of hindcasted and observed series is considered. Finally, the posterior joint distribution of all these trend parameters is studied using a conventional Gibbs sampler. This method is applied to compare hindcast and observed series of 10 min average wind speed at a deep buoy location off the Catalan coast (NE Spain, Western Mediterranean; buoy data from 2001; REMO wind hindcasting from 1958 on). Appropriate scale and domain of attraction are discussed, and the reliability of trends in time are addressed.

Should we Sample a time Series more Frequently?: Decision Support via Multirate Spectrum Estimation

“…Related work by Fokianos and Savvides () shows that the log‐ratio of the periodogram ordinates depends asymptotically on the log‐ratio of their respective spectral densities. Motivated by this fact, they proposed modelling the log‐ratio of two or more unknown spectral density functions for comparison and clustering; see also Savvides et al .…”

“…Obviously, model (8) in the paper belongs to this framework. In fact, any auto‐regressive moving average process can be approximated by expression (42) (see Fokianos and Savvides (), page 319). Some further work on model (42) can be found in Parzen (), Holan () and Holan et al .…”

“…Bayesian hindcasting methods have been developed in the fields of climatology and meteorology; see, for example, Solomon (), page 689, Werner and Holbrook (), Ortego et al . () and references therein. However, hindcasting using multiple sample rate assimilation via a multirate spectral approach as developed here appears to be new and fully coherent in the Bayesian sense.…”

Monitoring, cataloguing and weather scenarios of thunderstorm outflows in the Northern Mediterranean

“…Besides, the analysis of wind data has an ever more extensive field of application in the interpretation of the wind phenomena that occurred in the past as well as in forecasting the future ones, in climatological analyses and in climate change surveys (Ortego et al, 2014;Dawkins et al, 2016), in the study of wind hazards, damage and risks to which buildings, infrastructures and human activities in contact with Earth's atmosphere are exposed, until pollutant dispersion and wind energy production.…”