Keywords: extreme temperature events, generalised Pareto distribution, GPD, climate change, peak–over–threshold, POT, Ireland, modelling, risk assessment, global warming, seasonal models, hotter summers, mild winters
Modelling extreme temperatures in Ireland under global warming using a hybrid peak–over–threshold and generalised Pareto distribution approach
In the present paper, modelling extreme temperature to assess risk of global warming in Ireland is addressed. The approach used is a combination of peak–over–threshold (POT) - generalised Pareto distribution (GPD) in which the parameter of distribution is allowed to vary with a dominant feature of climate at the location. The dominant climatic feature at a location is approximated by climatic variables derived from the National Centre for Environmental Prediction (NCEP) reanalysis data. Data from six stations were used to develop seasonal models for winter, spring, summer and autumn. Future changes in extreme temperature values are generated by using climate variables derived from (HadCM3) GCM for the A2 emissions scenario. The software extRemes was used to develop the models as it serves the proposed modelling approach. Results indicate that significant changes in extreme temperature events are projected to occur in Ireland over the course of the present century. These include hotter summers and mild winters.