Title: Modelling volatility in daily air temperature on Svalbard
Authors: Sondre Holleland - University of Bergen (Norway) [presenting]
Hans Arnfinn Karlsen - University of Bergen (Norway)
Abstract: A lot of ink has been devoted to showing positive trends in air temperature over the last decades due to global warming and climate change. The effects are especially clear in the Arctic where the reduction in sea ice is a big issue. By considering daily average air temperature measurements from Svalbard airport dating back to 1975, we develop a model for the day-to-day conditional volatility. The model captures seasonal effects and a significant systematic decrease in the logarithmic volatility over the last four decades. A two-step iterative scheme is used between a mean model and the volatility model until convergence of the parameter estimates. The volatility model is related to exponential GARCH with seasonal time varying parameters and a linear trend in the logarithmic conditional volatility. We use the Template Model Builder (TMB) package in R to fit the model by maximum likelihood.