Markovian characteristics of dry spells over the Iberian Peninsula under present and future conditions using ESCENA ensemble of regional climate models

Abstract

The study analyses the annual structure of dry spells over the Iberian Peninsula through a second markov chain model based on the persistence concept under present and future conditions. The persistence of dryness can be considered as an indicator of drought, being useful for early warning drought systems. It can be studied as an isolated factor, here we focus the analysis on the probability of dry spells occurrence at each grid point of the domain by means of a two-state second order of Markov Chain model. Previous studies using this model along twentieth century observational data obtained successful results in the dry spells characterisation over several regions. First, the application of this theoretical model on three observational datasets over the domain, confirms the second order of markov chain characteristics of the observed dry spells. Then, the observational climate (1989–2007) EraInterim-forced regional climate model simulations from Spanish ESCENA project are successfully compared with those observational datasets results. Then, the study of climate change projections on the markovian behaviour is studied through the ECHAM5r2 global climate model and the regional models under present (1970–2000) and future (2021–2050) climate conditions. They indicate that the markovian characteristics of dry spells are kept in the future over the domain except some points in the South East coast. A decrease of the probability of occurrence of short dry spells (1–7 days) is obtained, while the longest dry spells (>12 days) are increased. These results are coherent with the analysis of the annual average of dry spells length.