Projected wave climate temporal variability due to climate change

This work proposes a new general procedure to stochastically analyze multi-model multivariate wave climate time series projections at different temporal scales. For every projection, it characterizes significant height, peak period and mean direction by means of univariate non-stationary distributions capable capturing cyclic behavior over reference interval duration. The dependence between the values given sea state previous short-term is described with vector autoregressive model (VAR). ensemble characterization based on compound distribution individual weighted averaged VAR model. methodology applied bias-adjusted derived using WaveWatch III forced wind field data from EURO-CORDEX models location close Mediterranean Spanish coast. Results are compared hindcast which shows clear bi-seasonal behavior. Different references were considered, starting 1-year overall changes in scales ranging days, months seasons respect historic conditions. results show that projected has very than data, delaying widening/shortening start duration two main including shorter term variations. Regarding energetic content states, variable highest percentiles height present lower (≈3−10%) during traditionally more severe (November–March) but higher (≈10−35%) calmer months. presents similar pattern while change historical bi-modal towards likely easterly waves throughout year. Additionally, 10-year analysis done find larger variabilities such as decadal variations associated North Atlantic Oscillation. observed variability yearly seasonal century addressed analysing 20-year rolling windows all variable. most periods summer end century.