Characterisation of extreme events waves in marine ecosystems: the case of Mediterranean Sea

Abstract

<p><strong>Abstract.</strong> We propose a new method to identify and characterise the occurrence of prolonged extreme events in marine ecosystems on the basin scale. There is a growing interest about events that can affect ecosystem functions and services in a changing climate. Our method identifies extreme events as peak occurrences over 99th percentile thresholds computed from local time series and defines an Extreme Events Wave (EEW) as a connected region including these events. The EEWs are characterised by a set of novel indexes, referred to initiation, extent, duration and strength. The indexes, associated to the areas covered by each EEW, are then statistically analysed to highlight the main features of the EEWs on the considered domain. We applied the method to the winter-spring daily chlorophyll field of a validated multidecadal hindcast provided by a coupled hydrodynamic-biogeochemical model of the Mediterranean open-sea ecosystem, with 1/12° horizontal resolution. This allowed to identify the maxima of chlorophyll as exceptionally high and prolonged <q>blooms</q> and to characterise their phenomenology in the period 1994–2012. A fuzzy k-means cluster analysis on the EEWs indexes provided a bio-regionalisation of the Mediterranean Sea associated to the occurrence of chlorophyll EEWs with different regimes.</p>