Community Research and Development Information Service - CORDIS


A2C2 Report Summary

Project ID: 338965
Funded under: FP7-IDEAS-ERC
Country: France

Mid-Term Report Summary - A2C2 (Atmospheric flow Analogues and Climate Change)

A2C2 investigates rare and extreme events of a dynamical system such as the climate. We focus on chaotic dynamical systems, which yield a strange attractor, and we aim at detecting how an alteration of the system leads to new, rare or extreme events. The general tools that we develop are good measures of pattern recurrence and assess when patterns have never been seen before. The tools are based on analogues of atmospheric circulation, which have been known for decades in the meteorological literature for weather forecast. We had demonstrated their potential in climate applications, especially for diagnostics of extreme events.
The A2C2 project has three workpackages (WPs). WP1 made a (previously unseen) link between statistics and the theory of dynamical systems, by relating extreme events to rare occurrences. This relation enabled us to devise tools to investigate the predictability of extreme climate events of spatial fields (rather than looking at time series). WP1 also aims at measuring how strange attractors can be altered. We found a link with optimal transport theory to quantify such alterations. WP2 was meant to devise efficient computer tools and web based applications so that the methods that are developed in WP1 can be easily used. We distribute the codes on the A2C2 web site (through a CECILL free licence). A continuous time analysis of atmospheric variability is also proposed on the A2C2 web page. We designed a web processing service with a visualisation tool to compute analogues on any atmospheric database. WP3 focused on the applications of analogues of circulation perform extreme event attribution. We have devised a methodology to compute dynamical and thermodynamical contributions to extreme precipitation events from large ensembles of simulations. We have also shown that such contributions can be determined from rather small datasets, at a fraction of the cost of the large ensembles.

Reported by

Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top