Global modeling of aggregated and associated chaotic dynamics

Spatially distributed systems are rather difficult to investigate due to two distinct problems which can be sometimes combined. First, the spatial extension is taken into account by monitoring the system evolution at different locations. Second, the dynamics cannot always be continuously tracked in time, and segments of data – sometimes recorded at different places – are only available. When the dynamics underlying a single marker is under consideration – as for instance the normalized difference vegetation index which can be used for assessing the vegetation canopy of a given area – a global model can be obtained from a single scalar time series built by aggregating the available time series recorded at different places and/or associating the segments of data recorded at different times (and possibly at different locations). We investigated how these two data preprocessing – common in environmental studies – may affect the model dynamics by using a system of spatially distributed Rössler systems which are phase synchronized or not.