The example is challenging, because selleckchem of the high space-time variability of currents caused by the dominance of tidal currents. We first describe the system, and then illustrate the performance. Then, we describe an application of such a “product” in the context of “search and rescue”. The system was developed in the framework of Coastal Observing SYstem for Northern and Arctic Seas (COSYNA) in recent years (Stanev et al., 2011). It uses radial current velocities from three high frequency (HF) radars. The employed assimilation method STOI (spatio-temporal optimal interpolation; Stanev et al., 2014) uses elements of assimilation

filters and smoother. The STOI method does not only interpolate, but also ‘extends’ in space the radar data, which makes possible to generate homogeneous mapped data Doramapimod in vivo series over areas larger than the observational array (Stanev et al., 2014). Surface currents are analyzed simultaneously using an analysis window of 13 or 24 h, thus continuous surface current trajectories over one or two M2 tidal cycles are obtained. In Fig. 3a, a snapshot of three different descriptions of a surface current field are displayed, namely HF radar observations (green), the result of the data assimilation using STOI (red) and a simulation with the same model, which is employed in STOI, but which is not constrained by the

HF radar observations (free run; blue). The data assimilation changes the description of the current in particular at near coastal grid points, e.g., in the Elbe estuary. Also, the region covered by the

analysis is larger than the area covered by HF radar observations. Fig. 3b shows radial velocities during a M2 tidal cycle for a point, as recorded from a HF radar station (black crosses), the analysis using STOI (green) and the free run mentioned above (blue). Note that the HF data are not available for the entire time – for a period of 4 h, no data have been recorded. Obviously, the data assimilated describe the observations very well, and are capable to “fill” the data gap consistently. An operational product based on this analysis system VAV2 may find an application in search and rescue operations. The utility is demonstrated by the large differences for the estimated transport trajectories, when unconstrained current simulations are used, compared to the trajectories derived from analyzed currents. In a transport model, many particles have been released in the center of every grid cell and were then moved with the surface currents derived from the STOI product. The mean travelled distances vary mostly between 2 and 4 km, but in some cases the distance amounts to 5 and more km. Fig. 4 shows 3-day trajectories emanating from six exemplary locations. The black one is run with unconstrained currents, the red one with constrained STOI currents. The wiggles in the trajectories represent the effect of tides.