productos

Model validation is esential to evaluate the reliability of simulated outputs and to get some knowledge that allows to improve the model set-up and its performance over different areas.

In order to configure the most accurate model set-up, and to validate the resulting ESEOAT application, a preliminary set of 1-year hindcast reanalyses were carried out. The "Prestige" event year was the one selected to be hindcasted, going the hindcast run from September 2002 to Augost 2003. The outputs from this hindcast run were validated by means of comparisons with different satellite and in-situ observations, providing us a measure of the skill shown by the model to simulate reliable 3-D ocean properties (e.g. currents, temperature, sea-level, etc.)


Figure 1. SST Monthly means corresponding to January-2003 derived from satellite observations (a) and from ESEOAT simulated outputs (b). (c) Differences between the above monthly mean fields (ESEOAT-observation). The satellite monthly means were provided by AZTI

Once verified the reliability of the ESEOAT-simulated spatial patterns, it was locally verified the ESEOAT performance. To this aim, observed SST time series, measured at buoys from the Puertos del Estado deep-water Buoy Network, were compared to the SST timeseries simulated with the ESEOAT application at the closest gridpoint of each of the 7 moore location. Figure 2 shows the time evolution of hourly modeled and observed SST along the period between October-2002 and July-2003 at the Silleiro and Cadiz buoys.

Figura 2. Time series of hourly SST obtained from in-situ observations (blue line) and from the ESEOAT application (red line). Left Panel illustrates SST evolution at the Puertos del Estado Silleiro Buoy (42º7.2'N,9º24.0W). The right panel shows the results corresponding to the buoy located at Cadiz (36º28.8'N, 6º57.6'W). In both cases, data is hourly sampled and the time period shown goes along ten months from October-2002 to July-2003.

Validation of the ESEOAT-simulated sea-level

The realism of the ESEOAT-simulated sea level was verified by means of comparing such simulated data with in-situ measurements from 7 stations of the Puertos del Estado Tide-gauge Network (REDMAR) localized at different Spanish harbours. Figure 3a shows that the ESEOAT System is able to realistically reproduce, without significant shifts the high and low waters, as well as spring and neap tides at a coastal location of the northern Iberian coast (Bilbao Harbour). Likewise, Figure 3b, shows an example of how the simulated meteorological residual realistically reproduce the ones obtained from tide-gauge observations,confirming the right inclusion in the ESEOAT system of wind and inverse barometer effects.

Figure 3. (a) Astronomical tide (in m) at Bilbao harbour (blue line) and elevations simulated at the ESEOAT closest grid point for November 2002. (b) Meteorological residual (in m) at La Coruña Harbour . Blue line represents the residual derived from measurement of the Puertos del Estado tidegauge, whereas the red one represents the ESEOAT simulated residual. The pure inverse barometer effect is marked by crosses.

Validation of the ESEOAT-simulated surface currents

Validation of the simulated surface currents was carried out by means of comparing the ESEOAT-simulated ones with current-meter observations obtained from the Puertos del Estado Deep-water Buoy Network. Figura 4 shows two examples of this local comparison exercise.


Figure 4. Time Series of surface current speed (in cm/s), obtained from in-situ observations (blue line) and from the ESEOAT run (red line). Left panel illustrates time evolution of surface currents at the Puertos del Estado buoy of Peñas and the right panel shows the results at the Silleiro Buoy. In both cases data is hourly sampled and the time period shown is of one month.