Fast-moving Xynthia starts to deepen

Around noon on the 26th, when passing south of the Azores, the storm was still showing much convective activity, with a prominent cloud head as well as an impressive dry intrusion. Although at this stage ASCAT only observed the western flank of the storm, it measured strong gale winds (Beaufort 9) up to 45 kt and, in a larger area, over 40 kt, thus still inside the reliable range for ASCAT winds (Figure 2.3.1). To the south a clear wind convergence line was found in the ASCAT wind product (depicted in yellow in Figure 2.3.1). It connects with a line of convection in WV6.2 (depicted as an instability line, also in yellow), but unfortunately the ASCAT winds do not reach this area. This line matched the sharp air mass gradient area that existed for the past 24 hours around the Tropic of Cancer (see Figure 2.1.4).

Figure 2.3.1 - ASCAT wind (colour wind barbs) between 12:29 and 12:35 UTC on 26/02/2010, Meteosat 9 WV6.2 at 12:30 UTC on 26/02/2010 and corresponding subjective surface analysis interpretation (in yellow).

In the ECMWF model, the wind speed threshold of 45 kt was clearly reached by the analysis and the 12 hours forecast, but not by the 24 and 36 hours forecasts where maximum wind speed were in the 35-40 kt range (Figure 2.3.2). Regarding the area of higher wind speeds, naturally the analysis field fits the ASCAT observations better (check how grey shades cover red, brown or magenta wind barbs). Minor discrepancies in the border of the 30 kt threshold area may be attributed to the 30 minutes time difference between ASCAT pass and analysis instants, as the depression has moved northeastward very rapidly, at a mean speed of around 35 kt during the previous 6 hours and 80 kt (!) in the following 6 hours, according to ECMWF analysis. This would amount to a difference of 1/3 to 2/3 of a degree of longitude in just half an hour!

Regarding forecasts, both the 12 and 24 hours forecast missed the intensity of wind speeds in Xynthia southwestern quadrant. The 36 hours forecast shifted the highest wind speeds almost 2 degrees of longitude to the west, well above the expected shift due to the non-coincident ASCAT pass time.

Figure 2.3.2 - ASCAT wind at 12.5 km sampling interval (colour wind barbs) on 26/02/2010, between 12:29 and 12:35 UTC, and ECMWF (0.25 spatial resolution) mean sea level pressure (black lines in hPa) and wind speed above 30 kt (grey shade and blue lines every 5 kt) on 26/02/2010 at 12:00 UTC: analysis (top left), 12h forecast (top right), 24h forecast (bottom left) and 36h forecast (bottom right).

Regarding the pressure field, the ECMWF MSLP analysis field on 26/02/2010 at 12:00 UTC showed a large low-pressure system, with two sub-lows but with the absolute minimum value (997 hPa) in the northeastern sub-low, around 29.5N 30W (Figure 2.3.2). This pattern was also suggested by the 24 hour forecast, but not by the 12 and 36 hours forecasts, which identified a main low-pressure system on the west side. At 12:00 UTC a drift buoy located close to the southwestern sub-low, at 28.5N 32.5W, reported 994.4 hPa, around 5 hPa deeper than ECMWF analysis for that location.

ECMWF analysis suggested a drop of 7 hPa in 6 hours, between 12:00 and 18:00 UTC. A constant pressure drop in this period would result in a decrease of ~0.6 hPa every 30 minutes. As the ASCAT surveyed the area slightly after 12:30 UTC, it could be acceptable to apply a pressure drop of 1 hPa to the ECMWF analysis and buoy values mentioned above. This would result in a MSLP of 996 and 993.4 hPa for the analysis and buoy, respectively, at the time of the ASCAT pass.

The UWPBL model output (using ASCAT wind as input) clearly depicts the curved isobars on the west side of the storm, most likely the southwestern sub-low (Figure 2.3.3). The minimum sea level pressure along the pass was 995.4 hPa at 28N 32W, therefore lying between the "time-corrected" values derived from the ECMWF analysis and buoy observation. Note, however, that neither ASCAT nor buoy caught Xynthia's centre.

Figure 2.3.3 - Surface pressure (hPa) from the ASCAT pass on 26/02/2010, between 12:31 and 12:34 UTC, obtained from the University of Washington Planetary Boundary Layer (UWPBL) model (Source: and mean sea level pressure (hPa), in dashed magenta lines, from the closest ECMWF analysis on 26/02/2010 at 12:00 UTC (0.125 spatial resolution).

At this time Xynthia still had its vertical trough axis slightly leaning west, although at higher levels the ECMWF analysis showed only an emerging wave embedded in the western flow. As mentioned, the frontal nature became clearer from model fields but was still better depicted in lower levels than in mid and higher levels (Figure 2.3.4). Xynthia continued its northeasterly path as a deep low with strong winds at all levels, always embedded in a very humid and warm air mass, with regions of intense warm advection (Figure 2.3.5). Based on subjective analysis, it is safe to say that Xynthia still remained a complex low for the following 24 hours, until it reached the Portuguese coast.

Figure 2.3.4 - Thermal Front Parameter (TFP) in the layer 1000-700 hPa (10-11 K/m2) and mean sea level pressure (hPa) analysis from the ECMWF model on 26/02/2010 at 12:00 UTC.

Figure 2.3.5 - Temperature Advection at 700 hPa (C/h) analysis from the ECMWF model on 26/02/2010 at 12:00 UTC.