Status:Ongoing analysis Material from: Linus, Esti, Fernando, Michael MG, ..
Show If | ||||||
---|---|---|---|---|---|---|
| ||||||
Discussed in the following Daily reports:
|
...
2. Description of the event
Copied from e-mail on tropical cyclone mailing list from Sebastien Langlade:
""
Here a few items from TC belal over La Reunion:
> Tropical cyclone BELAL made a direct hit on Reunion Island on January 15, 2024. Intensity estimates at time of landfall had been slightly reduced from real time estimates: 75 kts 10-min winds with estimated MSLP a little above 970 hPa likely based on pressure readings from AWS within the eye.
> After approaching the island from the northwest during the night of the 14th to the 15th, the eye of the cyclone finally rounded the island in the morning from the north, before skirting the northeast coast. By mid-day, it had entered land off the commune of Saint-Benoît (east coast). It emerged shortly afterwards, in the early afternoon, off the eastern slopes of the volcano. Weakened by its passage over the island, BELAL then moved very slowly for a few hours in the immediate vicinity of the southeast coast, before beginning to move away to the east-southeast in the late afternoon. All of this track deflection seems to be caused by the high mountains of La Reunion (3070 m and 2631m for the 2 volcanoes).
> The eyewall of BELAL, which concentrates the most degraded conditions associated with the cyclone, is strongly felt in the northern and eastern regions, and to a lesser extent in the southeast of the island. BELAL also featured a highly asymmetrical distribution of rain and wind within its circulation (due to some westerly shear), so that the front part of the meteor was frankly rainier and windier than the rear part.
> Gusts exceeding 150 km/h was recorded over wide portions of the island with a peak gust of 170 km/h recorded at the airport (north coast) during the eyewall. Gusts over 200 km/h were recorded at a height of more than 2000m (Maïdo), but may have also been felt over some exposed portions of the north and north-eastern island.
> Although total rainfall were significant (see AF), this is near of what can be expected for a "classical TC" over La Reunion. Rainfall amounts over quite short time period (from 1 to 12 hours) were really significant over the northwestern part and the inner part of the island.
> It has been more than 30 years since the eye of a tropical cyclone passed over Reunion Island (on January 19, 1993, with the passage of tropical cyclone COLINA).
> Forecast: track quite good (direct hit suggested 48 hours before) - intensity: significant overestimation ... if we expected the mid-shear to increase just before the landfall, i don't understand why this TC did not strenghen that much 24 to 36 hours before the landfall as all factors seems to be aligned ... (i would be happy to discuss this further if some of you may have a look ... currently i just understand that the shear started earlier than suggested by the guidance ... )
""
Precipitation totals and maximum wind gusts from MeteoFrance (from the same e-mail as above).
3. Predictability
3.1 Data assimilation
We were lucky to see two direct hits of ASCAT winds over cyclone BELAL before and after landfall in La Reunion.The acquisition times are very close to 18Z time (less than 30 min between the analysis and the data acquisition time) so I used the SCDA ODB to generate the plots displayed below. The first snapshot shows the cyclone before the landfall (La Reunion is to the SE of the storm). Overall the the wind arrows are consistent with the circulation suggested by the pressure contours in the analysis except a couple of cases for very intense winds (red arrows) close to the centre of the cyclone. The position errors in the analysis are a bit smaller than the first guess but the intensity pressure error is larger in the analysis. The analysis and first guess departures as a function of the distance from the storm centre is also presented below. Overall the mean bias and StDev is improved in the analysis. The second snapshot show the cyclone BELAL moving away from La Reunion. In this situations the representation of the ascat winds around the storm is poorer compared with the previous plot in particular near the centre of the storm (keep in mind we have a 30 min gap between the analysis and the acquisition data time). When we look at the ACTIVE data (used) things are not so bad except for one or two cases. Also note the wobbly effect on the mslp for a +6h forecast (right panel).
ASCAT- winds (arrows) and analysis/first-guess mslp (contour) and wind speed (shade) from before the passage of La Reunion. Analysis & first-guess wind speed departures as a function of distance from the storm centre.
ASCAT- winds and analysis/first-guess mslp and wind speed from after the passage of La Reunion based on all data.
ASCAT- winds and analysis/first-guess mslp and wind speed from after the passage of La Reunion based on used data.
3.2 HRES
The plots below show 1-day precipitation (14 January 12UTC - 15 January 12UTC) in observations (first plot), concatenated short forecasts (second plot) and ENS control forecasts with different lead times (all 00UTC runs). The box marks a 0.25x0.25 degree box centred on St Denis (20.9S, 55.5E).
...
The plots below show the tropical cyclone track for TC Belal for the operational ECMWF forecasts from 15 January 00UTC (first plot) to 8 January 00UTC (last plot). The symbols shows the position on 15 January 00UTC (hourglass for BestTrack). HRES (red), ENS CF (blue), ENS PF (grey) and BestTrack (black).
3.4 Monthly forecasts
3.5 Comparison with other centres
...
FuXi
ML centres comparison
TC track in ML models
Intensity evolution in ML models
4. Experience from general performance/other cases
5. Good and bad aspects of the forecasts for the event
- Early signal for genesis
- Left-hand track error until 13 January.
- Miss for extreme rainfall over Mauritius related to rainbands?