Status:Finalized Material from: Linus
Discussed in the following Daily reports:
http://intra.ecmwf.int/daily/d/dreport/2015/03/09/sc/
http://intra.ecmwf.int/daily/d/dreport/2015/03/10/sc/
http://intra.ecmwf.int/daily/d/dreport/2015/03/11/sc/
http://intra.ecmwf.int/daily/d/dreport/2015/03/12/sc/
http://intra.ecmwf.int/daily/d/dreport/2015/03/13/sc/
http://intra.ecmwf.int/daily/d/dreport/2015/03/16/sc/
1. Impact
The tropical cyclone Pam hit the Vanuatu Islands on 13 March 2015 with a devastating effect. The cyclone were the second strongest in the southern Pacific after Zoe (2002) and is believed to be the worst natural disaster on Vanuatu in history. The case is discussed in a Newsletter article http://www.ecmwf.int/sites/default/files/elibrary/2015/14588-newsletter-no144-summer-2015.pdf
The figure below shows the wind footprint of the cyclone (UCL/TSR wind field modelling using JTWC real-time track provided by Mark Saunders/UCL via Tropical-Cyclone mailing list). As the cyclone moved parallel to the band of Islands, a large part of the country was affected.
More sources:
http://en.wikipedia.org/wiki/Cyclone_Pam
http://www.bbc.co.uk/news/world-asia-31895231
2. Description of the event
The cyclone formed March 6 east of the Solomon Islands and was classified as a tropical storm on 9 March. The formation is believed to been caused by a strong MJO with a westerly wind burst in the western tropical Pacific. At the same time 3 other cyclones formed (Olwyn west of Australia, Nathan north-east of Australia and Bavi north of the equator as seen on a satellite image from Meteo-France below).
We are currently not receiving minimum pressure estimates from Best Track. But the plot below show different estimates on minimum pressure and maximum winds presented at http://tropic.ssec.wisc.edu (including Best Track). The minimum pressure from Best Track was 915 hPa and winds of 145 kt.
An animation of the MTSAT-2 IR images during the passage over Vanuatu can be found here:
http://cimss.ssec.wisc.edu/goes/blog/wp-content/uploads/2015/03/150312_mtsat_ir_Pam_anim.gif
3. Predictability
3.1 Data assimilation
The plots below show ASCAT diagnostics (analysis and first guess vs. observations) for three assimilation cycles during the life-time of Pam. In the first example (from 9 March 12z), there are a cluster of vectors that got the wrong direction assign due to the position error in the first guess. The second example is an example (11 March 00z) of a good first guess which led to good usage of the vectors. In the third example (12 March 00z) many vectors were rejected due to the strong wind speeds.
3.2 HRES
The plot below shows all HRES forecast valid 13 March 00z. All forecasts have a cyclone in the area although the first four forecasts (initialised 3-6 March 00z) had a weak cyclone. The forecasts from 7-10 March had the cyclone somewhat too the east of the observed position.
3.3 ENS
The plots below shows the operational TC plumes (ENS+HRES) for Pam. The first forecast is from 9 March 00z. In all the forecast HRES had a minimum pressure below 920 hPa on 13 March (except the last forecast that may suffered from too close to the initialisation). ENS had at least a minimum pressure 20 hPa higher.
TC Metgrams from Pam (Operational)
The next plot shows the tropical storm strike probability for 12-14 March from the forecast 8 March (unfortunately I forgot to save the forecasts from earlier dates).
3.4 Monthly forecasts
The plots below show the strike probability of tropical storms for the week 9-15 March.
Strike probability for tropical storms
The next set of plots are normalised tropical cyclone energy for the 9-15 March. The forecast from 26 February and onwards had more than a doubling of the energy compared to climatology.
Normailsed TC energy
The plots below show the MJO forecasts. The developments of the tropical cyclones during the week 9-15 March was connected to a very strong MJO event. The development of the MJO was in the forecast since 26 February.
MJO forecasts
3.5 Comparison with other centres
4. Experience from general performance/other cases
5. Good and bad aspects of the forecasts for the event
- Good intensity from HRES
- Position error present until forecast from 10 March for HRES but covered by the ensemble plume
- Increased tropical cyclone activity in Monthly forecast from 26 February and onwards, probably due to extreme MJO
- Good test case for ASCAT observations