Status:Ongoing analysis Material from: Linus, Mohamed


1. Impact

On 19 February heavy rainfall affected Brazil coastal areas near Sao Paulo causing floods and deadly landslides. More than 36 people lost their lives and a large number of people are homeless after the collapse of houses. This event followed days of rain that made the terrain more susceptible to landslides. Observed precipitation will be reported in the coming days but according to the media the 24h total exceeded 600 mm in few locations and the rain rates were intense. 

2. Description of the event

The plots below show analyses of MSLP and 6 hour rainfall from 17 February 00UTC to 20 February 00UTC, every 12th hour.

The plots below show analyses of z500 and T850 from 16 February to 20 February 00UTC, every 24 hour.

3. Predictability


3.1 Data assimilation


3.2 HRES

The plots below show observations (first plot), analysis from concatenated 6-hour forecasts (2nd plot) and HRES forecasts of 24-hour precipitation valid 19 February 00UTC - 20 February 00UTC, from different initial dates. The black box show a area around Sao Paulo. Note that no observation was available on GTS in the worst affected region.

The plots below show the same as above but for the e-suite.

3.3 ENS

The plots below show EFI and SOT for 1-day precipitation 19 February, from different initial times. 

The plots below show EFI for 1-day integrated water vapour flux 19 February, from different initial times. It is interesting to note that this event does not seems to be strongly forced by the water vapour flux.

The plot below shows the forecast evolution plot for 24-hour precipitation valid  19 February 00UTC - 20 February 00UTC for 0.5 degree box around outside Sao Paulo. Mean of observations - green hourglass, concatenated 6-hour forecasts - green dot, HRES –red, ENS blue box-and-whisker, Model climate – red box-and-whisker. Ensemble median as black box and ensemble mean as black diamonds. Triangle marks the maximum in the model climate based on 1200 forecasts. 48r1 e-suite is included in orange dot (HRES/ENS control) and purple (ENS distribution).

For the last forecast before the event the 48r1 ENS distribution is higher than o-suite, probably due to the higher resolution.

3.4 Monthly 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

6. Additional material