A%20case%20study%20of%20tropical%20extratropical%20interaction

1
A case study of tropical extratropical
interaction
Yucheng Song and Zoltan Toth Environmental
Modeling Center National Centers of Environmental
Prediction National Weather Service NOAA
2
Outline

• Ensemble forecasts from NCEP and ECMWF showing
  mean and spread of mean sea level pressure
• The ECMWF ensemble forecast is initialized 12
  hour earlier than NCEP
• Global IR image animation showing how the
  tropical convection interact with extra-tropical
  system
• Difference in NCEP and ECMWF analyses for the
  case
• Expected forecast error reduction in the
  verification region by NCEP and ECMWF ensembles
  using ETKF method
• Precipitation verification by observation and
  probability forecast from NCEP global ensemble
  forecast
• Animation of expected forecast error reduction by
  NCEP ensemble
• Summary and discussion

3
On Oct 14, 12Z two low pressure systems developed
4
Forecasts made starting with different initial
dates
NCEP Ensemble initialized on Oct 10, 00z was able
to predict the low pressure system near Gulf of
Alaska on a 4.5 day lead time
5
A map taken from the previous slide shows the
forecasts on a 4.5 day lead for NCEP forecast, 5
day lead for ECMWF forecast
6
Forecast loop initial time Oct 10 00Z
The animation on the left shows the forecast
initialized on Oct 10, 00Z for NCEP, Oct 9, 12Z
for ECMWF The NCEP ensemble spread shows tropical
influence to the Gulf of Alaska low
7
MSLP animation (12Z forecast added to mimic
analysis)
The animation on the left mimics how the two low
pressure systems in North Pacific formed
8
Global IR animation an example of tropical
extratropical interaction
9
Comparison of initial conditions NCEP vs ECMWF
- Mean Sea Level Pressure on Oct 10 00z, 2005
Differences in NCEP and ECMWF analyses can be
spotted at initial time
10
Difference of Mean Sea Level Pressure on Oct 11
00z, 2005 between NCEP vs ECMWF analysis
Differences in NCEP and ECMWF analyses can be
spotted at 24 hours later the region
corresponds well with expected forecast error
reduction in verification region spotted by ETKF
using NCEP ensemble well denoted by square boxes
in the two figures
11
Expected forecast error reduction in verification
region by ETKF method using 19-member NCEP
ensemble
Expected forecast error reduction in verification
region by ETKF method using 25-member ECMWF
ensemble
12
Precipitation verification heavy rainfall was
produced by the first system near Gulf of Alaska
13
Global PQPF animation - NCEP Global Ensemble
Animation shows how the first system near Gulf of
Alaska formed by NCEP global ensemble forecast
14
The blob shows the center of forecast error
reduction evolution with forecast lead time
15
Summary

• 
  
• Influence of organized tropical convection on
  extratropical circulation
• Tropical system picked up by westerly jet,
  wave evolves into strong low in Gulf of Alaska
• Strong downstream propagation of energy leads
  to explosive development of second low in front
  of original wave
• 
  
• Ensemble forecast performance
• NCEP ensemble captures
• First wave from Oct 7 12Z on
• Second wave from Oct 8 12Z on
• NCEP ensemble mean gives accurate forecast
  from Oct 10 00Z
• ECMWF ensemble captures developments at
  shorter lead times
• 
  
• 
  

16
Summary (continued)

• ETKF targeting method with ensemble from
• NCEP - identified area of tropical convection
  as main source of initial value uncertainty
• ECMWF - no clear signal
• What limits predictability for the first and
  second waves to 7 6 days? Errors in
• Initial conditions
• Observations
• Quality
• Quantity
• Model representation of tropical convective
  processes
• Other model problems?