A photographic collage depicting the societal, economic and ecological impacts of severe weather ass

1
David Parsons NCAR/EOL/MMM Co-chair, North
American Regional Committee US Interagency Lead
A photographic collage depicting the societal,
economic and ecological impacts of severe weather
associated with four Rossby wave-trains that
encircled the globe during November 2002.
2
What is THORPEX?
THORPEX a Global Atmospheric Research Programme
is an international research programme to
accelerate improvements in the accuracy of 1 to
14-day high-impact weather forecasts for the
benefit of society and the economy. THORPEX
will make progress by enhancing international
collaboration between the research and
operational-forecasting communities and with
users of forecast products. THORPEX is
coordinated within the World Meteorological
Organization. Fifteen countries (Australia,
Canada, China, France, Germany, Japan, Iceland,
India, Korea, Norway, Russia, South Africa,
Spain, UK, US) and the European Commission are
leading the THORPEX effort. Participation
includes developing (the 54 countries of African
for example) and developed world. Thirty six
countries will be represented at the 1st Intl
Science Symposium. The THORPEX web site is
http//www.wmo.int/thorpex The THORPEX
implementation phase begins 1 January 2005.  
  International Science Plan Version 2 September
2003  
3
THORPEX -- Global Perspective

• THORPEX A Global Atmospheric Research Programme
  is a once in our lifetime opportunity focusing
  the talents of the worlds research and
  operational communities to better understand and
  predict the atmosphere with profound implications
  for society.
• Past opportunities and efforts include..
• The International Geophysical Year (1957-58) that
  followed the advent of the upper-air network in
  the post-war era, the recognized needs to develop
  satellite measurements and the desire to solve
  geophysical research challenges through
  international collaboration.
• The first GARP effort that concluded with FGGE
  (First Global GARP Experiment) in 1978-79 and was
  motivated by the first opportunity for
  quantitative global measurements.
• These efforts were also recognized by and closely
  coordinated within the World Meteorological
  Organization.

NSF THORPEX Briefing, D. Parsons, 11/21/03
4
Why another GARP-type effort?

• GARP was successful leaving a legacy of a global
  observing system and global modeling for climate
  and numerical weather prediction (NWP). One of
  the grand scientific and technical
  accomplishments of the 20th century.
• However.
• Progress in global NWP has been relatively steady
  and slow.
• Failures still occur in the prediction of high
  impact weather and in efforts to mitigate
  disastrous weather events (18 billion in US)
• US research focus is elsewhere (drift toward
  climate and small-scale studies)
• Strong economic benefits in developed world
  (nearly ¼ of US GNP is sensitive to weather often
  impacting profit margins)
• Observations for even the 1-day forecast extends
  well beyond our national borders (therefore a
  international effort is required)
• Fundamental changes in global observing systems
  (satellite revolution)

5
Evolution of forecast skill for the northern and
southern hemispheres 1980-2001. Anomaly
correlation coefficients of 3, 5, and 7-day ECMWF
500-mb height forecasts for the extratropical
northern and southern hemispheres, plotted in the
form of running means for the period of January
1980-August 2001. Shading shows differences in
scores between hemispheres at the forecast ranges
indicated (from Holingsworth, et al. 2002).  
6
Evolution of forecast skill for the northern and
southern hemispheres 1980-2001. Anomaly
correlation coefficients of 3, 5, and 7-day ECMWF
500-mb height forecasts for the extratropical
northern and southern hemispheres, plotted in the
form of running means for the period of January
1980-August 2001. Shading shows differences in
scores between hemispheres at the forecast ranges
indicated (from Holingsworth, et al. 2002).  
7
57 billion dollar weather-related disasters
between years 1980 and 2003. Seven occurred
during 1998 alone and the 48 during the 1988-2003
period totaled unadjusted damages/costs of nearly
215 billion. In a typical year, between 300 and
400 people in the US die each from hazardous
weather (peak years 1000-10,000)
8
Perceived Forecast Failures

• 2000 and 2002 regionally over-predicted east
  coast snow falls
• Fall 1999, two storms from cut-off lows produce
  snow in N Vermont and N NY
• 16 Sept 1999, Hurricane Floyd after landfall (gt1
  billion in damage and 16 deaths) in NE US
• 5-9 Jan 1998 Montreal ice storm, in the US
  freezing rain occurs in areas where only rain is
  predicted
• 17-18 Oct 1998, SE Texas floods, some errors in
  location and magnitude, 1 billion in damage and
  31 deaths
• 19 Jan. 1997 Florida freeze, 100,000 farm workers
  unemployed or displaced, 300 million in crop
  damage in one region alone

• 7-8 Feb. 2002, Pacific NW cyclone with damaging
  winds, extensive power outages in Oregon
• 13-14 Dec. 2001, west coast mountains experience
  heavy orographic snows
• Above examples from 10-15 significant intensity,
  timing and location errors per year of storms in
  the Pacific NW
• 10 June 2001 Houston impacts of TS Allison, 22
  deaths, 4-5 billion in damage, medical emergency
  (5 hospitals in flood plain)
• 17 June 2001, Flooding from remnants of TS
  Allison on NE US
• 24-26 Jan. 2000 surprise snow , storm Carolina
  to NE with all-time record Carolina snowfall

9
Observational Requirements for US Weather
10
Moores Law for Intel
Satellite remote-sensing revolution
11
Four Interrelated and Coordinated THORPEX
Sub-programmes

• Observing Systems
• Data Assimilation and Observing Strategies
• Predictability and Dynamical Processes
• Societal and Economic Impacts.
• Combine efforts to work on grand challenges
  (design of the next generation global observing
  network, adaptive regional modeling efforts,
  operational forecasts from true combining of
  multi-national modeling efforts)

OFAP THORPEX Briefing, D. Parsons, 4/19/04
12
University Involvement

• University of Arizona
• Arizona State University
• U California Davis
• U of California at Santa Barbara
• U of Colorado
• Colorado State U
• Cornell U
• Florida State U
• Georgia Tech
• U of Hawaii
• U Harvard
• U of Maryland

• U of Miami
• MIT
• Naval Postgraduate School
• U New Hampshire
• U North Carolina at Charlotte
• Penn State U
• State U NY at Albany
• State U NY at Stony Brook
• Texas AM
• U Utah
• U Washington
• U Wisconsin

NSF THORPEX Briefing, D. Parsons, 11/21/03
13
THORPEX Research Challenges(Predictability)

• Activity on the mesoscale often, if not
  typically, requires an amplified or perturbed
  planetary-scale wave pattern with high-impact
  weather at the leading edge of Rossby wave
  packets.
• Understanding and predicting the triggering of
  wave trains and their role in producing severe
  weather would seem to be essentially to improved
  medium range forecasts

NASA THORPEX Briefing, D. Parsons, 4/15/04
14
1st Example Central European Floods
Prague
August 2002
15
The Storm
France
Italy
Dundee Satellite Station 1241 UTC 11 Aug. 2002
16
Dresden Germany
17
Central European Floods
Cyclogenesis off Japan
Courtesy of Shapiro and Thorpe
18
1
19
2
1
20
3
2
1
2
1
21
2
3
4
2
1
1
3
22
2
1
3
4
23
1
2
5
3
4
24
6
7
5
4
25
6
5
4
7
26
6
5
4
7
27
6
5
7
28
5
8
6
7
Flood
29
2nd Example Minnesota Flood 9-11 June 2002
Moderate drought on 1 June Widespread rainfall
in excess of 5 inches. Flood with gt340
million in federal disaster aid. 80 of homes
and businesses damaged in Roseau, MN Locally
most significant flood on record.
30
Minnesota Flood 9-11 June 2002
31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
34
(No Transcript)
35
(No Transcript)
36
Wisc. flood, previous wave packet ?
Convection along Mei-Yu Front
37
1st Downstream Cyclogenesis
38
Mn flood from frontal overrunning
2nd Downstream cyclogenesis
39
Rossby waves and spring flooding (not just for
the cool season anymore)?Activity on the
mesoscale often, if not typically, requires an
amplified or perturbed planetary-scale wave
pattern, even for some warm season convective
events?A link between mesoscale convective
complexes over continents and initial forcing
ahead of larger-scale wave patterns?
16 May
Record cold
PA severe weather
Wisc/Chi. flood
Mn flood
11 June
40
November 2002
41
Time/Long. Diagram 250-mb Meridional Wind (m
s-1) Latitude Belt ( 35-60 N)
6-28 November 2002
6 Nov.
Cyclogenesis
India/T.C.
Flood
Tornadoes
Cyclogenesis
Oil Tanker
Alps Flood/Wind
Snow/Ice Storm
Cyclogenesis
Flood
Cold-Air
Cyclogenesis
Shuttle Launch Delay
Moroccan Flood
Cold-Air
Alps flood




28 Nov.
UK
Japan
UK
Cal.
42
Time/Long. Diagram 250-mb Meridional Wind (m
s-1) 35-60 N 6-28 November 2002

9-27 Nov.




UK
Japan
UK
Cal.
43

THOPREX Research Challenges (Observing
Strategies)

• Implications of improved knowledge of
  predictability and improvements in forecast skill
• Targeting and the concept that all measurements
  are not created equally has been introduced and
  even applied operational but basic questions
  remain unexplored and difficult to answer from
  operational data sets

NSF THORPEX Briefing, D. Parsons, 11/21/03
44
Prediction of global sensitive regions for
November 2001
Error growth hot zones
Courtesy of Shapiro and Thorpe
45
Targeted sensitivity Washington snowstorm
Courtesy of Shapiro and Thorpe
46
Targeted sensitivity Washington snowstorm
47
Targeted sensitivity Washington snowstorm
48
Targeted sensitivity Washington snowstorm
49
Atlantic-TOST Potential Area of Interest
50
THORPEX Atlantic - European TOST

• Two efforts
• Start in Oct on Extratropical (ET) transition
  of tropical cyclones
• Begin shift to winter cyclones in Nov, ops to
  mid-Dec.
• 2nd planning meeting to be held on September
  8-9
• Description
• Scheduled for mid Oct / mid Dec 2003
• EUCOS and other European/NA observational assets
  will be targeted on sensitive areas
• Analysis complete by the end of 2004
• Observing Systems
• 600 European AMDAR aircraft
• 16 European ASAP ships
• Additional radiosonde ascents (Canada, US,
  Europe)
• Research aircraft - dropsondes and other
  measurements (Germany, US and Canada)
• Additional drifting buoys (Western and Eastern
  Atlantic areas)
• Driftsonde flights from US East coast
• Meteosat 6 - rapid scan winds

51
Observations in sensitive areas(TReC_024)
52
North American Participation in 2003 Atlantic
Regional ExperimentObservational Phase 13 Oct
12 Dec

• IOP - ran from 13 Oct. until 12 Dec.
• 21 TReC cases triggered observations

TReC_023 Heavy Mediterranean rainfall severe
flooding Marseilles 2nd December

TReC_026 prolonged, heavy snow gale-force
winds Snow in Boston 8th December
53
Different Techniques Yield Different Areas
Illustration of the differences between the
results arising from different targeting
algorithms. Two cases from the NORPEX field
experiment are shown the intent is to select the
observation location that will minimize the
expected 24-h forecast error in the box at right.
Colored regions indicate the sensitive regions
as determined by an ensemble-based filtering
approach contours indicate region of increasing
observation sensitivity as determined by an
adjoint-based singular vector approach. From
Majumdar et al., QJRMS, 128, p 2527.
54
Data Assimilation and Observing Strategies

• Improved use of observation
• Quantify observing system errors
• Develop methods for efficient utilisation of
  high-volume datasets
• Improve use of satellite observations
• Improve assimilation of physical processes
• Targeting techniques
• Refine targeting strategies
• Generalise existing targeting techniques
• Test targeting algorithms for a wide range of
  weather systems
• Design observational networks
• Adaptive Data Assimilation
• Improve background-error covariances in existing
  assimilation schemes
• Develop methods for cycling flow-dependent
  background errors
• Develop adaptive quality control
• Incorporate model uncertainty into data
  assimilation procedures

55
Primary Science Goals

•     Impact studies using ECMWF, UK Met, METEO
  France, Swiss, Danish, Italian, and German Met.
  Services, NRL (for ET Storms), NCEP
• Test ability to select appropriate cases for data
  targeting
•     Test ability to predict sensitive areas
•     Test ability to target these areas with
  additional observations
•     Define the benefits
•     Test of new instruments potentially useful
  for targeting and/or  future operations
  (driftsonde, wind lidar, etc)
•     Tests, design and validation studies of
  proposed experimental  satellite techniques

56
COSMIC and THORPEX

• THORPEX is officially very interested in
  COSMIC. An area where we could contribute to the
  satellite revolution, if COSMIC was interested
• Contribution
• Evaluation/validation studies of COSMIC data sets
• Studies of the relative impact on forecast skill,
  building on (UK) Met Office and recent ECMWF
  results
• OSEs and OSSEs studies for justification and
  design requirements for future networks
• COSMIC and radiosonde data studies (sonde biases,
  representivity errors, etc.)

57
COSMIC Soundings in a Day
Red dots are current radiosonde sites, green dots
are 2500 COSMIC soundings in 24-h.
58
THORPEX and international initiatives

• THORPEX is the meteorological component of the
  International Polar Year 2007-2008
• Atmospheric predictability in the Polar Regions
• Improving use of satellite and in situ data from
  high latitudes
• Strong North American interest in THORPEX field
  campaign during IPY, especially in Arctic and
  Pacific regions
• Hurricane-AMMA flights for targeting

59
THORPEX Observing System Research

• Develop and test new airborne delivery systems
  for deploying in situ sensors
• Carry out field-demonstrations of prototype
  remote -sensing systems for future airborne and
  satellite deployments

ATDs driftsonde development for THORPEX
60
Pre-Vorcore campaign Kiruna, Sweden, Jan. Feb.
2002
7 flights, most flights end because of crossing
latitude limit Longest flight 45 days
2 balloon sizes 8.5 and 10 m diameter
Overpressure 5.7 20.4 hPa
Flight levels .100 kg/m3, .110 kg/m3 and
.137 kg/m3
61
Pre-Vorcore campaign Kiruna, Sweden, Jan. Feb.
2002
7 flights, most flights end because of crossing
latitude limit Longest flight 45 days
2 balloon sizes 8.5 and 10 m diameter
Overpressure 5.7 20.4 hPa
Flight levels .100 kg/m3, .110 kg/m3 and
.137 kg/m3
62
Hibiscus campaign Bauru, Brasil, Feb. Mar. 2004
8 flights, most flights end because of autonomy
or equilibrium level too low Longest flight 79
days
2 balloon sizes 8.5 and 10 m diameter
Overpressure 0 20.5 hPa
Flight levels .090 kg/m3, .120 kg/m3 and
.135 kg/m3
Available weight up to 25 kg
63
Hibiscus campaign Bauru, Brasil, Feb. Mar. 2004
8 flights, most flights end because of autonomy
or equilibrium level too low Longest flight 79
days
2 balloon sizes 8.5 and 10 m diameter
Overpressure 0 20.5 hPa
Flight levels .090 kg/m3, .120 kg/m3 and
.135 kg/m3
Available weight up to 25 kg
64
3. US-Science Overview and Objectives
Key weather systems in the West African and
Tropical Atlantic regions
SAL
AEWs
TC
MCSs
65
2. IOP 2
The WAM system is an ideal laboratory to study
scale interactions in monsoons
66
TRACK GUIDANCE JUST PRIOR TO G-IV MISSION
67
TRACK GUIDANCE JUST AFTER G-IV MISSION
68
UKMET MODEL
WITHOUT SONDES
WITH SONDES
Courtesy Julian Heming, UKMO
69
Driftsonde Trajectories15-30 July 2000
Excellent synoptic Coverage from daily flights
at 100 hPa
70
IOP 3 Downstream component of AMMA
The societal need for improved predictions of
Atlantic hurricane activity A majority of
tropical cyclones form from African weather
systems but which one?
71
(No Transcript)