Tropical cyclone in a warmer climate: A global cloudsystem resolving approach

1
Tropical cyclone in a warmer climate A global
cloud-system resolving approach
K.Oouchi1, Y. Yamada1, M. Satoh1,2, H. Tomita1,
A. Noda1, and W. Yanase3 1. Frontier Research
Center for Global Change, JAMSTEC 2. Center for
Climate System Research, The University of
Tokyo 3. Ocean Research Institute, The University
of Tokyo 2nd International Summit on
Hurricanes and Climate Change

May 31 - June 5, 2009
2
background direction of this
study
3
MTSAT-1R
NICAM
Boreal winter simulation (snapshot at Day 10)
Miura et al. (2007,Science)
4
Why global-cloud-resolving model ?
Issue - how much the resolution should be fine
to capture typhoon dynamics and discuss its
climatology such as future change
Thunderstorms seen from the Space Shuttle
5
Pros and cons global warming study using GCRM

• Pros
• The use of global cloud-resolving model (GCRM)
  is a reliable
• way of getting around the uncertainty in
  cumulus convection
• scheme inherent to conventional climate
  models (Satoh et al. 2008)
• It allows to discuss change in structure of
  tropical cyclone by
• resolving cloud cluster of O (100 km )
  scale, and marginally resolving
• mesoscale convection of O (10km) scale
  and to derive more reliable
• estimate of future changes
• It is in harmony with the expected revolution
  in climate prediction
• endorsed in the Climate Summit Declaration
  (Shukla et al., 2009
• BAMS) and high-end computing
• ? hopefully to contribute to
  IPCC-AR5, with increased
• accuracy, reliability and details for the
  global warming issues

6
Pros and cons global warming study using GCRM

• Cons
• GCRM experiment is computationally
  demanding, with
• the forced compromise of short time
  integration and small
• sampling number
• ? unable to derive climate statistics should
  play a role in a
• different arena

Then what strategy should we take ?
case study approach for cloud process
and triggering process studies - TC
genesis process study (Fudeyasu et al. 2008)
- TC genesis preconditioning by MJO (Fudeyasu et
al. 2008
Oouchi et al.
2009a) expect to give more reliable
insight on the triggering disturbances and
mechanisms compared to conventional climate
models.
7

• Aim of the talk
• to discuss results on the general
• statistics of TCs in GCRM run, as a 1st
  step
• limited statistics in the sense that the
  sampling number
• in this case study approach is
  insufficient for rigorous
• statistics assessment
• basic statistics of future change
• track, frequency, intensity
• regional difference of the change
• large-scale environmental factors
• to discuss a strategy for future research

8

• Contents
• TC tracks, frequency, intensity change
• Environmental control (GPI analysis)
• - Emanuel and Nolan,
  2004
• Case study of the extremes

9
Model Experimental Design
10

• Nonhydrostatic ICosahedral Atmospheric
• Model (NICAM)

Satoh et al. (2008)
Tomita et al. (2004)
11
14-km mesh
Grabowski (1998)
12
NICAM development phase

• 1 Inclusion of moist processes and performance
  test as
• a global cloud-resolving model (15-30 day
  run)
• Aqua-planet experiment
• Tomita et al. (2005) , Miura et al. (2005) ,
  Nasuno et al. (2007),
• Iga et al. (2007)
• 2 Inclusion of realistic components (seasonal
  run)
• boreal winter experiment (30days)
• Miura et al. (2007) , Fudeyasu et al.
  (2008) Nasuno et al. (2009)
• Sato et al (2009)
• boreal summer experiment (3-5months)
• Noda et al. (2009) low-level cloud
• Oouchi et al. (2009a,b) triggering
  of TC by MJO / Asian monsoon
• 3 Global-warming experiment This study
• changes from the present-day simulation
• are discussed, with the seasonal run
  improved radiation
• scheme being the control case

13
Design of the global warming experiment

• CMIP3 multi model ensemble
  18 members
• sea surface temperature
• future SST
• Reynolds OI SST for 2004 CMIP3-based ?SST
• ?SST (future present) Important climate
  change forcing
• future ensemble average for the future
  25-year period (2075?2099)
• present ensemble average for the 25-year
  period (1979-2003) of
• the 20th century climate
  experiment
• sea ice concentration
• give a difference between future and present
  25years average
• 
  (Mizuta et al, 2008)
• greenhouse gases (SRES A1B scenario)
• CO2 doubled 100 years later
• ozone, other gases the same as the control
  (2004)

Thanks to Climate Research Division/MRI for the
sea ice dataset and processing codes
14
start
2004
Present-day climate
Global warming climate
Spin-up
20xx
May
June
July
August
September
July
June
May
August
October
September
15
Results
16
SST forcing
??
CTL
GW
GW - CTL
El Nino type
17
Global frequency and distributionof tropical
cyclone
JJASO
OBS CTL GW
18
Number of Tropical cyclone
JJASO
Reduced by 25
OBS
CTL
GW
OBS Unisys. Best Track
19
Possible change of the intensity
Surface wind (z10m)
JJASO
Surface pressure
The most intense TC 871 hPa
Under GW condition climate, we can see the
intensification of wind speed and minimum
pressure.
20
Environmental influence (GPI) Emanuel and Nolan
(2004), Camargo et al. (2007)
JJASO
GPI-CTL
GPI-GW
GW - CTL
Vertical shear
Relative Humidity
Absolute vorticity
Potential intensity
21
Environmental influence(GPI Atlantic)
JJASO
GPI-CTL
GPI-GW
GW - CTL
How about the change in weak incipient cyclonic
disturbance that has a potential to develop into
intense tropical cyclone ?
Vertical shear
Relative Humidity
Absolute vorticity
Potential intensity
22
Environmental influence (warm-core, wind speed
criteria removed)

• CTL ? GW
• ? 1 (regular tracking)
• 144 ? 135 (relaxed-criteria tracking)

23
GPI-analysis summary

• GPI provides potential genesis location with
• high fidelity useful metrics for capturing
  large-scale genesis control
• Over the western Pacific, eastward shift
• in the increased GPI region is evident,
  suggesting eastward shift in the genesis region
• Over the North Atlantic, the GPI shows
• two-tiered characteristics
• increases in the genesis region
• decreases in the development region

24
Relationship between minimum sea level pressure
and cloud top
Speculation (ref. Emanuel, 1986) (with no
consideration of moisture change) TC in GW -gt
cloud top becomes taller more
significantly for intense TC as thermodynamical
efficiency (TB-Tout)/TB (Emanuel, 1986)
becomes smaller for intense TC
enegy for development and maintainance of intense
TC is more increased in the stabilized warmer
climate than in the present-day climate
25
Summary and Future works
26

• Summary 1 The GCRM global-warming experiment
  suggests
• Increased intensity and decreased frequency in
  the globe
• - in consistent with the relevant
  statement in IPCC AR4
• Decreased frequency in the Atlantic more
  significantly, compared to the other oceanic
  basins
• - GPI increases in the genesis region,
  but does not in
• the development region
• - Frequency of weak incipient disturbance
  (including TD)
• almost remains unchanged ( 5
  reduction)
• - Large-scale control ultimately governs
  the frequency
• but process study is necessary to
  clarify the mechanism
• Higher cloud tops for more intense TC, in
  consistent with Emanuels theory (Emanuel 1986)
  to a first-order approximation more energy is
  necessary for intense TC
• in future warmer condition

27
Summary 2 Strength of the model (GCRM)
- Frequency of TCs in the western Pacific are
simulated better than that in
the conventional climate models
(e.g., Oouchi et al., 2006)
- This can be related to the fact that GCRM has
a strength in simulating the triggering
disturbances such as MJO in the region
(Fudeyasu et al. 2008 Oouchi et al.,
2009a)
-This suggests a high potential of GCRM for
better prediction of tropical cyclogenesis in
the western Pacific (the most intense TC
projected min. surface pressure of 871
hPa NW Pacific)
28

• Future works focus on the genesis stage and
  triggering
• mechanism
• to understand the change in stochastic processes
• including triggering disturbances, such as
  MJO and
• tropical waves
• to understand the change in the mesoscale
  processes
• and structure in the genesis and evolution
  of TCs

Vision further ahead

• ocean-atmosphere coupled high-resolution GCRM
• for TC study

• more enhanced interaction/collaboration with
  climate
• modeling community TCMIP by Prof. Walsh is a
  good way

29
(No Transcript)
30
Supplementary slides
31
Global frequency and distributionof tropical
cyclone (12.5, portion)
JJASO
OBS CTL GW
32
Global frequency and distributionof tropical
cyclone (17.5)
JJASO
OBS CTL GW
33
Trends in Tropical Atlantic Main Development
Region SSTs
Main Development Region
34
Conditions Associated With theActive 2004
Atlantic Hurricane Season
Source Chris Landsea (NOAA/National Hurricane
Center)
35
Temperature profile from equator to 30N
36
Relation between minimum sea level pressure and
radius of eye wall
37
1 Indian Ocean() 2 W.Pacific Ocean() 3 E.
Pacific Ocean(180-90W) 4 Atlantic(90W-0) ???????(
EQ-45N)
38
(No Transcript)
39
Preliminary analysis 2-3
Western Pacific ocean
Indian ocean
Eastern Pacific ocean
Indian,Atlantic,Eastern Pacific??????????????????.
Western Pacific??????????????GW????????????????.
TC?????????????.
Atlantic ocean
40
(No Transcript)
41
(No Transcript)
42
(No Transcript)
43
Supplementary - genesis
44
A simulated preconditioning of typhoon genesis
controlled by a boreal summer Madden-Julian
Oscillation event in a global cloud-system-resolvi
ng model
45
Different nature of convectiveorganizationsimulat
ed in NICAM
46

• Nonhydrostatic ICosahedral Atmospheric
• Model (NICAM)

Satoh et al. (2008)
Tomita et al. (2004)
47
NICAM development phase

• 1 Inclusion of moist processes and performance
  test as
• a global cloud-resolving model (15-30 day
  run)
• - Aqua-planet experiment
• (Tomita et al., 2005 Miura et al., 2005
  Nasuno et al., 2007)
• 2a Inclusion of realistic components (seasonal
  run)
• - sea surface temperature
• - topography
• - meridional asymmetry (complexity)
  
• 2b Improvement of model physics and climatology
• aiming at contributing to IPCC AR5
• - parameterization for subgrid-scale
  convective
• systems (for 14km-mesh model)
• - cloud microphysics

48
Aims

Understand the mechanism of typhoon genesis
influenced by the propagation of MJO
Advantages of using NICAM (1) better
simulation of MJO (e. g., Miura et al. 2007)
(2) resolves interactions among wide-ranging
scales of convection including
mesoscale(Yamasaki, 1983),
synoptic-scale to planetary-scale MJO
Understand mechanism of organization of
tropical convection in association with the
propagation of MJO
49
Why we chose the year 2004

• The year was marked by enhanced tropical cyclone
• activities
• 10 TCs made landfall on Japan
• (also famous for the typical El-Nino-Modoki
  year,and meteorologically interesting, Ashok et
  al.(2006))
• The season experienced clear signals of MJO
  propagation as well
• The year is among the target years of the CLIVAR
  intercomparison project to which our group plans
  to contribute

50
Experimental Design
51
JJA Mean FieldsResolution Dependency(14-km vs
7-km mesh)
52
JJA climatology (global)
JJA
Obs.
14km
7km
53
JJA climatology (global)
JJA
Obs.
14km
7km
54
JJA climatology(Asia)
UV850hPa Precipitation
14km
Obs.
JJA mean fields precipitation in purple
55
JJA climatology (Asia)
UV850hPa Precipitation
7km
Obs.
JJA mean fields precipitation in purple
56
An MJO eventand tropical cyclogenesis
57
zonal velocity200hPa (1 June12 Jul, 2004)
observation
NICAM
simulated eastward-propagating, planetary scale
velocity signal
58
zonal velocity200hPa (1 June12 Jul, 2004)
observation
NICAM
SCC1
modulation of warm-pool convection super
cloud cluster Hayashi and
Sumi (1986), Nakazawa (1988)
59
Case A
latitude-time section(1 June12 Jul 12, 2004)
precipitation
A2
SCC1
relative vorticity
zonal velocity
low-level westerly gt spawned vortical
disturbances gt TC formation
60

precipitation
vorticity 850hPa

zonal velocity

vorticity 200hPa

Upper-level cyclonic vorticity another
contiributor to TC formation
61
precipitation rate
A2
A1
pre-A2
A2 develops to the east of A1 horizontal
shearwestward-moving disturbances
62
precipitation rate
A2
A1
pre-A2
63
precipitation rate
A2
A2
The minimum Ps of A2 drops down to 970 hPa
64
- Comparison of the simulated track against that
of TC0407
TC resembles the observed TC0407 ?
Observation
NICAM
03
Jul01
28
Jun24
30
25
26
TRMM/NASA/NASDA
65

66
change in the zonal circulation before /after
MJO arrives
U200hPa
?
region of interest
U850hPa
lt0 (easterly shear)
67
change in the zonal circulation before /after MJO
arrives
upper tropospheric trough in the
subtropics
U200hPa
?
region of interest
U850hPa
lt0 (easterly shear)
68
Case A
latitude-time section (1 June12 Jul, 2004)
precipitation
A2
SCC1
relative vorticity
zonal velocity
before (1), during/after (2,3) the propagation
1
2
3
69
Summary of case study (1/2)

• 1. MJO cycle in June August, 2004 was
  reasonably
• simulated in the 14km-mesh GCRM run gt
  only 1st cycle was reported today
• - The phase of dynamical fields in the
  eastward propagation
• was basically in consistent with the
  observation
• 2. Genesis of tropical cyclone for the
  analysis period
• (01June 20July) was simulated more than
  two weeks
• after the start of the integration
• (Oouchi
  et al. 2009 in revision)
• - The genesis was in synchronous with
  enhanced cyclonic
• vorticity that originates from the
  equatorial low-level
• westerly area trailing behind SCC (MJO)
  - in consistent
• with twin-cyclone viewpoints
• e.g., Nakazawa
  (1986), Zhu et al. (2003)

70
Summary of case study (2/2)

• 3. The condition favorable for the genesis was
  suggested in
• terms of modulation of the warm-pool
  induced vertical
• shear by the MJO propagation
• - Genesis happens when the propagation of
  SCC/MJO
• acts to weaken the vertical shear of the
  zonal circulation
• centered over the Pacific warm pool
• - Upper tropospheric trough in the
  subtropics can play a role
• 
  

Further implication MJO modulates the warm-pool
induced vertical circulation, which controls the
tropical cyclogenesis and geographical preference
and maintenance of SCC. Cold pool, which is
well-known in the mesoscale meteorological
community, plays an important role in the
maintenance of SCC, and eastward-moving gravity
wave
71
Future works

• Origin of the MJO-associated organized
  convections
• including tropical cyclone and super
  cloud cluster
• especially in terms of
• favorable/unfavorable control to the
  specific
• organized form of tropical convection
  (tropical
• cyclone and SCC are among
  well-recognized form)
• role of the equatorial waves in the
  organization
• importance of cold pool in the
  tropical phenomena
• (extension of Oouchi (1999) and
  Oouchi and
• Yamasaki (2001) )

72
??????????? (Yanase et al. 2007)

• ??7? x 7????????0.5 hPa????????????TC??
• ??80 m height??????17m s-1??
• ??10? x 10???????????????(700, 500,300
  hPa??)?2K??
• ??????????(30S30N)???

73
TC genesis potential index (GPI)(Camargo et al
2007)
74
(No Transcript)
75
(No Transcript)
76
???

• ???
• 2004??????
• ?????????????????
• ???(??? Oouchi et al, 2006)??????
• ???????????????????
• ?????????????????(14km?7km???)
• 2004?????????
• ??????????????AR4????
• ????
• ?????????73ms-1, ????870hPa?TC???

77
??

• ???????????????
• - ???????????
• - ???????TC???
• ??????????????????
• - ?????????????????
• ??????? (???7km?????)
• ???????????????,????TC
• ???????
• ? ??????INDEX????, ?????
• ????????? (MJO, MRG?, ?????)

78
NICAM???TC?????????? ????
FRCGC NICAM ????

• NICAM?????????????? (2)
• ???????????????????
• ???????????????
• ????(MJO,?????)?????????
• ? ????????????????????
• (??????????????????????????)

79
NICAM???TC?????????? ????
FRCGC NICAM ????

• ????????
• (???????????????????????)
• ???????????? (??,??,????????)
• ??????
• ??????MJO????????????
• ????????????????????
• ?????????????

80
Precipitation movie (7-km run)
81
Hierarchical convections embedded in SCC
Hierarchy of SCC (Nakazawa, 1988)
SCC3
LCC (Large-scale Cloud Cluster) -Oouchi
and Yamasaki (2001)
82
Tropical cloud-resolving studies using NICAM

83
relative vorticity at 200, 500, 850, 925hPa (Jun
- Aug,2004)
2
1
3
4

• Cameroon highland
• Kenya/Uganda, Ethiopia
• Highland (incl. Kilimanjaro)
• 3. Maritime continent
• 4. Columbia/Ecuador mountains

vortex disturbances properties vary with height,
and trapped in specific area, seemingly
topographically determined
84
Case A
SCC1
A2
A1
Case A
SCC2
85
Case A(to the east of the maritime
continent)time evolution of TCs(6hourly
snapshots)
86
Another case study - MJO event in the boreal
winter, 2006
Miura et al., Science 318, 1763 -1765 (2007)
87
Case A
SCC1
A2
A1
Case A
SCC2
88
Case B(to the west of the maritime continent)
89
Zonal circulation (anomaly from zonal mean)
condensates
Case B
Easterly shear stronger than that in the eastern
branch of maritime continent, thereby enhancing
basic-state easterly shear upgradient momentum
transport
90
Case B
Case B
91
Case B
precipitation rate
92
Case B
precipitation rate
93
Case B
precipitation rate
94
2004?6? Walker Circulation anomaly
??
??
95
BL-scheme Summery and Future

• Summary
• Interseasonal experiment by 14km-mesh global
  model is now being conducted
• BL cloud in NICAM looks nice so far
• off the coast of California, Peru, Guinea,
  South-East of the Atlantic Ocean, South of Indian
  Ocean,
• Spatial development in the subtropics along the
  GPCI cross section
• maybe too much along the mid-latitude, though
• Future
• Improvement of the boundary-layer process
• Treatment of the subgrid-scale condensation for
  small-scale clouds
• Time-dependent turbulent closure for the
  high-resolution model
• Validation of Climatology and cloud development
• Diurnal variation
• Detailed comparison with GPCI
• Role of large-scale circulations
• Sensitivity experiments
• What is a key to reproduce the BL cloud?
• Influence of the vertical resolution