Transcript Tropical Cyclones and Climate Change in a High Resolution

Tropical Cyclones and Climate Change in a High
Resolution General Circulation Model, HiGEM
Ray Bell
Supervisors: Prof. Pier Luigi Vidale, Dr. Kevin Hodges and Dr. Jane Strachan
Introduction
Motivation
• Socio-economic impacts and changing risk with climate change.
• Impacts on the climate system, removing heat and moisture from the
ocean affecting large scale circulation.
Research Objectives
• Investigate the changes in TC activity (location, frequency, intensity,
structure and duration) with climate change.
• Investigate a change of natural variability mechanisms on TC activity
e.g. changing ENSO.
• [Investigate the impact of atmospheric resolution on TC activity with
climate change.]
• [How does TC activity change during a transient forcing vs. stabilised
forcing?]
What’s expected?
How will TC frequency change with climate
change?
(The number of TCs which form each year)
Stay the same
Decrease
Increase
“Warmer SSTs lead to more TCs”
“What controls TC frequency?”
time
Space
TCfreq= Σ Σ f(SST, dT/dz, mid level humidity,
absolute vorticity, vertical wind shear) +
Initial disturbances
Gray (1968); Emanuel (2005)
What’s expected?
How will TC intensity change with climate
change?
(The maximum intensity a TC can reach)
Decrease
•
Stay the same
Increase
TC intensity is governed by its immediate external environment
and internal processes
• TC intensity is hard to simulate in global models (constrained
by resolution)
- Large regional uncertainties (Knutson et al. 2010)
- Can we observe these changes? (Klotzbach and Gray, 2011)
Why do we get these changes?
TC frequency Decrease globally by 6-34% by 2100
• Distribution of SST change. Change in gradients -> Increase in VWS via
thermal wind balance (Vecchi and Soden, 2007)
• Weakening of the tropical circulation. Increase in dry static stability.
(Vecchi et al, 2006)
Knutson et al (2010)
TC intensity
Increase globally by 2-11% by 2100
• Decrease of TC frequency mainly from weaker storms.
When conditions are favourable for TC development it will be able to
utilize the energy available.
• Supported by theory (Emanuel, 1985) and idealised studies (Shen, 2000)
Idealised GCM simulations
HiGEM
UK’s new High-Resolution Global Environmental Model (Shaffrey et al, 2009)
HiGEM 4xCO2 30 yrs
1.25ox0.83o, ∆x50N = 90 km
HiGEM 2xCO2 30 yrs
HiGEM CTRL ~9x30 yrs
1/3o ocean model
HiGEM 1.1 CTRL 150 yrs
HiGEM 1.2 CTRL 117 yrs
Tracking algorithm (TRACK; Bengstton et al, 2007)
1) Locate and track all centres of high relative vorticity  35000/yr
2) Apply a 2-day filter to the tracks
 8000 storms / yr
3) Analyse vertical structure of storm for evidence of warm-core (tropical storm
structure)
 120 storms / yr
Assessing the model
At this model resolution we are
able to realistically capture
location and frequency
Strachan et al, (2012) in rev
Climate Change Simulations
Track density difference
2xCO2 - CTRL
4xCO2 - CTRL
Storms/month/106km2
Stippling if outside 9x30yr
CTRL variability
Climate Change Simulations
norm pdf
Grey shading is 9x30yr CTRL
variability
Increase in maximum intensity 
Max rel vor
Large scale forcing
Red: 2xCO2 - CTRL
Green: 4xCO2 - CTRL
50
NAtl
% change
-50
SST
VWS ppt RH700 ω500 TCfreq stronger TCfreq
50
% change
-50
NEPac
Conclusion
•HiGEM realistically captures the geographical location and TC
frequency compared to observations.
• HiGEM simulates a decrease of TC frequency in most regions
except for the North Indian Ocean and North Central Pacific
region.
• HiGEM simulates an increase of TC intensity, which only
becomes significant in the 4xCO2 experiment.
• An increase in VWS in the 4xCO2 over the North Atlantic
spreads to the North East Pacific and decreases TC freq.
• A weaker Walker circulation suppresses activity in the North
West Pacific and enhances activity in the North Central
Pacific.
Future work
•Continue Adding HiGEM1.2 onto my current study.
• Investigate the ENSO relation and different types of
El Niño and the impact they have on TC activity. How
these change with climate change.
• Apply my analyses to the different resolution
simulation (atmosphere only)
• Apply my analyses to the transient simulation
TC and climate change studies
Comparing
with
,
and
• Different tracking algorithms
• Model TCs are not exactly comparable to obs TCs
• Inhomogeneities of obs TCs in different basins and over
time.
• Different models resolutions/ different scenarios
- show different parameters to be of importance
Late 21st Century projections:
“storm-friendly”
Vecchi and Soden (2007)
Average of 18 models, Jun-Nov
“storm-hostile”
Vecchi et al., (2008)
Large-scale tropical
Atlantic climate changes
projected for late 21st
century by CMIP3 models
(A1B scenario). Average
SST change in MDR is
1.72oC with warming near
4oC in the upper
troposphere.
Temp anom
Knutson et al., (2008)
Shen (2000)
Emanuel Potential intensity
TC
Hurricanes
Small decrease of TCs. Small increase of major hurricanes
Increase in maximum
intensity 
MSLP-10m windspeed relation
5 member
ensemble for
N512, all 2005
Other models –
all years
Note: IBTrACS
uses 10min
winds, models
use instantan.
6hourly winds
TRACK Hodges (1995); Bengstsson et al. (2007)
• T42 ξ850 – Reduce noise. Comparison of different spatial resolution
data
• Minimum lifetime of 2 days and no constraint on the minimum
displacement distance. Capture more of TC lifecycle
•
•
•
•
Cyclogenesis (0-30oN over ocean)
Coherent vertical structure and warm core
Max T63 vor at each level from 850hPa to 250hPa
Intensity threshold T63 ξ850 > 6x10-5 s-1, ξ850 – ξ200 > 6x10-5 s-1 , for at
least 1 day (4 x 6hr).
• Search for warm core between p levels 850-500, 500-200hPa (+ ξ
value)
• Wind speed must attain 20m/s at 850hPa (change in slightly more
intense TCs) [att20 dataset]
• Statistical packages
Understanding natural variability
ENSO’s impact on geographical location
Understanding natural variability
ENSO’s impact on TC frequency
Climate Change Simulations
Change in TC frequency
Error bars are max and min of
9x30 yr CTRL variability
Change in SST
Zhao et al
(2009)
AMO ~= AMOC
Klotzbach and Gray (2011)
Sea Surface Temperature Difference
Jul-Oct
2xCO2 - CTRL
4xCO2 - CTRL
Sea Surface Temperature Difference (°C)
• Tongue of relatively less warm water compared to the rest of the tropics
• Grave results of TCs in this vicinity (NAtl).
• Leads to increased vertical wind shear (VWS) via thermal wind balance
Vertical Wind Shear Difference
Jul-Oct
2xCO2 - CTRL
4xCO2 - CTRL
Vertical Wind Shear difference (m/s)
• VWS spreads to the NEPac especially in the 4xCO2
• Detrimental affect on TCs.
• Reduced VWS in CPac favours development
Stippling if outside 5x30yr
CTRL variability
Walker Circulation Difference
Jul-Oct 0-10N°
2xCO2 - CTRL
4xCO2 - CTRL
-ω difference (Pa/s) and divU difference (m/s)
• Weakening of the tropical circulation inline with other studies (Vecchi and Soden,
2007)
• Favours development in the CPac and reduces TC frequency is the NWPac
Change in RH700
Vecchi et al (2007)
Change in –ω500
Change in ppt
Large scale tropical change
Climate Change Simulations
HadGAM – N96. 135km
HiGEM - N144. 90km
NUGAM - N216. 60km