Extreme Climate and Weather Events, Hurricanes and Climate

1
Extreme Climate and Weather Events,
Hurricanesand Climate

• Suzana J. Camargo

Guest Lecture, Mailman School of Public Health,
Columbia University, February 7, 2007.
2
Outline

• Extreme Events Basics
• Trends in Extreme Events
• Hurricanes Basics
• Hurricanes and ENSO
• Hurricanes and Climate Change

3
Weather and Climate Extremes

• Extremes - negative effects on society and
  ecosystems
• Floods, droughts, damaging high winds, extreme
  heat and cold, etc
• Some areas extreme events are beneficial
• Example northwest coast of Australia most of
  rainfall from sporadic tropical cyclones
  essential for water storage and supply.

Meehl et al. BAMS, 2000.
4
Concept of an Extreme Event

• Distribution of a weather phenomena
• Shading extreme events
• Tails of the distribution occur infrequently
  values far from the mean/median of the
  distribution

5
Common Types of Extreme Weather Events

• Heavy precipitation events - extreme daily
  precipitation exceeding 2 in. (50.8 mm).
• Flooding and landslides due to hurricanes (e.g.
  Mitch Central America, Katrina).
• Temperature extremes heat and cold waves (e.g.
  Europe summer 2004).
• Frost days important for crops
• Heavy wintertime events snow storms
• Tornados, thunderstorms, wind storms

6
Common Types of Climate Extreme Events

• Drought periods extended areas
• Extreme seasonal precipitation Venezuela, 1999
  landslides
• (B. Lyon, J. Climate 2003)

7
Extreme Events Issues

• Society has become more vulnerable to extreme
  events. Population and infrastructure increases,
  pollution event associated with storm run off,
  more coastal development, etc
• Lack of long-term climate data suitable for
  analysis of extremes. Few countries (United
  States, Australia, South Africa, Norway) have
  reliable precipitation data prior to World War
  II.
• How will (is) climate change affect extreme
  events? Models limitations spatial resolution,
  simulation errors, parametrizations.

8
Societal Impacts

• Frequency and/or intensity of extremes can cause
  major problems.
• Impact of climate change on society and
  ecosystems could be due to changes in the
  physical system or to changes in the
  vulnerability of society.
• Is the frequency of extreme events currently
  changing and/or is only the perception of an
  increase exacerbated by enhanced media coverage?
• Alternative definition of extreme event impact
  that an event has on society loss of life,
  economic losses, infrastructure destruction,
  pollution.

9
Possible Changes of Extremes
Change in Mean
10
Possible Changes of Extremes II
Change in Variance
11
Possible Changes of Extremes III
Change in Mean and Variance
12
Hurricanes
13
Atlantic 2004 Hurricane Season
14
2005 Atlantic Hurricane Season
15
2006 Atlantic Hurricane Season
16

• Impacts of hurricanes and typhoons on coastal
  regions can be very large both in life losses and
  economical cost.

Hurricane Katrina Track
17
Hurricane Katrina August 28, 2005
18
Katrina at landfall
Hurricane Katrina at landfall, August 29, 2005.
19
Where do hurricanes form?
Hurricanes
Typhoons
Severe Cyclonic storm
Severe Tropical Cyclone
General name Tropical Cyclone
20
Hurricane Tracks
21
Occurrence of tropical cyclones
Late summer, early fall Following the sun
warming of sea surface temperatures.
85 TCs per year 42 become hurricanes
Source K.A. Emanuel, Divine Wind, 2005.
22
Sea Surface Temperature - August
23
Sea Surface Temperature - February
24
Factors influencing hurricane occurrence

• Warm ocean
• Humid air
• Presence of spin in the atmosphere
• (vorticity)
• Convergence
• of humid air

25
Absence of vertical wind shear
Source K.A. Emanuel, Divine Wind, 2005.
26
Genesis Potential Index
Source K.A. Emanuel, Divine Wind, 2005.
27
Trigger Easterly Waves (Atlantic)
Source K.A. Emanuel, Divine Wind, 2005.
28
Easterly Waves (2)
Source K.A. Emanuel, Divine Wind, 2005.
29
Easterly Waves (3)
Source K.A. Emanuel, Divine Wind, 2005.
30
Factors controlling Hurricane Intensification

• Time 5-7 days to reach hurricane intensity.
• Possible disrupting factors
• Land
• Cold sea surface temperatures
• Ocean mixing (e.g. Ophelia)
• Vertical wind shear (symmetry)
• Dry air

31
Maximum Potential Intensity
32
El Niño Southern Oscillation (ENSO)
33
ENSO Typical Impacts
34
Atlantic Hurricanes and ENSO
35
Atlantic Hurricanes and ENSO (2)

• In El Niño (La Niña) years there are fewer (more)
  hurricanes in the Atlantic (Gray, 1984).
• Factors
• Larger vertical wind shear (Shapiro, 1987).
• Changes in atmospheric temperature and stability
  (Tang and Neelin, 2004).

36
Typhoons and ENSO
La Niña
El Niño
Shift in typhoon genesis location Source Chia
Ropelewski, J. Climate (2002)
37
Typhoon Intensity and ENSO
Source Camargo Sobel, J. Climate (2005).
38
Typhoons and ENSO
More intense and typhoons with long lifetimes in
El Niño years
Tropical Storms
Intense Typhoons
Typhoons
Tropical Cyclones Lifetimes
Source Camargo Sobel, J. Climate (2005)
39
Madden-Julian Oscillation and Tropical Cyclones

• Dominant model of variability in the tropical
  atmosphere with characteristic periods of 30-60
  days.
• Eastward propagation

40
MJO Phases
Source Wheeler Hendon, Mon. Wea. Rev. (2004).
41
Gulf of Mexico Hurricanes and MJO
Source Maloney Hartmann, Science (2000).
42
Hurricanes Eastern Pacific and MJO
Source Maloney Hartmann, J. Climate (2000).
43
Decadal variability in hurricane activity
Central Pacific
1966-1981
1982-1994
Source P-S Chu, J. Climate, 2002
44
Atlantic Multi-Decadal Signal
Source Goldenberg et al., Science (2001).
45
Climate change possible influences on
hurricanes

• Increase in sea surface temperature.
• Increase in the atmospheric temperature.
• Increase in the evaporation in the atmosphere
  (more humid atmosphere)

46
Hurricanes and Climate Change

• How could hurricanes change with global warming?
• Frequency
• Intensity
• Duration
• Precipitation
• Areas affected

47
Climate Change assessments

• Regions with hurricane occurrence is NOT expected
  to change.
• Increase in tropical cyclone precipitation (mean
  and peak) likely in some areas.
• No observed trends in the NUMBER of tropical
  cyclones.
• Changes in hurricane frequency unclear.

48
Hurricane Frequency

• Inconsistent results
• Different atmospheric models different outputs
  for changes in the number of hurricanes.
• Models differ in
• response (sensitivity)
• convection and other parametrizations
• resolution
• regional sea surface temperatures

49
Number of Tropical Cyclones
50
Hurricane intensityTheory
Hurricane as a Carnot engine
Source K.A. Emanuel, Nature (1987).
51
Hurricane intensity Models (1)
Source Knutson Tuleya, J. Climate (2004).
52
Hurricane intensity Models (2)
Source Knutson Tuleya, J. Climate (2004).
53
Hurricane Intensity (destructive power)
Source K.A. Emanuel, Nature (2005).
54
Hurricane Intensity (categories)
Source Webster et al., Science (2005).
55
Sea surface temperature trends
Source Webster et al., Science (2005).
56
Issues

• Quality of the data of tropical cyclone
  intensity.
• Length of the record (data starts in 1950).
• Most reliable data Atlantic and Western North
  Pacific more problems in other areas.
• NO trend in the number of LANDFALLING hurricanes.

57
How can we obtain past information on hurricanes?

• Archival studies
• Newspapers
• Official government records
• Ships logs
• Personal diaries

New England Hurricanes
Source Boose et a.l, Ecological Monographs
(2001).
58
Typhoon landfall history in southern China
Source K-b Liu et al., Ann. Assoc. Amer. Geog.
(2001).
59
Geological Proxies

• Sediment deposits produced by storm surge
  associated with a landfalling hurricane.

60
Corals cyclone frequency and intensity
(Australia)
Source Hayne Chappell (2001).
61
Questions

• Were there more or fewer hurricanes in various
  time in Earths past?
• Were they more intense or less?
• If the Earth continues to warm, especially the
  tropical oceans, due to anthropogenic global
  warming what will be the consequences on
  hurricane variability?
• What is the importance of hurricanes on Earths
  climate?

62
Tropical Cyclones Seasonal Forecasts

• Prediction of seasonal variability of tropical
  cyclones frequency, number of intense
  hurricanes, number of hurricanes that make
  landfall.
• Main methods
• Empirical Methods
• Dynamical Methods

63
Empirical Methods

• Existence of predictors affecting tropical
  cyclone formation.
• Up to one year in advance of the season.
• Historical data used to identify the predictors
  and estimate their weight in a statistical
  regression.
• Limitations
• Performance restricted by the number of past
  events.
• Assumption that future weather will behave in a
  manner similar to the past.

64
Empirical Methods (2)
Example December Predictors for Atlantic CSU
Forecasts
Source Klotzbach Gray, Weather and Forecasting
(2004).
65
Forecast Example
Source Colorado State University Atlantic
Hurricane Forecasts (Dec 2004).
66
Dynamical Seasonal Forecasts

• Use numerical models (atmospheric general
  circulation models).
• Two steps
• Prediction of the sea surface temperature
  anomalies.
• Identification and tracking of tropical
  cyclone-like structures in the atmospheric
  models.
• Issues with Dynamical forecasts
• Models low resolution
• No ocean-atmospheric interaction
• Models biases

67
Tropical Cyclones in Atmospheric General
Circulation Models

• Numerous studies showed that AGCMs can create
  model tropical cyclones with strong similarities
  to observed tropical cyclones (Manabe et al.,
  1970 Bengtsson et al., 1992 Vitart et al.,
  1997).

68
Typical AGCM Tropical Cyclone
69
Detecting and Tracking Tropical Cyclone-like
structures

• Using the output of the AGCM integrations,
  tropical cyclone-like structures are detected and
  tracked.
• Variables used in the detection and tracking
  algorithms
• Vorticity, sea level pressure, wind speed,
  temperature.

S.J. Camargo S.E. Zebiak, Wea. Forecasting,
2002.
70
IRI Tropical Cyclone Activity Experimental
Dynamical Forecasts
IRI Prediction Products http//iri.columbia.ed
u/forecast/tc_fcst
S.J. Camargo A.G. Barnston, in preparation.
71
IRI Tropical Cyclone Activity Experimental
Dynamical Forecasts
NTCNumber of named Tropical Cyclones ACEAccumula
ted Cyclone Energy , Location
centroid of all tracks.
72
How are the forecasts produced?

• Sea Surface Temperature forecasts produced.
• Atmospheric Model (ECHAM4.5) forced by sea
  surface temperature forecasts.
• Tropical Cyclone-like structures detected and
  tracked.
• Statistical corrections of the tropical cyclone
  activity based on the model climatology.
• Probabilistic forecasts of tropical cyclone
  activity.
• IRI Experimental Seasonal Tropical Cyclone
  Outlooks released

73
Atlantic 2005 TC Forecasts
74
Forecasting Hurricanes

• Weather forecasting models much higher
  resolution than climate models.
• Main Issues
• Chaotic nature of the atmosphere.
• Uncertainty in the initial conditions of the
  atmosphere.
• Model physics still need improvement
  (parametrizations).

75
Uncertainty in Forecasts
Source K.A. Emanuel, Divine Wind, (2005).
76
Forecast Errors
Source National Hurricane Center
77
Forecast Errors (2)
Source National Hurricane Center.
78
Summary

• Tropical cyclone seasonal forecasts can help
  guide preparatory measures for a hurricane
  season.
• Much research still needed in order to provide
  skillful probabilistic seasonal landfall
  forecasts.
• Forecasts of hurricane tracks have improved
  steadily in the last decades.
• Forecasts of hurricane intensity still need lots
  of improvement, depending on more understanding
  on the theory and dynamics of hurricane
  intensification.