Extreme Weather: Climate Change and Variability

1
Extreme Weather Climate Change and Variability
Greg Holland
Mesoscale and Microscale Meteorology Division,
ESSL National Center for Atmospheric Research
NCAR is Sponsored by NSF
2
Hypothesis

• Weather extremes respond to climate variability
  and change at much higher amplitude than the
  general weather-system population.
• Extreme weather may thus provide a bellwether
  for climate change

3
Summary

• Extreme Weather Variations with Climate
• Application of Extreme Value Theory
• Example Hurricanes in the North Atlantic

4
Extreme Weather Variations

• IPCC
• It is likely that heat waves have become more
  frequent over most land areas. WGI 3.8, SPM
• It is likely that the frequency of heavy
  precipitation events (or proportion of total
  rainfall from heavy falls) has increased over
  most areas. WGI 3.8, 3.9, SPM
• There is observational evidence of an increase in
  intense tropical cyclone activity in the North
  Atlantic since about 1970, and suggestions of
  increased intense tropical cyclone activity in
  some other regions where concerns over data
  quality are greater.

5
Extreme Tropical Cyclone Variations
Developed from a homogeneous satellite data
set Elsner et al (2008)
6
Cat 5 Atlantic Hurricanes
Change from Long-Term Mean
Amplitude extends well beyond any reasonable
issues with data errors
7
Extreme Value Theory


We utilize the Weibull distribution for which the
CDF and PDF are

Where parameters a and b determine the scale and
the shape, respectively.
8
Application to Atlantic Hurricanes
PDF (Normalized, HURDAT Smoothed)
CDF with Weibull fit a35, b1.9
9
Scale and Shape Variations
Shape Change b3.5
Original a40, b3
Scale Change a45
Scale, a, is related to the population mean and
variance and shape, b, almost exclusively to the
inverse of the variance (SD a a/b).
10
Hazard Rate Function
We define a Hazard (Failure) Rate Function by
(Hillier and Lieberman, 1986) For the Weibull
distribution this becomes For b1 the hazard
rate is independent of x, for bgt1 (lt1) it
increases (decreases) with x. (For Atlantic
cyclones b1.9).
11
Excedent Probability

• We are interested in the probability of exceeding
  a threshold event, Exgtc,
• The excedent likelihood decreases as the event
  becomes rarer (c/a increases), and/or the
  population less variable (b increases),.

12
Excedant Sensitivity

• Following Katz and Brown (1992), we define the
  relative sensitivity of P(Exgtc) to the scale
  and shape parameters as

13
Impact of Changes to the Scale and Shape
Parameters
220
0
230
380
14
1980-1994 vs 1994-2007
Comparing 1980-1994 with 1994-2007, a mean
intensity change of 3-4 m/s and a SD change of
2-3 m/s leads to an increased probability of Cat
5 hurricanes of 3 times! This is equivalent to
moving from 1 Cat 5 every 3-5 years to 1 every
year.
15
Comparisons
Change 1980/1994 to 1995/2007
16
Potential Impact of Climate Change

• Assume a mean wind intensity increase of 2-3 m/s
  per oC global ocean warming, then
• Cat 5 hurricane frequency increases by 70 for
  each oC warming
• Assume in addition that the variance also
  increase, say, by 2-3 m/s in standard deviation,
  then
• Cat 5 hurricane frequency increases by 150 for
  each oC warming!
• The mean and SD changes are below the current
  observational resolution.

17
Relationship a, b with PE69
18
Relationship a, b with PE69
Strong hysteresis over 1.5 cycles covering 70
years
19
Relationship SSTEA-GM with PE69
20
Consistency with Wave Accumulation
Decreased Shear
Strengthened Easterlies
Weakened Easterlies
Max SST Anomaly
East Pacific warming generates local Walker
Circulation
21
Summary

• Extreme value theory predicts that changes in
  extremes will be gtgtgt than those in the mean or
  variance (Hypothesis 1 is confirmed)
• The expected changes can be determined with some
  objectivity, given a knowledge of the current
  climate distribution and of the expected mean and
  variance changes
• Some interesting features of N. Atlantic changes
  emerge
• Hypothesis 2 is worthy of serious consideration
  in climate-change analysis
• Extreme weather may thus provide a bellwether
  for climate change