Transcript Presentation

On the Changing Characteristics
of Atlantic Hurricanes
Greg Holland
Summary:
Atlantic Changes and Relationship to SST
Natural Variability vs Climate Change
Based on Holland and Webster, 2006,
Proceedings of the Royal Society of London (revised)
Holland, Hawaii 102706
ESSL
The times they are a changin!
Bob Dylan
Criticisms
Webster et al 2005
(also Emanuel 2005)
• The data base is biased
(Landsea et al), but Kossin et al
(2006)
• Models and theory predict a
much smaller trend, hence the
data are wrong! (Landsea etal
2006), but Michaels et al 2006
state the opposite?
• All we are seeing is natural
variability (NOAA)
• The impacts are minor
compared to demographic and
related changes (Pielke et al)
Holland, Hawaii 102706
ESSL
Focus on the North Atlantic and
Hurricane Frequency
2005 Hurricane Season
25oN
Gulf
Eastern Atlantic
Western Atlantic
(NASA)
Holland, Hawaii 102706
ESSL
Atlantic Tropical Cyclone Trend
North Atlantic Tropical Cyclones 1905-2005
30
Tropical Cyclone Number
Individual Year
Aircraft
Recce
Satellites
25
20
TC3
15
10
TC2
TC1
5
9-Year Running Mean
0
1900
1920
1940
1960
1995
1980
2000
2020
Year
Holland, Hawaii 102706
ESSL
TC Number-SST Relationship
TC3
TC1
TC2
9-year running mean
Holland, Hawaii 102706
Changes between TC
climate regimes are
accompanied by similar
changes in eastern
Atlantic SSTs;
SST leads cyclone
changes and explains
>60% of the variance in
TC numbers (due
entirely to regime
changes).
ESSL
TC Climate Regimes
18
Regime 3
Number of Named Storms
16
14
~150% increase in
TC numbers
Regime 2
12
Regime 1
10
8
6
4
25.6
25.8
26
26.2
26.4
SST East Atlantic
5-y running mean
26.6
26.8
27
1 TC for each 0.1oC increase in SST
1 Hurricane for each 0.2oC increase in SST
Holland, Hawaii 102706
ESSL
Hurricane Proportions
Since 1900 and esp.
since 1945 hurricane
proportions have been
remarkably steady with
only a few % variation;
High hurricane
proportion prior to
1900 and occasional
year with only
hurricanes up to 1940
are indicative of a
tendency to miss
weaker systems in
earlier years.
Holland, Hawaii 102706
ESSL
Major Hurricane Variation
Period before 1900
dominated by data
issues;
No evidence of any
trend in proportions
since 1900;
Remarkable, out of
phase oscillation in
major and minor
hurricanes with multidecadal period.
Independent of Hurricane or Tropical
Cyclone Variation
Holland, Hawaii 102706
ESSL
However
• While there is no trend in the proportion of major
hurricanes, there is a distinct trend in the number
of major hurricanes.
Holland, Hawaii 102706
ESSL
Natural Variability vs Climate Trend?
We address the causal relationship:
North Atlantic Tropical Cyclones 1905-2005
30
Tropical Cyclone Number
Individual Year
25
SST signal
20
Relationship between SST and
Hurricanes
15
10
Natural variability or Climate Trend
5
9-Year Running Mean
0
1900
1920
1940
1960
Year
1980
2000
2020
(After Curry Webster and Holland 2006)
Holland, Hawaii 102706
ESSL
SST-Hurricane Relationship
• Webster et al (2005), Emanuel (2005) and Hoyas et al (2006)
noted a strong global relationship with climatological
intensity changes.
• Gray (1990), Landsea et al (1999), Goldenberg et al (2001),
Vitart and Anderson (2001) all find a strong relationship
between eastern North Atlantic SSTs and tropical cyclones.
• We have shown that Eastern North Atlantic SSTs explain
>60% of smoothed variance in Atlantic Tropical Cyclones
since 1905 (due entirely to regime changes);
• The SST-TC relationship is not direct, but arises from
related atmospheric environmental changes, such as
vertical wind shear (Delworth, 2006).
Holland, Hawaii 102706
ESSL
Surface Temperature Variability
Greenhouse Gases
Volcano
1970
Solar
“Natural” Forcing
There is no known natural forcing mechanism that
can explain the surface temperature increases since
1960 (Meehl et al 2004, 2006).
Sulfate
Ozone
Santer et al (2006) obtain similar result for the
Atlantic tropical cyclone development region.
Holland, Hawaii 102706
ESSL
First Conclusion
•
•
•
The strong relationship between increases in
storm and hurricane numbers and increases in
SSTs leads to the inescapable conclusion that
the majority of enhanced current TC and
hurricane activity is a direct result of greenhouse
warming.
Why the inconsistency with theoretical and
climate model projections (Henderson-Sellers et
al 1998, Knutson et al, several studies)?
Answer: The formation and intensification
regions have moved!
Holland, Hawaii 102706
ESSL
North Atlantic TC Location Changes
4
0.4
Subtropical
SST
2
0.2
0
0
-2
-0.2
SST Anomaly (C)
Storms by Month
Equatorial
9-y Mean
-4
1950
-0.4
1960
1970
1980
1990
2000
Year
SST variations explain 69% of the variance in
equatorial (easterly wave) developments.
Holland, Hawaii 102706
ESSL
Conclusion
• North Atlantic tropical cyclone (and major hurricane)
frequency changes due to anthropogenic climate change is
not in the future, a significant trend appears to have already
occurred.
• This is consistent with the climate change modelling study
by Oochi et al (2006).
North Atlantic Tropical Cyclones 1905-2005
30
Tropical Cyclone Number
Individual Year
25
20
15
10
5
9-Year Running Mean
0
1900
1920
1940
1960
1980
2000
2020
Year
Holland, Hawaii 102706
ESSL