A Review of 2005

1
A Review of 2005 2006 Hurricane Seasons and
Forecasts
Muthuvel Chelliah Climate Prediction Center,
NCEP/NWS
NOAA 31st Annual Climate Diagnostics and
Prediction Workshop Earth System Research
Laboratory, Boulder, CO. October 23, 2006.
2

• Two seasons 2005 2006
• Two Basins (at least) North Atlantic East
  Pacific (third Central Pacific)
• Two Outlooks May August for N.Atlantic and
  only May Outlook for E. Pacific.
• Verification ( of Tropical Storms TS,
  Hurricanes H, Major Hurricanes MH and
  Accumulated Cyclone Energy (ACE) Index.
• I will show even other years for comparison and
  historical behavior.

• NOAAs outlooks for N. Atlantic (since 1998)
  and E. Pacific (since 2003, Experimental in
  2003-2004) have been fairly accurate.
• This is a very preliminary analysis of what
  happened to the forecasts this 2006 season (not
  over yet, ends in November).

3

• All outlooks are official NOAA outlooks and are
  made by teams of scientists from
• Climate Prediction Center (CPC Lead Center) in
  Maryland
• National Hurricane Center (NHC)
• Hurricane Research Division (HRD) in Florida
• Hydrometeorological Prediction Center (HPC) in
  Maryland

North Atlantic Outlook Team CPC Gerald
Bell (Lead), Muthuvel Chelliah, Kingste Mo NHC
Christopher Landsea, Eric Blake and Richard
Pasch HRD Stanley Goldenberg East Pacific
Outlook Team CPC Muthuvel Chelliah (Lead),
Gerald Bell, Kingtse Mo NHC Christopher Landsea
and Eric Blake HPC Todd Kimberlain
4
NOAAs 2005 Seasonal Hurricane Outlooks Issued
16 May 2005
5
NOAAs 2005 Seasonal Hurricane Outlooks Issued
22 May 2006
6
2005 Atlantic Seasonal Activity Outlook - May
NOAAs Accumulated Cyclone Energy (ACE) Index
7
2005 Atlantic Seasonal Activity Outlook - August
8
NOAAs Updated 2005 Atlantic Hurricane Outlook
95 to 100 Chance of an Above-Normal
Season (Measured by forecasted numbers, strength,
and duration of named storms) Forecasted
Seasonal Totals 18-21 Tropical
Storms 9-11 Hurricanes
5-7 Major Hurricanes
180-270 of median ACE value
Forecasted for August-November 2005
11-14 Tropical Storms 7-9
Hurricanes 3-5 Major
Hurricanes 110-200 of median ACE
value
Already during 2005 7 Tropical Storms, 2 Major
Hurricanes (Dennis and Emily) ACE index 68 of
median
9
For 2006 Less North Atlantic Activity Expected
Than 2005

• 2006 Atlantic Hurricane Outlooks
• May Outlook called for less activity than 2005
• August Outlook called for even less activity than
  May outlook

9
10
East Pacific Hurricane ActivityMay 2005 Outlook
NOAAs Accumulated Cyclone Energy (ACE) Index
11
East Pacific Hurricane ActivityMay 2006 Outlook
12
North Atlantic Hurricane Season
TS - 27 (18-21) H - 15 (9-11) MH - 7 (5-7)
2005
ABOVE-NORMAL (Hyper-Active)
ACE - 285 (180-270)
2006
NORMAL
0.0
TS - 9 (12-15) H - 5 (7-9) MH 2 (3-4)
ACE 90 ? (110-170)
0.5?
Tracks Graphics Courtesy of Unisys
13
East Pacific Hurricane Season
TS - 15 (11-15) H - 7 (6-8) MH - 2 (2-4)
2005
BELOW NORMAL
ACE 65 (45-95)
2006
NORMAL
Tracks Graphics Courtesy of Unisys
TS - 16 (12-16) H - 8 (6-8) MH - 5 (1-3)
ACE 95 (45-85)
14
2004
2004
North Atlantic
ACE gt250
1995
ACE gt250
-0.5
15
1993
1993 Aug-Sep-Oct Nino 3.4 Index 0.4 (from CPC)
officially NEUTRAL
1994 Aug-Sep-Oct Nino 3.4 Index 0.7 (from CPC)
officially El Nino
1994
16
Basis and Procedures for the Seasonal Hurricane
Outlooks

• Interannual and multi-decadal extremes in
  tropical North Atlantic, East Pacific and
  Central Pacific hurricane activity are not
  random.
• Instead, they result from a set of coherent and
  inter-related atmospheric and oceanic conditions
  controlled by tropical climate patterns, and some
  possibly related to increases in ocean surface
  temperatures.
• NOAAs makes seasonal hurricane outlooks by first
  analyzing and predicting these leading recurring
  patterns of climate variability in the Tropics,
  and then predicting their impacts on hurricane
  activity.

The Two dominant climate factors that
influence/control seasonal hurricane activity in
all three regions are El Niño/ Southern
Oscillation (ENSO) Gray (1984) Tropical
multi-decadal climate variability Chelliah and
Bell (2004) Bell and Chelliah (2006)
17
Regional North Atlantic Conditions associated
with Active Hurricane Season
Warmer SSTs Lower Wind Shear Lower Surface
Pressure (Red Area)
Higher Pressure in Upper Atmosphere
Favorable Winds From Africa
Upper-level Easterlies Expand Westward (Green
arrows)
Weaker Low-Level Easterly Winds (Blue arrow)
18
_
200-hPa Velocity Potential Loading Patterns
during ASO Contours shown for active
Atlantic hurricane eras. East-west see-saw in
tropical convection between West Africa and
Amazon basin

_

19
Composite Conditions Associated with Active/
Inactive Atlantic Hurricane Seasons
Anti- Cyclonic

• Regional Anomalies Exhibit Larger-Scale
  Convective Signature
• Inter-hemispheric symmetry
• Baroclinic structure to wind anomalies (reversal
  in anomalies between 200-hPa and 850-hPa).

20
Lower Shear
Higher Shear
Enhanced easterly vertical shear produces
3-celled pattern of anomalous shear magnitude
21

• Since 1995 (thorough 2005) 9 out of 11 years
  are above-normal active hurricane seasons in
  N.Atlantic except for two El Nino years 1997
  and 2002.
• The generally negative correlation between
  the N.Atlantic and E. Pacific basin hurricane
  activities has strengthened in the last couple of
  decades.

22
The opposite phase relation between the two
basins activities seems to extend far back. BUT
the East Pacific Data considered unreliable
before 1965/1970
23
The Transition to Weak El Niño Conditions was
not Predicted
The Climate Forecast System (CFS) forecast issued
May 9th 2006 indicated slightly below-average
SSTs over central equatorial Pacific, thus
failing to predict the impending El Niño.
CFS Forecast SST Departures (oC) Forecast Issued
May 9th 2006
Observed SST Departures (oC)
15
24
Very Warm Waters Returned during August 2006
During the peak of the 2006 season, tropical
Atlantic SSTs were generally 1-2oF above
Average. Departures were sometimes comparable
to last year (2-3oF).
1
2
1
2
1
2
16
25
Forecasts of current El Ninoin APRIL 2006
As measured by Nino 3.4 SST Anomaly
ENSO ADVISORY La Niña conditions are expected to
continue during the next 1-3 months.
26
Forecasts of current El Ninoin MAY 2006
ENSO ADVISORY ENSO-neutral conditions are
expected to prevail during the next 3-6 months.
27
Forecasts of current El Ninoin JUNE 2006
ENSO ADVISORY ENSO-neutral conditions are
expected to prevail during the next 3 months.
28
Forecasts of current El Ninoin JULY 2006
ENSO ADVISORY ENSO-neutral conditions are
expected to prevail during the next 3 months.
29
Forecasts of current El Ninoin AUGUST 2006
ENSO ADVISORY ENSO-neutral conditions are
expected to continue for the next one to three
months, with a 50 chance that weak El Niño
conditions will develop by the end of 2006.
30
Forecasts of current El Ninoin SEPTEMBER 2006
ENSO ADVISORY ENSO-neutral conditions are
expected to continue for the next one to two
months, with weak warm episode (El Niño)
conditions likely by the end of 2006.
31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
34
(No Transcript)
35
(No Transcript)
36
(No Transcript)
37
In Closing..

• So, what happened to the Above Normal North
  Atlantic Hurricane activity AND the Below
  Normal East Pacific outlook ? The season is not
  over yet (ends in November) But, most likely,
  the season for both basins will end as NORMAL.
• Not just NOAA, but all other outlooks did
  indicate Above Normal activity for the North
  Atlantic basin during 2006 season. NOAA has been
  issuing Hurricane outlooks since 1998 and this is
  the first year the outlook may turn out to be not
  so pretty..
• Preliminary culprit The unexpected and sudden
  emergence of El Nino right in the middle of
  the season - which was not forecast by any of
  the dynamical or the statistical models. Needs
  further analysis
• After 2005 being such a record active year, to be
  followed by such a bla.. 2006 season,
  hopefully brings some sanity to the global
  warming hurricane connection debate - at least
  demonstrates that just local ocean warm waters
  alone (thermodynamic argument GCM style
  thinking) is not enough, and that atmospheric
  circulation factors such as vertical wind shear,
  cyclonic vorticity, etc. (due to local and remote
  influences) are just as important in deciding the
  seasons activity.

38
(No Transcript)
39
(No Transcript)
40
(No Transcript)
41
Summary

• I have not addressed cause(s) of the LF/
  multi-decadal variability, nor has this issue
  been addressed in Bell and Chelliah (2006). Our
  work focuses on understanding the observations.
  This is particularly important if we are to
  understand how combinations of climate factors
  (multi-decadal variability and GH warming) will
  impact hurricane activity in the future.
• The leading tropical multi-decadal modes of
  convective variability account for many key
  circulation and SST anomalies associated with
  seasonal and multi-decadal extremes in Atlantic
  hurricane activity since 1950.
• The transition to current active Atlantic
  hurricane era shows strong link to phase change
  in tropical multi-decadal signalsee-saw in
  tropical convection between West Africa and
  Amazon Basin.
• This phase change accounts for many conditions
  known to favor active hurricane seasons, and is
  consistent with transition to warm phase of AMO.
• Warmer Atlantic SSTs since 1995 are associated
  with significantly more hurricanes in large part
  because of their common association with
  overlying circulation anomalies.

42
Some Unresolved Issues of Concern

• We see large tropical circulation/ convection/
  SST changes occurring between active/ inactive
  hurricane eras.
• Numerical simulations indicate little to no wind
  changes across tropical Atlantic with increasing
  CO2. Seems inconsistent with observations,
  suggesting weak (at best) CO2 impact at this
  time.
• How can minimal impact on winds be maintained if
  SST increases substantially in a future CO2
  world? Is this realistic?
• If not, how might CO2 increases affect leading
  tropical modes.
• Ultimate cause(s) of tropical multi-decadal
  variability
• We should study potential CO2 impacts
  simultaneous with increasing our understanding of
  the tropical multi-decadal variability.

43
AEJ
Tropical disturbances remain in extended region
of cyclonic shear while moving westward into low
vertical shear, warm SST environment of central/
western MDR.
44
Measuring total seasonal activity The
Accumulated Cyclone Energy (ACE) index The phrase
"total seasonal activity" refers to the
collective intensity and duration of Atlantic
named storms and hurricanes occurring during a
given season. The measure of total seasonal
activity used by NOAA is called the Accumulated
Cyclone Energy (ACE) index. The ACE index is a
wind energy index, defined as the sum of the
squares of the maximum sustained surface wind
speed (knots) measured every six hours for all
named systems while they are at least tropical
storm strength. NOAA uses the ACE index, combined
with the numbers of named storms, hurricanes, and
major hurricanes, to categorize North Atlantic
hurricane seasons as being above normal, near
normal, or below normal.
NOAAs Atlantic hurricane season
classifications Above-normal season An ACE index
value well above 103 x 104 kt2 (corresponding to
117 of the median ACE value or 110 of the
mean), or an ACE value slightly above 103 x 104
kt2 combined with at least two of the following
three parameters being above the long-term
average number of tropical storms, hurricanes,
and major hurricanes. Near-normal season An ACE
index value in the range 66-103 x 104 kt2
(corresponding to 75-117 of the median or
71-110 of the mean), or an ACE index value
slightly above 103 x 104 kt2 but with less than
two of the following three parameters being above
the long-term average numbers of tropical
storms, hurricanes, and major hurricanes. Below-n
ormal season An ACE index value below 66 x 104
kt2, corresponding to below 75 of the median or
71 of the mean.
45
NOAAs 2005 East Pacific Hurricane Outlook
Issued 16 May 2005
Season and 2005
2005 Climatological
Activity Type Outlook
Observed Mean (1971-2000)
Chance Above Normal 10 Below
Normal Chance Near Normal 20 Chance
Below Normal 70
Tropical Storms
11-15 15
15.5 Hurricanes
6-8 7
9.2 Major Hurricanes 2-4
2
4.5 ACE of Median 45 - 95
65 110
46
ASO Tropical Atlantic SSTA
Regressed
Observed
SST warming in MDR is consistent with phase
change in tropical multi-decadal signal and
associated weaker trade winds.
47
Weaker trade winds, along with reduced flow of
dry, subsiding northeasterly flow coming into MDR
from east side of Bermuda High.
48
Less Activity was Expected and Forecast This
Year in 2006 Than 2005

• 2006 Atlantic Hurricane Outlooks
• May Outlook called for less activity than 2005
• August Outlook called for even less activity than
  May outlook

9
49
(No Transcript)
50
Observed vs. Regressed
Explained variance 82.
Current active Atlantic hurricane era associated
with phase change in tropical multi-decadal
signal.
51
ASO 5-year LP-Filtered Variance Explained by
Leading Tropical Multi-Decadal Modes

LF variability in all KEY circulation features
strongly related to leading tropical
multi-decadal modes.
52
February-March 2005 SST Departures
Very warm ocean temperatures continue in
Tropical Atlantic
53
Red shading shows lower vertical wind shear
between 200-850-hPa in the main hurricane
development region (black box). Blue shading
shows higher than normal vertical wind shear. The
3-celled pattern of anomalies between the eastern
tropical Pacific and Africa has been in place
since 1995. This pattern has resulted in more
Atlantic hurricanes and fewer eastern Pacific
hurricanes.
54
(No Transcript)
55
AEJ
Anti-Cyclonic Cyclonic
56
Less Activity Expected This Year Than 2005

• May 22, 2006 Seasonal Hurricane Outlook
• Fewer hurricanes expected than 2006
• The combinations of conditions was not expected
  to be the same as in 2005
• La Niña-like rainfall pattern was not expected to
  be as pervasive as last year
• Seasonal Atlantic SSTs were not expected to be as
  warm
• Upper-level and lower-level wind and air pressure
  patterns not as favorable.

10
57
Less Activity Expected This Year Than 2005

• August 8, 2006 Updated Seasonal Hurricane Outlook
• Fewer hurricanes than May outlook
• Transition from La Niña-like rainfall patterns to
  ENSO-neutral conditions occurred in June -- more
  rapidly than expected
• Seasonal sea-surface temperatures in the tropical
  North Atlantic were not as warm as previously
  expected.
• Atmospheric wind patterns over the eastern
  Atlantic were not as conducive as originally
  expected, partly due to the transition into El
  Niño conditions.

11
58
August Outlook Called for Fewer Hurricanes Than
May Outlook Why?
Transition away from La Niña-like Rainfall
Pattern during June
A transition from La Niña-like conditions to
ENSO-neutral conditions occurred by mid-June, and
El Niño conditions developed by September.
Therefore, the critical combination of 1) an
active hurricane era and 2) La Niña was no longer
possible. Central Pacific became wet and the
shear increased over the Caribbean and Western
Atlantic (neither supports an active hurricane
season)
Mid-June through mid-July 2006 Conditions
12
L
59
ASO 5-yr LP-Filtered Surface Temperature
Variance ()
Leading tropical multi-decadal modes capture much
surface temperature variability associated with
AMO and WAFR monsoon.
60
August Outlook Called for Fewer Hurricanes Than
May Outlook
Sea-surface temperatures in the tropical North
Atlantic were not as anomalously warm as
previously expected
Sea-surface temperatures during July 2006 were
0.5o-2oFahrenheit above average. Significant
warmth was confined to Caribbean Sea.
13
61
Cause of the Apparent Change in Atlantic Ocean
Temperatures between 2006 and 2005
Differences in tropical easterly trade winds and
associated pressure pattern
2006 has weaker easterly trade winds in the
hurricane formation areas (green box), and a
reduced flow of moist, tropical air. These
conditions cause increased evaporative cooling
and cooler sea-surface temperatures. Low
pressure (like in 2005) favors hurricane
development. A detailed understanding of the
reasons behind these changes requires more
research.
Anomalous Surface Winds and Air Pressure (Lower
pressure indicated by blue shading)
14