Mark OMalley

1
Arctic Oscillation (AO)/North Atlantic
Oscillation (NAO) and Applications to Medium
Range Forecasting

Mark OMalley WFO Pleasant Hill Kansas City
Chapter of the AMS December 4, 2006
2
Arctic Oscillation (AO) Definition
Arctic Oscillation (AO) is defined as the leading
mode of Empirical Orthogonal Function (EOF)
analysis of monthly mean 1000mb height during
1979-2000 period.
Based on height anomalies poleward of 20o
latitude in the Northern Hemisphere from the
NCEP/NCAR reanalysis dataset at a horizontal
resolution of (lat,lon)(2.5X2.5). The seasonal
cycle has been removed from the monthly mean
height field.

AO is sometimes referred to as the northern
annular mode
3
AO Positive Warm Phase
Characterized by intensification of the polar
vortex with lower than normal pressures over the
Arctic region. Meanwhile, higher than normal
pressures become evident over the Central
Atlantic (and to some degrees the North Central
Pacific) inducing an enhanced belt of zonal
westerlies across the northern latitudes (near
45o latitude).

Illustration from National Geographic Magazine,
2000
4
AO Negative Cool Phase
Characterized by higher than normal pressures
over the Arctic region and a breakdown of the
polar vortex. Lower than normal pressures develop
over the Central Atlantic and Central Pacific.
This leads to a more meridional flow pattern
across the Northern Hemisphere.

Illustration from National Geographic Magazine,
2000
5
Vertical Cross section of the Northern Hemisphere
Polar Vortex
1000mb height anomalies does not necessarily
correlate directly with midlevel height anomalies
(particularly in non-cold season). In general,
the stronger the anomaly, the more correlation
through the depth of the atmosphere will be seen.

AO Index measured as 1000mb height anomaly from
monthly mean. Negative anomaly Positive AO value
6
Average DJF 500 mb heights and Anomalies based on
AO phase
Enhanced Polar vortex and contraction of the
midlatitude westerlies

Breakdown of the Polar vortex with relaxation and
increased meridional component of the midlatitude
westerlies.
7
DJF Temperature Precipitation Anomalies based
on AO Phase

8
Controversy Is the Arctic Oscillation (AO)
really just the North Atlantic Oscillation (NAO)
in disguise?
9
AO/NAO Controversy Temperature Response

10
AO/NAO Controversy Arguments For vs. Against a
Separate Teleconnection Mode

• Separate Mode

AO center of action covers more of the Arctic,
with a larger horizontal scale and more zonal
symmetry than the more regionalized NAO.
(Thompson and Wallace, 1998)
The AO signal has a strong interconnection with
troposphere/stratosphere variability, while the
NAO is primarily confined to the troposphere.
(Kodera and Kuroda, 2004) (Wang et al., 2005)

Correlation between AO and NAO is near 74
interannually and near 75 during the winter
season. There also exists several periods where
AO and NAO measures are nearly out of phase
leading to slightly different spatial variability
of surface air temperature. (Wettstein and
Mearns, 2002) (Wang et al., 2005)
11
AO/NAO Controversy Arguments For vs. Against a
Separate Teleconnection Mode

• The Same Mode

The Empirical Orthogonal Function (EOF) of the AO
pattern has no straightforward interpretation as
a covariance structure and is mainly a
reflection of similar behavior in the Pacific and
Atlantic basins. Therefore, AO cannot be truly
viewed as a teleconnection pattern. (Ambuam et
al., 2001)

Tests of Rotated Principle Component Analysis
(RPCA) on the AO pattern fail to show correlation
between one-point teleconnection maps and the
centers of action and the AO pattern is an
artifact of the EOF analysis and not a true
teleconnection pattern. (Livezey, 2006)
It follows that the NAO and AO are synonyms
they are different names for the same
variability, not different patterns of
variability. The difference between the terms is
in whether that variability is interpreted as a
regional pattern controlled by Atlantic sector
processes or as an annular mode whose strongest
teleconnections lie in the Atlantic sector.
(Wallace, 2000)
12
AO/NAO Compositing Methodology

• Utilized normalized monthly AO and NAO indices
  for the cold season months of Nov-Mar with
  monthly average temperature departures from the
  1971-2000 normals. Period of record 1950-2006.

• Average temperature departures were categorized
  as above, near, or below average based on a
  tercile ranking of 1971-2000 normals.

• Monthly AO and NAO indices were characterized as
  above, near, or below normal based on
  approximately 1 standard deviation from the
  normalized value of 0. Consideration was also
  given to provide an even distribution. (AO
  Below lt -1.5, AO Above gt 1.0) (NAO Below lt -1.0,
  NAO Above gt 1.0) This approach agrees well with
  previous work done by Wettstein and Mearns, 2002
  ( 1.0) and Wu et al., 2005 ( 1.2).

13
AO/NAO Compositing Methodology

• Determine the number (and probability) of
  occurrences of above, near, or below average
  temperatures for each category of AO/NAO.

• Determine Statistical Significance/Risk Analysis

• Assume a null hypothesis that the distribution
  is random and has no dependence on AO/NAO

• Perform a Student-t test to look at all possible
  distributions and determine the likelihood
  (probability) that the given AO/NAO distribution
  could be a random distribution (heavily dependant
  on sample size)

14
AO/NAO Compositing Methodology

• Results show that the AO/NAO distribution has
  strong statistical significance with less than a
  5 chance that the distribution is random

15
AO Compositing Results

16
NAO Compositing Results

17
Applications to Medium Range Forecasting

• Assumptions

• In order for monthly AO/NAO values to result in
  greater than 1 standard deviation above or below
  normal, there must significant stretches within
  the month (on a weekly basis) where AO/NAO values
  are much above or below the normalized average

• In order for monthly temperatures to fall in the
  below or above tercile, there must exist
  significant stretches within the month (on a
  weekly basis) where temperatures are much above
  or below average

• Given the strong statistical significance on a
  monthly basis and the conceptual model of what
  defines periods of strongly positive and negative
  AO/NAO, time periods shorter than a month must
  also have a strong correlation between AO/NAO
  phase and temperature departure

18
Applications to Medium Range Forecasting Example
MCI Daily Temperature Data

Period of Oct 9-22 characterized by highly
negative AO
19
Applications to Medium Range Obtaining Data
CPC Website
http//www.cpc.noaa.gov/products/precip/CWlink/dai
ly_ao_index/teleconnections.shtml
Different Ensemble solutions

Correlation coefficient generally poor past 7
days. Performance best during the cold season.
AO measure typically increases and decreases
faster than predicted by ensemble members
20
Summary

• Monthly AO/NAO phase shows a strong correlation
  with monthly temperature departures across the
  Lower Missouri River Valley during the cold
  season.

• It is assumed that this correlation can be
  applied to shorter time scales on the order of
  several days to weeks.

• There exists a strong possibility that the
  combination of forecast AO/NAO phase with
  situational awareness and pattern recognition
  could be successfully used to improve temperature
  forecasts in the medium range days 5-7 against
  the climatologically weighted GFS guidance.

21
References
Ambaum, Maarten, B.J. Hoskins, and D.B.
Stephenson, 2001 Acrtic Oscillation or North
Atlantic Oscillation? J. Climate, 14, 3495-3507.
Higgins, R.W., Y. Zhou and H.-K. Kim, 2001
Relationships between El Niño-Southern
Oscillation and the Arctic Oscillation A
Climate-Weather Link. NCEP/Climate Prediction
Center ATLAS 8.
Livesey, Robert, 2006 North Atlantic and Arctic
Oscillations. COMET Climate Variability Course.
Boulder, August 2006. 9pp.

Thompson, D.W.J, and J.M. Wallace, 1998 The
Arctic Oscillation signature in the wintertime
geopotential height and temperature fields.
Geophys. Res. Lett., 25, 1297-1300.
Thompson, D.W.J, and J.M. Wallace, 2000 Annular
modes in the extratropical circulation. Part I
Month-to-month variability. J. Climate, 13,
1000-1016.
Thompson, D.W.J, and J.M. Wallace, 2000 Annular
modes in the extratropical circulation. Part II
Trends. J. Climate, 13, 1018-1036.
22
References
Wallace, J.M., D.W.J. Thompson, and Z. Fang,
2000 Comments on Northern Hemisphere
Teleconnection Patterns during Extreme Phases of
the Zonal-Mean Circulation. J. Climate, 13,
1037-1039.
Wallace, J.M., 2000 On the Arctic and Antarctic
Oscillations. 2000 NCAR Advanced Studies Program
Summer Colloquim on Dynamics of Decadal to
Centennial Climate Variability. 40pp.
Wang, Dongxiao, C. Wang, X. Yang, and J. Lu,
2005Winter Northern Hemisphere surface air
temperature variability associated with the
Arctic Oscillation and North Atlantic
Oscillation. Geophys. Res. Lett, 32, in print.

Wettstein, Justin and L.O. Mearns, 2002 The
Influence of the North Atlantic-Arctic
Oscillation on Mean, Variance, and Extremes of
Temperature in the Northeastern United States and
Canada. J. Climate, 15, 3586-3600.
Wu, Aiming, W.W. Hsieh, A. Shabbar, G.J. Boer,
and F.W. Zwiers, 2005 The nonlinear association
between the Arctic Oscillation and North American
winter climate. Climate Dynamics, 26, 865-879,
doi10.1007/s00382-006-0118-8.