WH

MM

M

MM

M

M MM M

MM

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

0.2

0.0

-0.2

-0.4

-0.6

-0.8

-1.0

Monthly internode correlations of the Mobile (M), Rogers (WR), and Hurrell (WH) NAO indi-ces for 1948-1997. The 99% (95%) significance level is 0.36 (0.28).

WH

WH

WH

WHWH

WH

WH

WHWH

WHWR WR

WR

WR

WRWR

WR

WR

WR

WR WR

WR

WH

The "mobile" subtropical high nodes are mainly clus-tered in the center of the Atlantic basin, with the exceptionof June and July (located over the Gulf Stream) and Janu-ary (eastern Atlantic). There is a general southeast-to-northwest migration of the subtropical node, from winterto summer, and a return from summer to winter, with theexception of southwestward excursions in February-March and especially November. The "mobile" subpolarand subtropical nodes undergo nearly "parallel" month-to-month displacements, so that their latitudinal separationgenerally remains constant. Sensitivity tests of the nodallocations were done by repeating the calculations forthree other SLP data-sets from 1958 onward and for twoadditional 40-year intervals: (1878-1917 and 1918-1957).The general conclusion of these sensitivity tests is thatnodal NAOm locations are robust with respect to the SLPdata source for a single period of 40-50 years, but that thelocation can change from one such period to another.

The monthly internode correlation plot indicates thatthe NAO signal weakens somewhat in summer, but thatthe oscillation is present year-round. The winter-basedindices of Hurrell and Rogers capture the winter NAO sig-nal well, but only the NAOm captures the NAO signal forevery calendar month.

Seasonality of the NAO

Low-frequency time series of the Mobile NAO index for contrasting seasons of (a) winter(DJF)/summer (JJA) and (b) spring (MAM)/autumn (SON). Superimposed is the low-frequency annual time series.

Temporal VariationsApplying the NAOm nodal locations

developed from the NCEP reanalysisdata-set to the University of East Anglia(UEA)/Jones data-set, we constructed alow-frequency annual time series. Theannual time series has a century-scalevariation that began with a strong nega-tive epoch in the 1870s-1880s, afterwhich it was positive for an extendedperiod (1900-1950s) before transitioninginto a negative epoch in the 1960s. Thelatter was followed by an upward trend inthe 1980s. A variable feature of theselow frequency epochs is a couplingbetween the seasons that dominate theannual cycle (e.g. summer/autumn forthe negat ive event o f 1875-1885;winter/spring for the positive event of1900-1914; autumn/winter for the nega-tive event of the 1960s). A primary con-clusion is that low-frequency variation oft h e N A O h a s s t r o n g s e a s o n a ldependence. All seasons can signifi-cant ly ampli fy or diminish the low-frequency annual values of NAOm.

While the internode correlation plot indi-cates that seasonal mobility is important ina two-point representation, this does noteliminate the possibility that a 2D patternframework (e.g. EOF) may be required tocapture the essence of NAO variations.We have tested the adequacy of our 2-point index (NAOm) by evaluating its abilityto capture the North Atlantic surface west-erly airflow. Specifically, we have correlat-ed NAOm, Rogers' winter NAO index, andthe EOF-based index maintained byNOAA's Climate Prediction Center with (i)Barry and Perry's index of westerlies overthe 35°N-55°N / 70°W-10°E region and (ii)a similar westerly index used by H. H.Lamb. We conclude that, at least for repre-sentations of the North Atlantic surfacewesterlies over the above domain, the EOFrepresentation does not add to the utility of2-point indices.

Monthly correlations of the Mobile (M), Rogers (WR), andan EOF-based index from NOAA's Climate PredictionCenter (C) with westerly indices (a) Barry and Perry and(b) H. H. Lamb for the period 1958-1999. The 99% (95%)significance level is 0.39 (0.30).

e. Jul-NAOm f. Jul-WH

g. Nov-NAOm h. Nov-WH

+.2 +.4 +.6 +.8-.2-.4-.6-.8

Spatial correlations (1948-1997) of reanalysis SLP with the mobile NAO index andthe winter based Hurrell (WH) index highlight the large-scale variability that each indexrepresents. The comparisons here are for March, May, July, and November since theyparticularly illustrate the seasonal differences in signal, strength and associatedteleconnections. The NAOm index, which has nodes mostly clustered in the centralAtlantic basin, generally has correlation maxima in the central North Atlantic, while W Hgenerally has correlation maxima along the eastern Atlantic rim. This is particularly evi-dent in March. It is also noteworthy that the March subpolar NAOm node is the stron-gest and broadest of all calendar months and extends deep into the Arctic. In May, theNAOm exhibits a strong teleconnection to Mexico and the southwestern US, when apositive correlation feature is located over the area of the July-August "monsoon" of theU.S. Southwest. The interannual variability of this monsoon has been linked to the latespring circulation over Mexico and the southwestern U.S. In July, the NAOm has itswesternmost nodal locations for the year, with a strong correlation dipole over the west-ern Atlantic, but of smaller spatial scale than in late winter. The Novemberteleconnection field for NAOm has the same general character as in May, except thatthe November pattern does not include teleconnections with Mexico and the south-western U.S.

Monthly sea level pressure (SLP) data from the NCEPreanalysis for 1948-1999 are used to develop a seasonally andgeographically varying “mobile” index of the North AtlanticOscillation (NAOm). NAOm is defined as the difference betweennormalized SLP anomalies at the locations of maximum negativecorrelation between the subtropical and subpolar North AtlanticSLP. The subtropical nodal point migrates westward and slightlynorthward into the central North Atlantic from winter to summer.The NAOm index is robust across data sets, and correlates morehighly than EOF coefficients with historical measures of westerlywind intensity across North Atlantic midlatitudes. As measured bythis “mobile index”, the NAO’s nodes maintain their correlation fromwinter-to-summer to a greater degree than traditional NAO indicesbased on fixed stations in the eastern North Atlantic (Azores,Lisbon, Iceland). When the NAOm index is extended back to 1873,its annual values during the late 1800s are strongly negative due tonegative contributions from all seasons, amplifying fluctuationspresent in traditional winter-only indices. In contrast, after themid–1950s, the values for different seasons sufficiently offset eachother to make the annually-averaged excursions of NAOm smallerthan those of winter-only indices. Global teleconnection fieldsshow that the wider influence of the NAO -- particularly in thewestern North Atlantic, eastern North America, and Arctic -- is moreapparent during spring-summer-autumn when the NAOm is used tocharacterize the NAO.

Diane H. Portis 1, 2, John E. Walsh 2,Mostafa El Hamly 1, 3, and Peter J. Lamb 1, 4

1 Cooperative Institute for Mesoscale MeteorologicalStudies, The University of Oklahoma, Norman, Oklahoma73019

2 Department of Atmospheric Sciences, University ofIllinois, Urbana, Illinois 61801

3 Moroccan Direction de la Météorologie Nationale(DMN), Casablanca, Morocco

4 School of Meteorology, The University of Oklahoma,Norman, Oklahoma 73019

This paper will appear in the J. Climate in mid-2001.

The framework we chose for a seasonal-ly dependent NAO index was the identifica-tion of the NAO nodal locations that anchorcoherent areas of maximum negative corre-lation of SLP in each calendar month. Twobroad boxes encompassing potential nodallocat ions were selected. Both boxesspanned the same longitudinal range(70°W-0°), with the latitudinal range for thenorthern (southern) node being 55°N-80°N(20°N-45°N). For each calendar month, theSLP value at each grid point within thenorthern box was correlated with eachgridded SLP value in the southern box. Thetwo most negatively correlated north-southpoints were selected as the NAO nodes forthat calendar month. If more than one pairof nodes had similarly high negative correla-tions (i.e., the correlations differed by lessthan 0.05), the final selection criterion wasproximity to the long-term monthly meanlocations of the subpolar low and the sub-tropical high.

Validation of 2-Point NAOIndex

MethodAbstractSeasonality of the NAO

Hemispheric Teleconnections of the NAO

1

23

124

56

7 8

9 10

11

11

7

3 4&62

129

10

8

1&5WR

WR

WH

WH

(a)

(b)

M MM

M

MM

MM

M

M

M M

C

C

C

C

C

C

C

C

C C

CC

-0.4-0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

MM

M M

MMWR

M

M

M MMC

C

C

C

C

C

C C

CC

C

C

-0.4-0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

WR

WR

WR

WR

WRWR WR

WR WR

WR

WR

M

WR

WRWR

WR WR WRWR

WRWR

WR

WR

WR

(a)

(b)

,

J F M A M J J A S O N D

J F M A M J J A S O N D

AGU Chapman Conference on "The North Atlantic Oscillation", November 28 - December 1, 2000, Orense, Galicia, Spain.

Locations of nodes defining the three NAO indices considered --Mobile, Rogers, and Hurrell. The monthly nodal locations of theMobile index (computed for 1948-1999) are labeled by monthnumber; the station-based nodes of the Rogers' and Hurrell's indi-ces are located by "WR" and "WH" respectively. Arrows indicatethe annual march of the NAOm nodes.

+.2 +.4 +.6 +.8-.2-.4-.6-.8

a. Mar-NAOm b. Mar-WH

d. May-WHc. May-NAOm