IntroductionThe positive phase of the Southern Oscillation whichcontinued through summer 2000/01 (Fawcett 2002)returned to a neutral phase by autumn 2001 (Pahalad2002). This neutral phase persisted through winter2001 (Fawcett and Watkins 2002) and spring 2001

(Trewin and Fawcett 2002). Summer 2001/02 alsoshowed predominantly neutral conditions, howeversubsurface oceanic analyses were starting to showsigns of potential El Niño development.

This summary reviews the southern hemisphereand equatorial climate patterns for summer 2001/02,with particular attention given to the Australasian andPacific Regions. The main sources of information forthis report are the Climate Monitoring Bulletin

Aust. Met. Mag. 52 (2003) 63-72

63

Seasonal climate summary southernhemisphere (summer 2001/02):a continuation of near-normalconditions in the tropical Pacific

Felicity M. Gamble National Meteorological and Oceanographic Centre, Commonwealth Bureau of

Meteorology, Australia(Manuscript received January 2003)

Southern hemisphere circulation patterns and associatedanomalies for the austral summer (December to February)2001/02 are reviewed, with emphasis given to the Pacific Basinclimate indicators and Australian rainfall and temperature pat-terns. After three successive positive phases (from spring 2000to autumn 2001) of the Southern Oscillation, near neutral con-ditions persisted through the remainder of 2001, and the sum-mer of 2001/02.Summer rainfall was above to very much above average

through most of the Northern Territory, eastern WesternAustralia and northwestern South Australia. Below to wellbelow average rainfall was recorded in parts of the Gascoyneand Pilbara regions of Western Australia and on the centraleast coast border region extending west and then south to thetri-state border of New South Wales, Victoria and SouthAustralia. For the rest of the country, summer rainfall was nearaverage. Summer was cooler than average across most of thecontinent with respect to both maximum and minimum tem-peratures, particularly near the tri-state border of SouthAustralia, Northern Territory and Western Australia. Aboveaverage temperatures were recorded along the Queenslandcoastal regions.

Corresponding authors address: Felicity Gamble, NationalMeteorological and Oceanographic Centre, Commonwealth Bureauof Meteorology, GPO Box 1289K, Melbourne, Vic. 3001, Australia.Email: F.Gamble@bom.gov.au

(Commonwealth Bureau of Meteorology, Australia)and the Climate Diagnostics Bulletin (ClimatePrediction Center, Washington). Further detailsregarding sources of data are given in the Appendix.

Pacific basin climate indicesThe Troup Southern Oscillation Index*The set of largely neutral monthly values of the TroupSouthern Oscillation Index (SOI) for spring 2001(+1.4, –1.9, +7.2) (Trewin and Fawcett 2002) was fol-lowed by –9.1 (December), +2.7 (January) and +7.7(February), indicating the continuation of a neutralstate of the Southern Oscillation. The rise in the SOIover summer 2001/02 was mostly due to an increasein magnitude of positive pressure anomalies at Tahiti(particularly during January and February) than tonegative anomalies at Darwin.

The 30-day average values of the un-normalisedTahiti minus Darwin pressure difference (not shown)generally remained near the climatological averagefor most of summer 2001/02. However, for a briefperiod in December the pressure difference droppedbelow one standard deviation from the mean.Similarly, during January the index rose to near onestandard deviation above the climatological mean. Bythe end of February the pressure difference hadreturned to near normal values.

The neutral SOI values were reflected in theDecember/January and January/February values ofthe Climate Diagnostics Center (CDC) MultivariateEl Niño-Southern Oscillation (ENSO) Index (MEI).Both periods displayed weakly negative MEI values,thereby indicating conditions marginally on the LaNiña side of neutral. In comparison to the previous 52years (the total number of years on record), theDecember/January and January/February values wereboth ranked 29th, only just above the middle ranking,confirming their neutral status. As the MEI is derivedfrom a number of atmospheric and oceanic indicators,it reinforced the truly neutral state suggested by thesummer 2001/02 values of the SOI.

Figure 1 shows the monthly SOI values fromJanuary 1998 to February 2002. A curve of five-month moving averages has been superimposed onthe graph.

Outgoing long wave radiationFigure 2, adapted from the Climate Prediction Center

(CPC), Washington (CPC 2001, 2002), shows themonthly standardised anomaly of outgoing long waveradiation (OLR) from January 1998 to February 2002,together with a three-month moving average. Thesedata, compiled by the CPC, are a measure of theamount of long wave radiation emitted from the equa-torial region centred about the date-line (5ºS to 5ºNand 160ºE to 160ºW). Tropical deep convection inthis region is particularly sensitive to changes in thephase of the Southern Oscillation. During warm (ElNiño) ENSO events, convection is generally more

64 Australian Meteorological Magazine 52:1 March 2003

*The Troup Southern Oscillation Index (SOI) used in this article isten times the standardised monthly anomaly of the difference inmean sea-level pressure between Tahiti and Darwin. The calculationis based on a sixty-year climatology (1933-1992).

Fig. 1 Southern Oscillation Index, from January1998 to February 2002. Means and standarddeviations used in the computation of the SOIare based on the period 1933-1992.

-30

-20

-10

0

10

20

Jan Jan Jan Jan Jan| | | | |

1998 1999 2000 2001

Monthly SOI5-month weighted average

Fig. 2 Standardised anomaly of monthly outgoinglong-wave radiation averaged over the area5ºS to 5ºN and 160ºE to 160ºW, from January1998 to February 2002. Negative (positive)anomalies indicate enhanced (reduced) con-vection and rainfall in the area. Anomalies arebased on the 1979-1995 base period. AfterCPC (2001, 2002).

-3

-2

-1

0

1

2

3

1998 1999 2000 2001 2002| | | | |

Monthly 3-month weighted average

OLR Anomaly from 5N-5S & 160E-160W(after CPC)

prevalent resulting in a reduction in OLR. This reduc-tion is due to the lower effective black-body tempera-ture and is associated with increased high cloud anddeep convection. The reverse applies in cold (LaNiña) events, with generally less convection than nor-mal in the region.

The monthly values –1.5 (December), +0.2(January) and –1.1 (February) of this index showed adecline in the OLR with predominantly negative val-ues, December’s being the most negative valueobserved since October 1997, during the last El Niñoevent. The anomalously low OLR values observedwould appear to be related to an active phase of theMadden Julian Oscillation (MJO), which movedacross the equatorial Pacific during late Novemberand December, reaching just east of the date-line bythe end of the month.

A large area of below average OLR (i.e., increasedhigh cloudiness) was observed along the equator, eastof Borneo (120°E), to the date-line during December.Below normal OLR values extended, though withsmaller magnitude, right across the southwest Pacific,suggesting a shifted and more intense South Pacificconvergence zone (SPCZ) than normal. This patternis consistent with the MJO event described above. Astrong positive OLR anomaly in December over thetropical Indian Ocean alludes to a weaker monsoontrough in early summer. The passage of the MJO isalso consistent with the OLR anomalies observednorth of Australia during January. Decreased highcloud is observed for most regions north of the conti-

nent, whilst increased high cloud is apparent in theIndian Ocean, south of the Bay of Bengal. ByFebruary, negative OLR anomalies are again presentalong the equator west of the date-line. Similarly, neg-ative anomalies (increased high cloud) were presentin a band across central Australia, stretching fromsouth of the Kimberley region to the New SouthWales coast. This band was at least in part related totropical cyclone Chris, which briefly reachedCategory 5 intensity with wind gusts to 290 km/h,crossing the coast 160 km northeast of Port Hedland(Western Australia) on 6 February, bringing somerecord rainfall totals and local flooding. The band wasalso related to good rains in parts of New South Walesduring February, and contributed significantly to therains in central and western inland Australia (asshown in Fig. 15).

Oceanic patternsSea-surface temperaturesFigure 3 shows the summer 2001/02 sea-surface tem-perature (SST) anomaly in degrees Celsius (°C). Thecontour interval is 0.5°C. Positive anomalies areshown in orange and red shades, while negativeanomalies are shown in blue shades.

Summer 2001/02 sea-surface temperaturesremained above average in the western equatorialPacific, with a band of anomalies along the equator0.5°C above average between 160°E and 160°W. This

Gamble: Southern hemisphere climate summary summer 2001/02 65

Fig. 3 Anomalies of sea-surface temperature for summer (December, January, February) 2001/02 (ºC). The contourinterval is 0.5°C.

was reflected in the monthly NINO4 SST index val-ues, as calculated by the National Meteorological andOceanographic Centre (NMOC), Melbourne, whichwere +0.215°C (December), +0.567°C (January) and+0.492°C (February). These values arrested a pre-dominantly downward trend that occurred in spring2001 (Trewin and Fawcett 2002).

In the eastern equatorial Pacific, sea-surface tem-peratures did not vary far from average, with onlyslight negative anomalies just south of the equatorand slight positive anomalies just to the north. TheNMOC NINO3 SST index values for the summermonths were weakly negative.

The Indian Ocean was mostly warmer than aver-age apart from a small area of negative anomalies at30°S. Most areas were up to 0.5°C to 1.0°C aboveaverage, however there were also some small patcheswhose anomaly exceeded +1.0°C. The positive anom-alies were stronger than those observed in spring2001 (Trewin and Fawcett 2002) but weaker and lesswidespread than those in winter 2001 (Fawcett andWatkins 2002).

Cooler than average sea-surface temperatures,resulting in anomalies in the range from –0.5°C to–1.0°C, were observed in the Great Australian Bightand extending around the southeast coastline to 30°Son the east Australian coast. North of this however,the SSTs were above average stretching northwards tojoin the positive anomalies in the western equatorialPacific.

The negative anomalies observed throughout 2001in a band around Antarctica between 40°S to 60°Scontinued, with the only positive anomalies occurringbetween 50°W and 10°E. The strength of the negativeanomalies increased again after weakening in spring2001 (Trewin and Fawcett 2002) between 140°W and70°W. Wind anomalies in this region suggest thecooler SSTs may be related to cold Antarctic conti-nental air drawn up by stronger than average south-westerlies. This corresponds with a low in the geopo-tential anomalies (see Figs 8 and 9) also implying acolder than average air mass in the region of stronglynegative SST anomalies. A low is also evident in thisarea in the MSLP analysis and MSLP anomalies (seeFigs 6 and 7).

Subsurface patternsFigure 4 shows a time-longitude diagram of theanomaly in metres of the depth of the 20°C isothermalong the equatorial Pacific Ocean between January1996 and February 2002, as calculated by the Bureauof Meteorology Research Centre (BMRC). Thisisotherm is generally situated very close to the equa-torial ocean thermocline, the region of greatest tem-perature gradient with respect to depth. The thermo-

cline can also be regarded as the boundary betweenthe upper ocean warm water and the deeper oceancold water. An abnormally shallow thermocline in theeastern Pacific Ocean is characteristic of La Niñaevents. Positive anomalies correspond to the 20°Cisotherm being deeper than average, and negativeanomalies to it being shallower than average.

The main feature observed in the summer 2001/0220°C isotherm depth was a significant deepening ofthe thermocline (warmer than average subsurfacetemperatures) and a shift in the anomalies eastwardsfrom the central Pacific, around 150°E towards the fareastern Pacific. This was a clear illustration of thepropagation of a down-welling oceanic Kelvin wave,a result of a significant westerly wind burst (itselfassociated with the November/December MJO event)observed in December 2001-January 2002. The pat-tern was very similar to the Kelvin wave initiated inDecember 1996 (prior to the 1997/98 El Niño),although not of the same intensity.

66 Australian Meteorological Magazine 52:1 March 2003

Fig. 4 Time-longitude section of the monthly anom-alous depth of the 20°C isotherm at the equa-tor from January 1996 to February 2002. Thecontour interval is 10 m.

Figure 5 shows a sequence of equatorial Pacificvertical temperature anomaly profiles for the fourmonths ending February 2002, also obtained fromthe BMRC. In the figure, red (blue) shades indicatesubsurface waters which are warmer (cooler) thanaverage. December and January showed significantwarming (of up to 4°C) in many locations, and aneastward propagating trend in the anomaliesbetween 150°E and 140°W, however this weakenedsomewhat in February as oceanic and atmosphericconditions failed to become truly coupled. Thecross-section for December also shows a cooling ofthe subsurface temperatures in the eastern Pacific,between 150°W and 90°W, however the cool anom-alies were short lived, returning towards normal inJanuary and February.

Atmospheric patternsSurface analysesThe summer 2001/02 mean sea-level pressure(MSLP) across the southern hemisphere is shown inFig. 6, with the associated anomalies shown in Fig. 7.These anomalies are the departures from an eleven-year (1979-1989) climatology obtained from theEuropean Centre for Medium-range WeatherForecasts (ECMWF). The MSLP analysis itself hasbeen computed using data obtained from the Bureauof Meteorology’s Global Assimilation and Prediction(GASP) model daily 2300 UTC analyses.

The low pressure anomaly over high elevationparts of South America and the high pressure anom-aly over the Antarctic plateau are persistent features,and appear to be the result of a systematic differencebetween the GASP and ECMWF models and the waythey extrapolate pressure values down to mean sealevel. Therefore, the indicated low pressure over partsof South America should not be treated as a meteoro-logical consequence of the state of the El Niño-Southern Oscillation.

The Antarctic circumpolar trough over the summermonths was stronger and narrower than average withnegative anomalies extending around the whole ofAntarctica. Two substantial minima located at 90°Wand 20°E were observed in the pressure pattern, bothbeing approximately 6 hPa below average. A strongerand more southerly subtropical ridge with two anom-alous high centres, one southwest of Australia and theother southeast of New Zealand, being 8 hPa aboveaverage, forced the decrease in trough width, particu-larly in eastern latitudes.

The positive anomaly southeast of New Zealandand a negative anomaly over the Tasman Sea are con-sistent with the stronger than average Blocking Index

(see Fig. 10) also observed in this region. A broad areaof pressures 1 to 2 hPa below average was observedin the Indian Ocean.

Over Australia, MSLP values were generallyslightly above average except along the east coastwhere a negative anomaly extending from the lowover the Tasman Sea decreased pressures. MSL pres-sures in the equatorial Pacific region were generallybelow average except over Papua New Guinea andIndonesia where no major deviation from averagewas evident.

Mid-tropospheric analysesThe mean 500 hPa geopotential height patterns forsummer 2001/02 are shown in Fig. 8, with anomaliesshown in Fig. 9. The geopotential flow was largelyzonal with no distinct troughs, although a 3 to 4 wavepattern is vaguely evident. The 500 hPa geopotentialheights were broadly consistent with the MSLP pat-tern, however the negative anomaly in the TasmanSea spread a lot further west over southern Australiaat this level, corresponding to the region of belowaverage temperatures (Figs 16 and 17).

There are subtle splits in the mid-latitudinal flowover New Zealand, South America and in the south-ern Indian Ocean at around 75°E consistent with theBlocking Index in these regions.

BlockingFigure 10 is a time-longitude section of the dailysouthern hemisphere mid-level Blocking Index (BI),

Gamble: Southern hemisphere climate summary summer 2001/02 67

Fig. 5 Four month November 2001 to February 2002sequence of vertical temperature anomalies atthe equator for the Pacific Ocean. The contourinterval is 0.5°C.

BI=1/2[(u25+u30) – (u40+2u45+u50) + (u55+u60)].

Here, uλ indicates the 500 hPa level zonal wind com-ponent at λ degrees of southern hemisphere latituderanging from 0° at the equator to +90°S at the SouthPole. The BI measures the strength of the 500 hPaflow at the mid-latitudes (40°S to 50°S) relative tothat at subtropical (25°S to 30°S) and high (55°S to60°S) latitudes.

Taken across the entire season in the form of a sea-sonal mean (Fig. 11), southern hemisphere blocking

was above average for almost all longitudes, theexception being between 80°W to 100°W where therewas a strong zonal flow at high latitudes. Positiveanomalies in the blocking index were strongest inJanuary and February in the Australian region. A peakin the blocking index in the New Zealand sector wasa result of stronger than normal anticyclones locatedfurther south than the climatological average shownin the MSLP anomaly (Fig. 7) and also in the split midlevel pattern at 40°S depicted in the mean geopoten-tial height (Fig. 8).

68 Australian Meteorological Magazine 52:1 March 2003

Fig. 7 Summer 2001/02 mean sea-level pressureanomaly (hPa). The contours are spaced at 2hPa intervals between –12 hPa and +8 hPa.

Fig. 8 Summer 2001/02 500 hPa mean geopotentialheight (m). The contours are spaced at 80geopotential metre intervals from 5040 gpm to5840 gpm.

Fig. 9 Summer 2001/02 500 hPa mean geopotentialheight anomaly (m). The contours are spacedat 40 gpm intervals from –80 gpm to +80 gpm.

Fig. 6 Summer 2001/02 mean sea-level pressure(hPa). The contours are spaced at 4 hPa inter-vals between 976 hPa and 1020 hPa.

WindsLow-level (850 hPa) and upper-level (200 hPa) windanomalies for summer 2001/02 are shown in Figs 12and 13 respectively. The low-level summer windanomalies (Fig. 12) showed a marked cyclonic circu-

lation over the Tasman Sea with the upper levelsshowing a similar form with strong anomalous west-erlies over southern Australia and easterlies to thesouth of the continent. This pattern is consistent withthe strong BI recorded in this region and may also beassociated with the below average temperaturesrecorded over southern Australia. The cyclonic anom-alies observed are in stark contrast to summer2000/01 (Fawcett 2002), when strong anticyclonicanomalies were observed in this region. As a result ofthe cyclonic anomalies, low-level flow over southernQueensland and northern New South Wales showedanomalous westerly characteristics implying a weak-ening of the southeast trade winds.

Particularly strong anomalous westerlies were alsorecorded in the upper levels over the eastern equator-ial Pacific. These winds, in association with the east-erly anomalies at low levels in the central Pacific,pointed to a marginally stronger Walker Circulationdespite weak and varied anomalies at low levels in thewestern equatorial Pacific. Low-level anomalouswesterlies north of Australia, a result of westerly windbursts in December and January, generated zonal con-vergence near the date-line and contributed to thepropagation of the oceanic Kelvin wave seen in Fig.4.

In the tropical Indian Ocean, areas of moderateconvergence in the lower levels were observed withcorresponding divergence in the upper levels, point-ing at a slightly stronger than normal monsoonaltrough just south of the equator. An area of stronglow-level anomalous northwesterly winds in thenorthwestern tropical Indian Ocean was observed,with a broader region of anomalous southeasterlyreturn flow in the upper levels. A strong anticyclonicanomaly was located southwest of Australia reflectingweaker than normal mid-latitude westerlies.

Australian regionRainfallFigure 14 shows the summer rainfall totals forAustralia, while Fig. 15 shows the summer rainfalldeciles, where the deciles are calculated with respectto gridded rainfall data for all summers from 1900/01to 2001/02.

Summer rainfall was above average through south-ern parts of the Northern Territory, eastern WesternAustralia and the northwestern half of South Australia,with a substantial area of highest-on-record rainfallrecorded near the tri-State border of Western Australia,South Australia and the Northern Territory. This waslargely attributed to moist air masses accompanyingactive low pressure systems, generally embedded in the

Gamble: Southern hemisphere climate summary summer 2001/02 69

Fig. 10 Summer 2001/02 daily blocking index: time-longitude section. The horizontal axis mea-sures degrees of longitude east of theGreenwich meridian. Day one is 1 December.

Fig. 11 Mean southern hemisphere blocking index forsummer 2001/02 (bold line). The dashed lineshows the corresponding long-term average.The horizontal axis shows degrees east of theGreenwich meridian.

WA heat trough or the monsoon trough extendingsouthwards from the tropics. The majority of this rain-fall fell during December and February, January bring-ing mostly average rainfall across Australia. Above towell above average rainfall on the southwest tip ofWestern Australia resulted from the development of a

low in that region during December.Lowest-on-record rainfall occurred on the mid-

western coastal region near Carnarvon, with areas ofbelow to well below average rainfall being recordedacross southeast Queensland, northern New SouthWales and extending over northwest Victoria into

Fig. 12 Summer 2001/02 850 hPa vector wind anomalies (m s–1).

Fig. 13 Summer 2001/02 200 hPa vector wind anomalies (m s–1).

70 Australian Meteorological Magazine 52:1 March 2003

southeast South Australia. These dry conditions werelargely due to the maintenance of a ridge over south-east Australia for most of this period. February how-ever saw this region, particularly near the east coast,receiving very much above average rainfall resultingfrom troughlines extending down from the monsoon-al trough and interacting with cold fronts movingacross the Great Australian Bight.

Although experiencing a slow start to the wet sea-son, tropical cyclone Bernie in January, as well as thestrong presence of the monsoon trough in February,brought northern Northern Territory rainfall figuresback to near normal over the summer.

TemperaturesFigures 16 and 17 show the maximum and minimumtemperature anomalies, respectively, for summer2001. The anomalies have been calculated with

respect to the 1961-1990 period.Summer maximum temperatures were below

average over most of the country apart from thenortheast quadrant of the continent and the northern-most tips of Western Australia and the NorthernTerritory. Anomalies of between +1°C to +2°C, withpeaks of up to +4°C, were recorded over most ofQueensland and northeast New South Wales. Theseabove average maximum temperatures were largelydue to warm conditions in December and February,with anomalies of up to +5°C recorded in centraleastern Queensland during December. A warm, drycontinental airmass was pushed towards theQueensland coastline from inland areas with theincreased northwesterly wind anomalies (see Fig.12), and was likely to have caused the positiveanomalies. Summer maximum temperatures werethe highest of the post-1950 period for much of the

Gamble: Southern hemisphere climate summary summer 2001/02 71

Fig. 14 Summer 2001/02 rainfall totals (mm) inAustralia.

Fig. 15 Summer 2001/02 rainfall in Australia: decilerange values based on grid-point values overthe summers 1900/01 to 2001/02.

Fig. 16 Summer 2001/02 maximum temperatureanomalies (°C) for Australia based on a 1961-1990 mean.

Fig. 17 Summer 2001/02 minimum temperatureanomalies (°C) for Australia based on a 1961-1990 mean.

southern to central east coast of Queensland.A large area of negative anomalies in the range of

–2°C to –3°C was recorded over the central and west-ern interior extending down through most of SouthAustralia and southern Victoria. Similarly large nega-tive anomalies were also recorded in eastern centraland southern coastal, Western Australia, reaching 4°Cbelow normal. These strongly negative anomalieswere largely the result of cooler than average maxi-mum temperatures in December and February (bothpositive and negative anomalies being less extreme inJanuary) with a small patch of –6°C anomalies in cen-tral Western Australia during February. These anom-alies were a continuation of cooler than average tem-peratures through the central and western interiorobserved since October 2001. As a result, Victoria andSouth Australia recorded their lowest areally-aver-aged summer maximum temperatures of the post-1950 period, and Tasmania and Western Australiatheir third lowest.

Minimum temperature anomalies over summer2001/02 followed a similar pattern to the maxima,although of lesser magnitude. Anomalies from –1°Cto –2°C were recorded in an area stretching from cen-tral and southern Western Australia through southernparts of the Northern Territory and most of SouthAustralia into western Victoria, with peak anomaliesreaching –3°C in some small patches. Central easternQueensland recorded anomalies of +1°C to +2°C,with a small area of +2°C to +3°C just inland fromRockhampton. Areally-averaged minimum tempera-tures over summer 2001/02 were the lowest of thepost-1950 period for South Australia and second low-est for Western Australia.

As seen in the MSLP analysis (Fig. 6), the domi-nant high located in the Great Australian Bight overthe summer months generated a steady, cool andmoist southeast to southwesterly airstream over mostof the southern parts of the continent, most likely con-tributing to the lower than average temperatures.

AcknowledgmentThe author gratefully acknowledges Dr RobertFawcett and Dr Andrew Watkins, who provided use-ful comments and guidance for this study.

ReferencesClimate Prediction Center 2001, 2002. Climate Diagnostics Bulletin,

December 2001, January 2002, February 2002. US Departmentof Commerce, National Oceanic and AtmosphericAdministration, Washington D.C.

Commonwealth Bureau of Meteorology 2001, 2002. ClimateMonitoring Bulletin, December 2001, January and February2002 issues. National Climate Centre, Bur. Met., Australia.

Fawcett, R.J.B. 2002. Seasonal climate summary southern hemi-sphere (summer 2000/01): a third successive positive phase ofthe Southern Oscillation continues. Aust. Met. Mag., 51, 49-57.

Fawcett, R.J.B. and Watkins, A.B. 2002. Seasonal climate summarysouthern hemisphere (winter 2001): near-normal conditions inthe tropical Pacific continue. Aust. Met. Mag., 51, 181-90.

Pahalad, J. 2002. Seasonal climate summary southern hemisphere(autumn 2001): a return to near-normal conditions in the tropicalPacific. Aust. Met. Mag., 51, 117-24.

Trewin, B.C. and Fawcett, R.J.B. 2002. Seasonal climate summarysouthern hemisphere (spring 2001): persistence of near-normalconditions in the tropical Pacific. Aust. Met. Mag., 51, 251-60.

AppendixData sources used for this review were:• National Climate Centre, Climate Monitoring

Bulletin - Australia. Obtainable from the NationalClimate Centre, Bureau of Meteorology, GPO Box1289K, Melbourne, Vic. 3001, Australia.

• Climate Prediction Center (CPC), ClimateDiagnostics Bulletin. Obtainable from the ClimatePrediction Center (CPC), National WeatherService, Washington D.C., 20233, USA.

72 Australian Meteorological Magazine 52:1 March 2003