57 (2008) 255-268 Seasonal climate summary southern ...

Aust. Met. Mag. 57 (2008) 255-268

255

IntroductionThe latter half of 2006 brought extremely dry and very warm conditions to Australia, the result of an El Niño event in the equatorial Pacific. However there was much optimism for a return to more favourable

conditions for Australia when 2007 commenced with a cooling trend in the Pacific Ocean and a warming trend in the eastern Indian Ocean, and hence the possibility of a La Niña event later in the year (Braganza 2008). However the cooling in the Pacific halted as westerly wind-bursts in the western Pacific during April, May and July weakened trade winds, and effectively reset the clock for La Niña development (Bettio 2008).

Seasonal climate summary southern hemisphere (spring 2007): La Niña

finally takes holdPandora Hope

Centre for Australian Weather and Climate Research – A partnership between the Australian Bureau of Meteorology and CSIRO, Australia

and

Andrew B. WatkinsNational Climate Centre, Bureau of Meteorology, Australia

(Manuscript received September 2007)

Atmospheric and oceanic conditions in the southern hemi-sphere are reviewed for the austral spring of 2007. Particular emphasis is given to the Australian and Pacific regions. The transition from El Niño conditions in 2006 to La Niña condi-tions in 2007 took approximately ten months, arriving in ear-ly to mid-spring. Sea-surface temperatures in the central to eastern equatorial Pacific were generally cool for the season, as was the subsurface, while the Southern Oscillation Index gradually rose, reaching near one standard deviation above normal by season’s end. Despite the La Niña conditions, mean temperatures remained warm over Australia, with spring 2007 being the third warmest since high-quality records com-menced in 1950. Rainfall was very low early in the season, but by November central and eastern Australia received above-average totals. This rainfall did little to alleviate the six-year drought for the agriculturally vital Murray-Darling Basin. In the Antarctic, the sea-ice area of 15.2 million km2 was the larg-est area observed for any month since the sea-ice pack was first observed in its entirety by satellites in 1979, and appeared consistent with changing circulation patterns. The Antarctic ozone hole reached an area of 25.1 million km2; considerably less than the record 2006 area but still well above the 1979-2006 mean peak value of just under 21 million km2.

Corresponding author address: Pandora Hope, Centre for Australian Weather and Climate Research, Australian Bureau of Meteorology, GPO Box 1289, Melbourne, Vic. 3001, Australia.Email: [email protected]

256 Australian Meteorological Magazine 57:3 September 2008

As a result, it was not until September that weakly cool conditions in the equatorial Pacific subsurface, and then surface, fed back into the atmosphere. Hence a truly coupled La Niña event was finally declared in October. It was still not until November that much of central and eastern Australia received the much anticipated above-average rainfall, and temperatures, at least for northern Australia, cooled to near or below average. This summary reviews the southern hemisphere and equatorial climate patterns for spring (the calen-dar months of September, October, November) 2007, with particular attention to the Australasian and Pacif-ic regions. The main sources of information used for this summary are analyses prepared by the Bureau of Meteorology’s National Climate Centre, the Bureau’s Monthly Significant Weather Summaries, and the Cli-mate Diagnostics Bulletin (Climate Prediction Center, Washington D.C., USA). Information on the Antarctic ozone hole, included in the spring seasonal summary in previous years, is provided in a separate paper (Tully et al. 2008).

Pacific basin climate The Southern Oscillation The weakly positive Troup Southern Oscillation In-dex (SOI*) anomalies at the end of winter (Fig. 1) persisted and increased through the austral spring. In September 2007 the SOI value was +1.5, in October it rose to +5.4 and in November it reached +9.8, close to one standard deviation above the mean (the Troup SOI has a standard deviation of 10). The resulting spring SOI value was +5.6 (Fig. 1). Throughout the season, running 30-day SOI values were always within one standard deviation of the mean. Although these anomalies are not large, they reflect a coupling of the atmosphere with the previously cool eastern Pacific Ocean indicating the progress towards a fully coupled La Niña. A La Niña was declared by the Australian Bureau of Meteorology in October. Mean sea-level pressure (MSLP) anomalies at Tahiti and Darwin were generally close to their long-term mean during September. Weak positive anoma-lies at Tahiti in early October and stronger negative anomalies at Darwin in November contributed to the rising values of the SOI. The shift towards a La Niña in the tropical atmo-sphere, as suggested by the values of the SOI, was sup-ported by the Climate Diagnostics Center (CDC) Mul-tivariate ENSO Index (MEI) (Wolter and Timlin 1993, 1998), an index derived from a number of atmospheric

and oceanic parameters typically associated with El Niño and La Niña. Negative or low-ranked values of the MEI indicate cooler (La Niña-like) conditions and positive, high-ranked values indicate warmer (El Niño-like) conditions. The MEI ranked values during the season dropped from 10 to 8 (out of 58 years); a drop in rankings from those observed during winter (19 and 16). An MEI ranking of 1 to 8 is considered to be a strong La Niña event.

Tropical convection After sunlight is absorbed and re-radiated from the Earth’s surface, it is termed long-wave radiation. Neg-ative (positive) anomalies of outgoing long-wave radi-ation (OLR) suggest cooler (warmer) black-body tem-peratures, which tend to be associated with an increase (decrease) in high cloud amount and a good proxy for tropical convection. (Negative anomalies of OLR may also signal increased rainfall.) Studies have shown that during El Niño events, OLR is generally reduced (i.e. convection is generally enhanced) along the equato-rial Pacific, particularly near and east of the date-line. During La Niña events, OLR is often increased (i.e., convection is often suppressed) over the same region (Hoerling et al. 1997; Vincent et al. 1998). The time series of monthly standardised OLR anomalies for the region centred on the date-line (5ºN to 5ºS, 160ºE to 160ºW) is shown in Fig. 2. Through-out austral winter 2007 OLR anomalies were weakly positive. The OLR was also positive throughout spring 2007, with values in October rising above one standard deviation. The positive values of the OLR anomaly suggest decreased convection around the date-line, a pattern indicative of a La Niña event. Figure 3 shows the spatial pattern of OLR anomalies observed during the season. This shows that the region

Fig. 1 Southern Oscillation Index, January 2003 to November 2007 inclusive. Means and standard deviations based on the period 1933-92.

* The SOI used here is ten times the monthly anomaly of the differ-ence in mean sea-level pressure between Tahiti and Darwin, divided by the standard deviation of that difference for the relevant month, based on the period 1933-92.

Hope and Watkins: Seasonal climate summary southern hemisphere (spring 2007) 257

of increased OLR (suppressed convection) extended far beyond the region closest to the date-line. The area of suppressed convection extended east, north of the equator, to the South American coast. In the far west-ern equatorial Pacific, decreased OLR (increased con-vection) was apparent. The South Pacific convergence zone (SPCZ) was generally south of its climatological position. (During El Niño events the SPCZ shifts north and east of its average position, while during La Niña events it tends to be shifted south and west (Vincent 1994).) The observed pattern of OLR anomalies is typical of a La Niña pattern in the Pacific.

In the eastern Indian Ocean, which would usually display negative OLR anomalies accompanying those in the western Pacific, positive anomalies persisted, suggesting suppressed convection and the weakly negative eastern pole of the Indian Ocean Dipole pattern (see following section). Besides Papua New Guinea, parts of Indonesia, and equatorial western Africa, all land areas had positive anomalies of OLR. As described later, the enhanced OLR over land was generally associated with dry conditions. Winter 2007 brought only a weak Madden-Julian Oscillation (MJO) (Madden and Julian 1971, 1972) event in July, with no activity in August nor in the fol-lowing September or October (Gottschalk and Bell 2008), though a period of anomalous westerlies in the western Pacific during late September did spark a weak oceanic Kelvin wave. In November, however, an MJO event propagated from the Indian Ocean into the Pacific, resulting in weak westerly 850 hPa wind anomalies south of the equator, west of the date-line. This westerly wind-burst triggered a downwelling oceanic Kelvin wave, which crossed the equatorial Pacific during late November and December, leading to some, but not total, erosion of the cool La Niña-associated anomalies below the surface.

Ocean patternsSea-surface temperature (SST) patternsThe spring 2007 SST anomaly distribution (Fig. 4) displayed a classical La Niña pattern in the eastern and central Pacific. Cool anomalies of up to -3°C

Fig. 2 Standardised anomaly of monthly outgoing long-wave radiation averaged over 5°N-5°S and 160°E-160°W, from January 2003 to November 2007. Negative (positive) anomalies indicate enhanced (reduced) convection and rainfall. Anomalies are based on a 1979-95 base period. After CPC (2007).

Fig. 3 Anomalies of outgoing long-wave radiation for spring 2007 (Wm–2), based on a 1979-98 base period.


were evident east of about 160°E along the equator, extending down the South American coast. The ma-jority of the rest of the South Pacific Ocean showed warm anomalies. Around New Zealand and off the northeast coast of Australia, weak negative anomalies were evident, the latter being somewhat atypical for a La Niña event. Warm SST anomalies also extended around south-ern Africa, with values up to +1.5°C. Similar values were also seen along the east coast of South America centred at about 30°S. Niño indices were all weakly negative through win-ter 2007, with the exception of Niño4. During spring, all Niño indices, including Niño4, were negative, with increasingly large magnitude anomalies. These indi-ces of ocean temperature, combined with the posi-tive SOI and low ranked MEI, indicated that a fully coupled La Niña event had finally emerged during the austral spring, after several months of promising con-ditions. Indeed, throughout spring, the La Niña pat-tern strengthened in the Pacific Ocean with growing negative anomalies in the eastern equatorial region. However, towards the end of the season the broad area of cool anomalies that extended out from the South American west coast began to be eroded. In the Indian Ocean, equatorial SST anomalies were generally warm in the central to western regions in early September, but eased towards normal through-out the rest of the season. In the eastern Indian Ocean, negative SST anoma-lies to the north of Australia in late June and July, which coincided with an exceptional cold outbreak over northern Australia centred on 17-22 June (NCC 2007), moved to the northwest during August and Sep-tember. The resulting weak anomalies contributed to a cool eastern ‘pole’ of the Indian Ocean Dipole (IOD), and hence a positive IOD index for the spring, which brought with it reduced evaporation and convection in

this region, and thus a reduced influence from north-west cloudbands across northern and southeastern Aus-tralia (see later section: Australian region, rainfall). The movement of cool anomalies from north of Australia to south of Sumatra was not suggestive of the typical upwelling formation of cool anomalies in this region, and in turn suggests that this event was different in origin than classic cool episodes in this re-gion. Indeed, it has been noted that it is extremely rare for a positive IOD to occur during a La Niña period (Meyers et al. 2007), with the only prior occurrence since 1950 being August and September 1967. East-erly winds in the central equatorial Indian Ocean in May (e.g. Braganza 2008) were suggested by Behera et al. (2008) as generating an upwelling Kelvin wave, which later reinforced the cool anomalies along the southern coasts of Sumatra and Java and to the north-west of Australia. These weak cool anomalies persist-ed until the end of the season, when this area started to warm with the shift in the monsoon winds and the breakdown of the positive IOD event. Late in the season, in concert with the warming off the northwest coast of Australia, the east coast also saw warm anomalies, with much of the Tasman Sea displaying positive anomalies of up to +3°C. Further afield, by late September there was some warming evident off the east coast of South America, though during the season cool anomalies persisted around southern regions. Off the southern coast of Africa a localised warm anomaly persisted from mid- September until it moved to the southwest by mid- November and another, weaker area of warm water developed off the south coast. At the high latitudes (south of 60°S) SST values were generally within 0.5°C of their long-term mean, except for a region north of the Ross Sea, where cool anomalies reached values of greater than one degree cooler than the mean.

Fig. 4 Anomalies of sea-surface temperature for spring 2007 (°C). Contour interval is 0.5°C. Base period is 1961-90.


For the sea-ice itself, the individual months of Sep-tember and October (not shown) saw higher concen-trations, and an associated more northerly extent of the sea-ice edge, particularly north of the Ross Sea and Dronning Maud land (east of the Ross Sea), with a smaller region of anomalous northerly sea-ice off East Antarctica. By November, a rapid melt meant that the sea-ice edge was only further north than its climato-logical position equatorward of the Ross Sea. The most notable southern hemisphere sea-ice feature of 2007 was the record sea-ice area (defined as the total area covered by ice; areas of open water in between the floes are not included) (Massom et al. 2008). In Septem-ber 2007, the sea-ice area of 15.2 million km2 was the largest observed for any month since the sea-ice pack was first observed in its entirety by satellites in 1979, be-ing some six per cent above the long-term mean. Simi-larly, the sea-ice extent area (i.e. the total area, including leads, that lies within the 15 per cent sea-ice isopleth) of 19.2 million km2 was the second-largest area since satel-lite records began, being three per cent above the mean. This record sea-ice state appears to be the result of an anomalously strong wave number three pattern in the surface pressure field during September, a rela-tionship shown by Reid et al. (2008). They note that the climatological low-pressure centres near 20°E, 90°E and 150°W were unusually deep in September 2007, leading to anomalously strong equatorward sur-face movement and cold air advection on the broad western flanks of the regions outlined above and thus an expanded sea-ice pack in these regions. Reid et al. (2008) also noted the similarity between these areas of anomalously low surface pressures in 2007, and those areas which have experienced the largest (decreasing) trend in mean sea-level pressure since 1979 for the in-dividual months of July, August and September. Globally averaged, 2007 austral spring SSTs were 0.33°C above the 1880-2006 mean (NCDC 2007), the tenth-warmest September to November period on re-cord. In the southern hemisphere oceans were 0.27°C above average; the twenty-first warmest on record.

Subsurface ocean patternsThe Hovmöller diagram for the 20°C isotherm depth anomaly along the equator from January 2001 to No-vember 2007 is shown in Fig. 5. The 20°C isotherm depth is generally situated close to the equatorial ocean thermocline, the region of greatest temperature gradi-ent with depth and the boundary between the warm near-surface and cold deep-ocean water. Changes in the thermocline depth may act as a precursor to future changes at the surface. The negative 20°C isotherm depth anomalies in the eastern equatorial Pacific at the end of winter (Bet-tio 2008) continued and intensified through the spring, reaching almost −40 metres at times during October

and November. In the western equatorial Pacific, a downwelling oceanic Kelvin wave was triggered by a relaxation in the trade winds in the western Pacific in late September. This Kelvin wave briefly warmed areas near the thermocline in the central and western Pacific during October and early November, however this was not enough to completely erode the existing cool anomalies. In fact the rapid return to enhanced trade winds by mid to late October across the cen-tral equatorial Pacific resulted in enhanced upwell-ing which rapidly cooled the near-surface once more. This was further reflected in the westward spread of significantly cool anomalies at the surface by mid-November. This continued positive reinforcement, and thus the fully coupled nature of the Pacific ocean-atmosphere system, made it clear that La Niña condi-tions had become established during the season. Anomalies of ocean temperature along the equator throughout the ocean depth for the individual spring months are shown in Fig. 6. As described above, the upward displacement of the thermocline was evi-dent from the date-line to the South American coast throughout the season. However the region of maxi-mum anomaly values (of mid-range strength, around −3.5°C) near the thermocline earlier in the season ex-tended towards the surface in October.

Fig. 5 Time-longitude section of the monthly anoma-lous depth of the 20°C isotherm (m) at the equa-tor for January 2001 to November 2007. Base period: 1979-89. Contour interval is 10 m.


Overall, the cool subsurface anomalies in the cen-tral to eastern equatorial Pacific, combined with the average to slightly warm anomalies in the western Pacific, produced an anomaly gradient indicative of a maturing La Niña event.

Hemispheric surface pressure patterns The southern hemisphere spring 2007 MSLP pattern, computed from the Australian Bureau of Meteorol-ogy’s Global Assimilation and Prediction (GASP) model, is shown in Fig. 7, with the corresponding anomaly pattern shown in Fig. 8. These anomalies are the difference from a 22-year (1979-2000) climatology obtained from the National Center for Environmental Prediction (NCEP) II reanalysis data (Kanamitsu et al. 2002). It is difficult to accurately calculate MSLP over regions of high topography, and these are classed as missing data and are shaded grey in Fig. 8. The spring MSLP pattern shown in Fig. 7 was largely zonal in structure, overlaid with a wave num-ber three pattern. Seasonally averaged mid-latitude troughs were located at 30°E, 170ºW and 60ºW. As expected from climatology, the subtropical ridge and the Antarctic high were the dominant high pressure features, whilst the Antarctic circumpolar trough was the strongest low pressure feature. In terms of the southern annular mode (SAM), the seasonal mean

value of −0.46 (CPC 2008) indicated a slight equator-ward shift in the westerlies; however, the mean value disguises somewhat a decreasing trend during the sea-son, with the index falling from 0.031 to −0.434 and then −0.984 for the months of September, October and November respectively.

Fig. 6 Four-month, August to November 2007, se-quence of subsurface temperature anomalies at the equator. Contour interval is 0.5°C.

Fig. 7 Mean sea-level pressure for spring 2007 (hPa). The contour interval is 5 hPa.

Fig. 8 Anomalies of the mean sea-level pressure from the 1979-2000 National Centers for Environ-mental Prediction reanalysis II climatology, for spring 2007 (hPa). Contour interval is 2.5 hPa.


The MSLP anomaly distribution (Fig. 8) for spring 2007 shows the greatest anomalies at the mid to high latitudes, as may be expected owing to this region hav-ing the highest variability. In the Indian Ocean sector, spring 2007 MSLP was anomalously low, meaning that the circumpolar trough was deeper in this sector and the mid-latitude high weaker than usual. The high-pressure centres in the Pacific and Atlantic had higher MSLP values than usual, with positive anomalies of greater than +2.5 hPa. The Tasman Sea also had higher than usual MSLP. Between the Tasman Sea and the high in the eastern Pacific, a strong negative anomaly, extending from the Antarctic Circle to the mid-latitudes, result-ed in a slight northward extension of the circumpolar trough. Over Australia the MSLP anomaly was slightly positive, as reflected in the generally positive OLR values across eastern Australia over the season. Stron-ger negative anomalies were evident in the MSLP over South America and southern Africa. One notable synoptic aspect of spring 2007 was tropical cyclone Guba, the first named cyclone of the 2007-08 season, forming off Papua New Guinea on 13 November, reaching Category 3 on 16 November, and threatening Cape York Peninsula, before turning northwards and weakening without reaching the Aus-tralian coast. (There has never been a recorded landfall on the east coast in November.) Flooding in Oro Prov-ince of Papua New Guinea unfortunately resulted in a number of deaths.

Mid-tropospheric analysesThe 500 hPa geopotential height (an indicator of the steering of surface synoptic systems) across the south-ern hemisphere is shown in Fig. 9, and anomalies in the 500 hPa field are displayed in Fig. 10. The analysis of the 500 hPa height suggests a split in the flow in the Tasman Sea and in the eastern South Pacific. A further split was apparent to the east of Uruguay (South America).These splits reflect slightly higher pressures at the surface, associated with block-ing systems in these regions (Fig. 8). This is further discussed in the following section. The major 500 hPa height anomalies were gener-ally centred over the same locations as their MSLP counterparts. Combined, this suggests a largely equiv-alent barotropic atmospheric structure. However, there are some slight differences in the pattern of weaker anomalies, leading to a more zonal structure in the 500 hPa height anomalies than in the MSLP anoma-lies. The height anomalies formed a ring of low values around Antarctica, with weak positive anomalies to their north.

BlockingThe time-longitude section of the daily southern hemi-sphere blocking index (BI)* is shown in Fig. 11. This index is a measure of the strength of the zonal 500 hPa

* BI = 0.5[U25 + U30 – (U40 + 2U45 + U50) + U55 + U60] where Uxx is the westerly component of the 500 hPa wind at latitude xx°S.

Fig. 9 Mean 500 hPa geopotential heights for spring 2007 (gpm). The contour interval is 100 gpm.

Fig. 10 Anomalies of the 500 hPa geopotential height from the 1979-2000 National Center for Envi-ronmental Prediction reanalysis II climatology, for spring 2007 (gpm). The contour interval is 30 gpm.


flow in the mid-latitudes relative to that at subtropi-cal and high latitudes. Positive values of the block-ing index are generally associated with a split in the mid-latitude westerly flow centred near 45°S and mid-latitude blocking activity. Throughout September there were weak positive values of the BI in the Pacific Ocean sector, the region climatologically favoured for blocking (Trenberth and

Mo 1985; Sinclair 1996). The BI was greater than 60, representing a relatively strong blocking event off the west coast of South America in mid-September, then the blocking shifted over the continent later in the month. A blocking event was also evident just east of the dateline late in September. There was some fur-ther blocking around the dateline in early October, and some in the central Indian Ocean in mid-October. The latter blocking extended across Australian longitudes and lasted into early November, but there was little blocking after that. The seasonal mean blocking index is shown in Fig. 12. Although the time-longitude plot showed some events, the seasonal mean reflected little evidence of blocking during this season. It was positive only in the region of the Tasman Sea (150°E to 180°E) and just west of the coast of South America (280°E/80°W).

Low and upper-level windsSpring 2007 low-level (850 hPa) and upper-level (200 hPa) wind anomalies (from the 22-year NCEP II cli-matology) are shown in Figs 13 and 14 respectively. Regions where the surface rises above the level of the 850 hPa winds are shaded grey in Fig. 13. Generally, the low-level wind anomalies were weak, indicating that the circulation was close to the climatological average. Wind anomalies in the west-ern equatorial Pacific were westward, in line with the developing La Niña conditions, while northward anomalies were evident at 90°W, corresponding to a strengthened gyre in the eastern Pacific, as shown by the positive MSLP anomalies in this region (Fig. 8). Anomalous anticyclonic circulation was evident in the Tasman Sea, again associated with a high-pressure anomaly over this region during the season. Strength-ened easterlies to the northwest of Australia, a signal usually associated with El Niño conditions (Dros-dowsky and Williams 1991), may have been associ-ated with the anomalous cooling of the surface waters in this region, as described earlier. At high latitudes, the seasonal mean low-level wind showed an easterly anomaly close to the Ant-arctic coastline, with some anomalies as great as half the magnitude of the total wind field in that location (not shown). This suggests a weakening or northward shift of the circumpolar westerlies. The regions of low MSLP north of the Ross Sea and south of Africa are regions of enhanced cyclonic circulation in the 850 hPa wind anomalies. As discussed earlier, circulation associated with these unusually deep low-pressure centres led to anomalously strong equatorward surface movement and cold air advection on the broad western flanks of these regions and thus an expanded sea-ice pack, which reached a record area in September.

Fig. 11 Spring 2007 daily blocking index (m s-1) time-longitude section. The horizontal axis shows the degrees of longitude east of the Greenwich me-ridian. Day 1 is 1 September.

Fig. 12 Mean southern hemisphere blocking index for spring 2007 (solid line). The dashed line shows the corresponding long-term average. The horizontal axis shows the degrees east of the Greenwich meridian.


The upper-level wind anomalies (Fig. 14) clearly show the strengthened return flow of the Walker Cir-culation in the equatorial Pacific associated with La Niña conditions. At high latitudes, the 200 hPa wind anomalies reflect the anomalies at the surface, with cyclonic circulation north of the Ross Sea and south of Africa. This indicates an approximate equivalent barotropic structure at high latitudes.

Australian regionRainfallThe distribution of Australian rainfall totals for spring 2007 is shown in Fig. 15, whilst Fig. 16 shows the as-sociated decile ranges based on gridded rainfall data for all springs from 1900 to 2007. Highest rainfall totals for the season were recorded

Fig. 13 Spring 2007 850 hPa vector wind anomalies. The contour interval is 5 m s-1, with values above 5 m s-1 stippled.

Fig. 14 Spring 2007 200 hPa vector wind anomalies. The contour interval is 5 m s-1, with values above 5 m s-1 stippled.


in western Tasmania, where in excess of 400 mm fell for the season. On the mainland, the coast of northern New South Wales and southern Queensland received totals in excess of 200 mm. The southern tips of the continent and the wider southeast Queensland region all received rainfall totals greater than 100 mm. Along the east coast and in southeast Queensland precipitation was generally in decile 8 and 9 (above av-erage), with some parts of Queensland being very much above average (decile 10: i.e., rainfall totals in the high-est ten per cent of historical observations). Along the south coast however, spring 2007 rainfall totals were in the average (deciles 4-7) or below average (deciles 2-3) range. The near-average rainfall in the southwest of Western Australia continued the relatively flat spring-

time trend in this region over the last 50 years, despite strong drying trends in autumn and winter. Unlike the well-below average rainfall in spring 2006 (Qi 2007), the spring 2007 rainfall was generally close to average across much of northern and western Australia, with State averages in the middle tercile for all States except for Victoria, South Australia and Tas-mania, where rainfall was in the lowest tercile. Two notable regions where rainfall totals were very much below average (decile 1) for the season were in the vicinity of both Melbourne and Hobart water catchments (Fig. 16). Tasmania recorded its driest No-vember in 108 years of record. Despite generally average rainfall for the season as a whole, for the southern regions it was in fact a season of marked contrasts. For southeastern Austra-lia, September was exceptionally dry, with New South Wales recording the fourth driest September, South Australia the eighth driest, Victoria ninth driest and significantly, the Murray Darling Basin its driest-ever September in 108 years of record. Similarly, south-western Australia also recorded dry conditions for the first month of spring, with the region recording its fifth-driest September on record. These dry conditions (which for Victoria continued into October, which was its seventh driest on record) were countered by a relatively wet November. The Murray-Darling Basin recorded its tenth-wettest November in the 108 years of record, whilst New South Wales recorded its ninth wettest. Tasmania, however, did not experience such a reversal and endured its driest November on record, with a State average of only 19 mm; the first time Tas-mania had received less than 20 mm for that month. The increase in rainfall during the season, particu-larly across eastern Australia, paralleled the develop-ment of La Niña conditions in the Pacific Ocean, as expected from known relationships (McBride and Nicholls 1983). However the drier than expected condi-tions in the southeast early in the season may have had their origins in the cool anomaly to the northwest of the continent, and hence the broader positive IOD pattern (see; Nicholls 1989, Simmonds and Rocha 1991). The dry conditions of September and October for the southeast also marked the six-year anniversary of the Murray-Darling Basin drought. November 2001 to October 2007 was the equal driest such 72-month period on record with an average of only 389 mm/year, approximately 20 per cent below the long-term average of 480 mm, and equal to the previous record dry of 1939-45. Maximum temperatures over the 72 months were also at record levels, being some 1.3°C above the 1961-90 average, further increasing water demand. Furthermore, 2007 was the first time in the meteorological record dating from 1900 that an El Niño drought in the Murray-Darling Basin was not been followed by at least one three-month period with

Fig. 15 Rainfall totals over Australia for spring 2007 (mm).

Fig 16 Spring 2007 rainfall deciles for Australia: decile range based on grid-point values over the spring periods from 1900 to 2007.


above normal (Basin average) rainfall by the end of the following winter (Watkins and Trewin 2008). Similarly, October 2007 also marked the eleventh anniversary of dry conditions through southern Vic-toria, southeastern South Australia and northern and eastern Tasmania. The 11-year drought in Victoria was its driest such 11-year period since at least 1900. Southeastern Australia (including Tasmania) received just 6011 mm for the 11-year period (i.e., 546 mm/year) ending October 2007. When compared to the long-term average for this region of 6650 mm (605 mm/year), southeastern Australia had missed out on over one year of normal rainfall over the course of this long-term dry. Spring also marks the end of the April-October main crop and pasture growing season for much of southern Australia. Over the Grains Research and De-velopment Council’s (GRDC) cropping region it was 22nd driest season on record as a whole (221 mm), coming on the heels of 2006, which was sixth driest with only 162 mm. However, for the southern crop-ping zone it was 17th driest (212 mm; while 2006 was the fourth driest, 156 mm) and for the western region 11th driest with 219 mm (2006 was the third driest: 182 mm). For the Australian Bureau of Agricultural and Resource Economics (ABARE) pastoral zone, rainfall totals were 31st driest. While such rankings may sound reasonable, the late winter and early spring dry period came at a crucial time in Australia’s crop and pasture growing cycle and this, coupled with high temperatures later in the growing season, had a devas-tating impact upon crops, with the total crop estimated to be 38 per cent below the five-year average (ABARE 2007). Overall, spring 2007 was ranked 52nd out of the 108 years of Australia-wide records (i.e. near the me-dian), with a mean rainfall of 62.9 mm, far wetter than

spring 2006 (Qi 2007), which was the sixteenth driest on record with an Australia-wide mean rainfall of 43.6 mm. However, although 2007 was a La Niña year, it was drier than spring 2005 (Arblaster 2006), which had a mean rainfall of 95 mm. Table 1 summarises the seasonal rainfall extremes and rankings on a national and state basis.

TemperaturesThe spring 2007 mean maximum and minimum tem-perature anomalies, calculated with respect to the ref-erence period 1961-90, are shown in Figs 17 and 18 respectively. The Australia-wide mean temperature anomaly (relative to the 1961-90 mean) for spring 2007 was +1.06°C; the third-warmest spring in the high-quality Australian record (data since 1950). The warmest spring on record was in 2006 (+1.46°C), followed by 2002, both years associated with, albeit weak, El Niño events. For Victoria, virtually all ar-eas north of the Great Dividing Range recorded their warmest spring on record, resulting in a Victoria-wide mean spring temperature anomaly of +1.29°C, more than 0.3°C above the previous hottest spring for Vic-toria which occurred in 1961. Similarly, the mean southeastern Australian temperature was also hottest on record (+1.49°C), while the Murray-Darling Basin experienced its fourth-warmest spring on record, some 1.5°C above the 1961-90 average. High temperatures were not simply confined to the southeast; Western Australia recorded its second-warmest spring on re-cord in 2007 (+1.07°C) following its exceptionally warm spring in 2006 (+1.79°C). This warm mean temperature was the result of both warm maxima and minima: the Australia-wide mean maximum spring 2007 temperature anomaly was +1.38°C, the fourth highest on record, and the Australia-wide mean minimum spring 2007 tempera-

Table 1. Seasonal rainfall ranks and extremes on a national and State basis for spring 2007.

Highest seasonal Lowest seasonal Highest 24-hour Area-averaged Rank total (mm) total (mm) fall (mm) rainfall (AAR) of (mm) AAR *

Australia 930 at Mount Read (TAS) Zero at several locations 225 at Monvale (Qld) 6 September 64.7 53WA 373 at Karri Valley Zero at several locations 94 at Mornington 8 November 28.6 31NT 393 at Southport Zero at Walungurru 105 at Carmor Plains 5 November 65.0 61SA 230 at Mount Gambier 1 at Parakylia Station 96 at Braemar 24 October 34.9 32QLD 907 at Bellenden Ker 2 at Orientos Station 225 at Monavale 6 September 94.6 72 Top StationNSW 790 at Mount Seaview 15 at Tibooburra 167 at Port Macquarie 10 November 113.3 53VIC 554 at Weeaproinah 20 at Patho West 180 at Tanybryn 4 November 134.8 30TAS 930 at Mount Read 35 at Lisdillom 105 at Mt Read 15 October 246.0 23

* The rank goes from 1 (lowest) to 108 (highest) and is calculated for the years 1900 to 2007 inclusive.


ture anomaly was +0.74, making it Australia’s eighth- warmest spring. Only the northeast of the Australian continent displayed negative maximum temperature anomalies during the spring. This pattern of tempera-ture anomalies generally reflected a negative correla-tion with the rainfall – dry conditions across the south of the continent and wet in the northeast. Warm minimum temperature anomalies were also widespread across much of Australia, though the mag-nitude of the anomalies was far smaller than for the

maximum temperatures. In decile terms, however, much of southern Australia was well-above aver-age (deciles 8-10), with parts of central New South Wales and northeast South Australia being warmest on record. Although this may at first appear some-what contrary to what may have been expected given that monthly OLR anomalies (not shown) suggested cloudiness was suppressed over much of the continent during September and October, which would lead to greater night-time heat loss, it is consistent with the relatively larger increase in the magnitude of the maximum temperature anomalies, and hence a greater mean diurnal temperature range than normal. Table 2 summarises the seasonal maximum tem-perature ranks and extremes on a national and state basis, whilst Table 3 shows the corresponding sum-mary for the seasonal minimum temperatures.

AcknowledgmentsThe authors would like to thank Eunpa Lim, Matthew Wheeler and Tim Cowan for their valuable input and helpful suggestions with this manuscript.

ReferencesABARE 2007. Australian Crop Report, 144. sighted 26/08/2007

[http://www.abareconomics.com/interactive/cr_dec07/] Arblaster, J.M. 2006. Seasonal climate summary southern hemisphere

(spring 2005): a wet and warm season across much of Australia. Aust. Met. Mag., 55, 231-44

Behera, S.K., Luo, J.-J. and Yamagata, T. 2008. Unusu-al IOD event of 2007. Geophys. Res. Lett., 35, L14S11, doi:10.1029/2008GL034122.

Bettio, L. 2008. Seasonal climate summary southern hemisphere (winter 2007): A developing La Niña and a positive Indian Ocean Dipole. Aust. Met. Mag., 57, 155-66.

Braganza, K. 2008. Seasonal climate summary southern hemisphere (autumn 2007): La Niña emerges as a distinct possibility in 2007. Aust. Met. Mag., 57, 65-75.

Climate Prediction Center (CPC) 2007. Climate Diagnostics Bulletin. Climate Prediction Center, National Weather Service, Washington D.C., USA.

Climate Prediction Center (CPC) 2008. http://www.cpc.noaa.gov/products/precip/CWlink/daily_ao_index/aao/aao.shtml (sighted 10/09/2008).

Drosdowsky, W. and Williams, M. 1991. The Southern Oscillation in the Australian region. Part I: Anomalies at the extremes of the oscillation. Jnl Climate, 4, 619–38.

Gottschalck, J. and Bell, G.D. 2008. The Madden-Julian Oscillation, in State of the Climate in 2007. D.H. Levinson and J.H. Law-rimore eds. Bull. Am. Met. Soc., 89, S66-7.

Hoerling, M.P., Kumar, A. and Zhong, M. 1997. El Niño, La Niña, and the nonlinearity of their teleconnections. Jnl Climate, 10, 1769–86.

Kanamitsu, M., Ebisuzaki, W., Woollen, J., Yang, S.-K., Hnilo, J.J., Fiorino, M. and Potter, G.L. 2002. NCEP-DOE AMIP-II Re-anal-ysis (R-2), Bull. Am. Met. Soc., 83, 1631-43.

McBride, J.L. and Nicholls, N. 1983. Seasonal relationships between Australian rainfall and the southern oscillation. Mon. Weath. Rev.,

Fig. 17 Spring 2007 maximum temperature anomalies for Australia based on a 1961-90 mean. (°C).

Fig. 18 Spring 2007 minimum temperature anomalies for Australia based on a 1961-90 mean. (°C).


Table 2. Seasonal maximum temperature ranks and extremes on a national and State basis for spring 2007.

Highest Lowest Highest Lowest Anomaly of Rank seasonal seasonal daily daily area-averaged of mean (°C) mean (°C) recording (°C) recording (°C) mean (AAM) AAM* (°C)

Australia 40.4 at Fitzroy 8.3 at Mt Wellington 46.2 at Port Hedland −1.6 at Mt Wellington +1.38 55 Crossing (WA) (TAS) and at Roebourne (WA) (TAS) 3 September 20 NovemberWA 40.4 at Fitzroy 18.6 at Albany 46.2 at Port Hedland 10.6 at Manjimup +1.46 56 Crossing and at Roebourne 27 October 20 NovemberNT 38.7 at Timber 30.7 at McCluer 43.9 at Curtin Springs 16.5 at Curtin Springs +1.12 51 Creek Island 30 November 4 NovemberSA 31.8 at Oodnadatta 16.5 at Mt Lofty 43.5 at Marla 7.2 at Mt Lofty +2.15 57 30 November 11 SeptemberQLD 37.6 at Century 22.4 at Applethorpe 43.6 at Boulia 10.4 at Applethorpe +0.65 43 Mine 28 October 5 SeptemberNSW 30.0 at Mungindi 8.4 at Thredbo Top 45.0 at Ivanhoe −1.1 at Thredbo +1.94 55 Station 20 November Top Station 4 SeptemberVIC 26.2 at Mildura 8.4 at Mt Hotham 43.0 at Echuca −0.9 at Mt Baw Baw +2.08 57 and Kerang 3 September 20 NovemberTAS 19.2 at Bushy Park 8.3 at Mt Wellington 34.5 at Swansea −1.6 at Mt Wellington +1.15 58 20 November 3 September

* The temperature ranks go from 1 (lowest) to 58 (highest) and are calculated for the years 1950 to 2007 inclusive.

Table 3. Seasonal minimum temperature ranks and extremes on a national and State basis for spring 2007.

Highest Lowest Highest Lowest Anomaly of Rank seasonal seasonal daily daily area-averaged of mean (°C) mean (°C) recording (°C) recording (°C) mean (AAM) AAM* (°C)

Australia 26.2 at Troughton 0.9 at Thredbo 32.0 at Wittenoom −8.7 at Thredbo +0.74 51 Island (NT) Top Station (NSW) (WA) Top Station (NSW) and Mt Wellington 27 November 4 September (TAS)WA 26.2 at Troughton 6.4 at Wandering 32.0 at Wittenoom −2.0 at Southern Cross +0.67 52 Island 27 November 1 September NT 25.9 at McCluer 14.5 at Territory 30.2 at Ali Curung −0.9 at Arltunga +0.26 33 Island Grape Farm 21 November 6 September SA 16.4 at Marree 6.7 at Munkora 28.4 at Coober Pedy −3.1 at Murray Bridge +1.35 55 19 November 5 SeptemberQLD 24.8 at Horn Island 9.0 at Stanthorpe 29.4 at Boulia 0.5 at Stanthorpe +0.77 52 28 October 21 SeptemberNSW 16.1 at Yamba 0.9 at Thredbo 28.3 at Broken Hill −8.7 at Thredbo +1.03 54= Top Station 20 November Top Station 4 SeptemberVIC 11.8 at Gabo Island 1.7 at Mt Hotham 25.7 at Mildura −7.7 at Mt Hotham +0.51 52 20 November 4 SeptemberTAS 10.7 at Swan Island 0.9 at Mt Wellington 19.6 at Flinders Island −6.3 at Liawenee on +0.43 48= Airport 17 November 4 September and 23 October

* The temperature ranks go from 1 (lowest) to 58 (highest) and are calculated for the years 1950 to 2007 inclusive.


111, 1998-2004.Madden, R.A. and Julian, P.R. 1971. Detection of a 40-50 day oscil-

lation in the zonal wind in the tropical Pacific. J. Atmos. Sci., 28, 702-8.

Madden, R.A. and Julian, P.R. 1972. Description of global scale cir-culation cells in the tropics with a 40-50 day period. J. Atmos. Sci., 29, 1109-23.

Massom, R.A., Reid, P., Barreira, S. and Stammerjohn, S. 2008. Sea ice extent and concentration in State of the Climate in 2007. D.H. Levinson and J.H. Lawrimore (eds), Bull. Am. Met. Soc., 89, S103-4

Meyers, G., McIntosh, P., Pigot, L. and Pook, M. 2007. The years of El Niño, La Niña and interactions with the tropical Indian Ocean. J. Climatol., 20, 2872-80.

National Climate Centre (NCC) 2007. Exceptional cold and unsea-sonable rain in the Australian tropics, 17-22 June 2007, Special Climate Statement 13, Bur. Met., Australia.

National Climatic Data Center (NCDC) 2007. http://www.ncdc.noaa.gov/oa/climate/research/monitoring.html (sighted 20/08/2008).

Newman, P.A. 2008. Ozone depletion in State of the Climate in 2007, D.H Levinson. and J.H. Lawrimore eds. Bull. Am. Met. Soc., 89, S104-S105.

Newman, P.A., Nash, E.R. Kawa, S.R., Montzka, S.A. and Schauf-fler, S.M. 2006. When will the ozone hole recover? Geophys. Res. Lett., 33, L12814, doi:10.1029/2005GL025232.

Nicholls, N. 1989. Sea surface temperatures and Australian winter rainfall. J. Climatol., 2, 965-73.

Qi, L. 2007. Seasonal climate summary southern hemisphere (spring 2006): a weak El Niño in the tropical Pacific – warm and dry conditions in eastern and southern Australia. Aust Met. Mag., 56, 203-14.

Reid, P.A., Massom, R.A., Marsland, S.J., Zweck, C. 2008. Septem-ber 2007 - Circulation Driven Southern Hemisphere Sea Ice Area

Maximum, Polar Research - Arctic and Antarctic Perspectives in the International Polar Year, SCAR/IASC IPY Open Science Conference, St. Petersburg, Russia July 8th – 11th 2008 in asso-ciation with the XXX SCAR Meeting.

Simmonds, I. and Rocha, A. 1991. The association of Australian win-ter climate with ocean temperatures to the west. J. Climatol., 4, 1147-61.

Sinclair, M.R. 1996. A climatology of anticyclones and blocking for the Southern Hemisphere. Mon. Weath. Rev., 124, 245-63.

Trenberth, K. and Mo, K.C. 1985. Blocking in the southern hemi-sphere. Mon. Weath. Rev., 113, 3-21.

Tully, M.B., Klekociuk, A.R., Deschamps, L.L., Henderson, S.I., Krummel, P.B., Fraser, P.J., Shanklin, J.D., Downey, A.H., Gies, H.P. and Javorniczky, J. 2008. The 2007 Antarctic ozone hole. Aust. Met. Mag., 57, 279-98.

Vincent, D.G. 1994. The South Pacific Convergence Zone (SPCZ): A review. Mon. Weath. Rev., 122, 1949-70.

Vincent, D.G., Fink, A., Schrage, J.M. and Speth, P. 1998. High- and low- frequency intraseasonal variance of OLR on annual and ENSO timescales. Jnl Climate, 11, 968-86.

Watkins, A. and Trewin, B. 2008. Annual Climate Summary: 2007. Bull. Aust. Met. Ocean. Soc., 21, 8-16.

Wolter, K. and Timlin, M.S. 1993. Monitoring ENSO in COADS with a seasonally adjusted principal component index. Proc. of the 17th Climate Diagnostics Workshop, Norman, OK, NOAA/N MC/CAC, NSSL, Oklahoma Clim. Survey, CIMMS and the School of Meteor., Univ. of Oklahoma, 52-7.

Wolter, K. and Timlin, M.S. 1998. Measuring the strength of ENSO - how does 1997/98 rank? Weather, 53, 315-24.

World Meteorological Organization (WMO) 2007. WMO Antarctic Ozone Bulletins [available from: http://www.wmo.ch/pages/prog/arep/gawozobull07_en.html].

57 (2008) 255-268 Seasonal climate summary southern ...

Documents

Transcript of 57 (2008) 255-268 Seasonal climate summary southern ...