1. Introduction
a. Background
The South Pacific convergence zone (SPCZ) is an extensive Southern Hemisphere atmospheric circulation feature that contains one of Earth’s most expansive and persistent convective cloud bands (Vincent 1994). The SPCZ merges with the intertropical convergence zone (ITCZ) over the western equatorial Pacific warm pool (Trenberth 1991) and stretches along a zone orientated from northwest to southeast across the southwest Pacific (10°S, 160°E–30°S, 140°W; see Fig. 1). The morphology of this large-scale atmospheric circulation feature is controlled by the land–sea distribution in the western Pacific, and by the interplay of ocean and atmospheric circulation in the subtropics and midlatitudes of the Pacific (Kiladis et al. 1989; Takahashi and Battisti 2007; Widlansky et al. 2011).
Enhanced infrared satellite photograph of the southwest Pacific region taken at 2015 UTC 28 Oct 2010 (http://www.ghcc.msfc.nasa.gov/GOES/globalir.html; synoptic photograph is courtesy of the NASA Global Hydrology and Climate Center using data from the NCEP Aviation Weather Center, Kansas City, Missouri). Color shading shows the position of clouds and rainfall. Solid light blue lines circumscribe the ITCZ astride the equator and the SPCZ extending diagonally from Papua New Guinea toward South America. The locations of island groups contributing data to the South Pacific Rainfall Atlas (SPRAT) project that are used in this study are also shown. An approximation of the austral summer [December–February (DJF)] climatological position of the SPCZ extending to 20°S is indicated by the red dashed line (labeled DJF; from the 1982–2008 OLR analysis of Widlansky et al. 2011), while the austral warm season climatological position (based on Folland et al. 2002) is shown as the pink dashed line labeled November–April.
Citation: Monthly Weather Review 140, 11; 10.1175/MWR-D-11-00228.1
High baroclinicity and quasi-stationary convective activity characterizes the SPCZ. Since baroclinic instabilities are fast moving and dispersive, the convective activity related to the convergence zone spreads over a large area. The SPCZ portion energized from the Maritime Continent region experiences convective pulses arising from the Madden–Julian oscillation (MJO; Madden and Julian 1971), and these pulses are capable of propagating from the tropics into the subtropics through the SPCZ (Matthews 2012). SPCZ motions are linked to the El Niño–Southern Oscillation (ENSO; Vincent 1994), the Interdecadal Pacific Oscillation (IPO; Salinger et al. 2001), the Walker circulation (Trenberth 1991; Streten and Zillman 1984; van Loon and Shea 1987), the MJO (Madden and Julian 1994), and the southern annular mode (SAM; Thompson and Wallace 2000). The vast spatial extent of convection, cross-equatorial flow induced by a northwest–southeast orientation and position in the western tropical–subtropical Pacific highlights the significance of the SPCZ impact on the planetary climate system.
About half of the variability in SPCZ convective activity occurs on an interannual time scale (Widlansky et al. 2011), which is partially linked to SST and atmospheric modulations generated by ENSO (Webster 1982; Kiladis et al. 1989; Vincent et al. 2011). The spatial variability in the western tropical limb of the SPCZ is low in contrast to the temporal variability seen in convective pulses that arise from the MJO (Matthews 2012). Convective activity in the SPCZ is elevated in the austral warm season (November–April, also known as “the wet season”), with the convergence extending as far as 140°W. In the austral winter convective activity becomes weaker, and the SPCZ usually contracts northwestward toward Papua New Guinea along with a decreased angle of incidence into the subtropics (see Fig. 1 in Widlansky et al. 2011). The MJO modulation of the SPCZ activity on a 40–50-day time scale and the combination of interannual (ENSO) to multidecadal (IPO) variability generates distinct geospatial patterns in the position of the convection zone (Folland et al. 2002; Vincent et al. 2011). A more northward SPCZ position relative to climatology is observed in El Niño (positive IPO) episodes, while a southward bias is seen in La Niña (negative IPO) episodes. “Asymmetric” orientations of the SPCZ in a near-parallel alignment with the equator are also noted for very strong El Niños (Vincent et al. 2011).
SPCZ dynamics and associated precipitation variability play a major role in the climate of southwest Pacific Island nations. Significant SPCZ rainfall impacts recently observed in Vanuatu (in April 2008) and Fiji (January 2009) have included sustained heavy rainfall and severe precipitation that caused floods, which ruined subsistence and commercial agriculture crops (S. Kaniaha 2008, personal communication), and severely damaged civil infrastructure (S. McGree 2009, personal communication). Displacement of the SPCZ also generates anomalous wet periods in the Solomon Islands, Tonga, and Niue, as well as droughts in parts of Vanuatu, Samoa, and the low-lying islands of Tuvalu and Tokelau (Lorrey and Renwick 2011). Nearby, seasonal and ENSO-related variability of the SPCZ can also affect the climate of northern eastern Australia and even New Zealand. Overall, the SPCZ location is paramount for dictating the genesis area of tropical cyclones that regularly impact the region (Vincent et al. 2011) with extreme rainfall on some islands. This means synoptic to seasonal understanding of SPCZ motions are highly relevant for regional preparedness.
b. Purpose
Southwest Pacific rainfall variability, including that related to ENSO and the SPCZ, is currently monitored by the Tropical Rainfall Measuring Mission (TRMM) satellite measurements (examples of TRMM data usage can be found in Masunaga et al. 2005; L’Ecuyer et al. 2006; Nesbitt et al. 2006) and indirectly by outgoing longwave radiation (OLR) measurements. These remotely sensed measurements illustrate SPCZ-related precipitation, the position of intense convective foci, and interactions with tropical cyclones. However, the satellite monitoring period (1969–present; see Diamond et al. 2012 for explanation) is relatively short. Extending our records of SPCZ activity before the satellite era has the potential to improve our understanding of how the SPCZ operates and why it exists.
Here, we make use of surface rainfall observations from across the southwest Pacific region to estimate ENSO-related rainfall anomalies and the mean SPCZ location (e.g., Basher and Zheng 1998). Combining such analyses with records of tropical cyclone (TC) tracks and with atmospheric reanalysis fields allows us to ask the question: can a reliable estimation of SPCZ behavior back to at least the mid-twentieth century be made from in situ rainfall measurements?
A La Niña case study is provided here to demonstrate the performance of using station-based rainfall anomalies for identifying 1) past rainfall patterns associated with ENSO patterns, and 2) SPCZ location nested within those patterns during the presatellite era. This case study essentially highlights a “bucket on the ground” approach for estimating SPCZ position. Our findings suggest this approach is a viable option for identifying past SPCZ position, and it can also be used to test extended atmospheric reanalysis datasets such as those currently being developed from surface pressure data assimilations (Compo et al. 2006).
2. Data and methods
Monthly rainfall totals were sourced from 64 stations in the southwest Pacific sector (10°N–40°S, 130°E–130°W), via the Island Climate Update (ICU). Access and permissions to use the historical rainfall totals in the ICU database were courteously provided by Pacific Island National Meteorological Services, the Australian Bureau of Meteorology (BoM), National Oceanic and Atmospheric Administration (NOAA)/National Climatic Data Center (NCDC), the National Institute of Water and Atmospheric Research (NIWA), and Météo-France. Statistics on each station’s temporal coverage for the period of study were calculated and are shown in Table 1. Reference patterns for rainfall associated with different ENSO event types were then generated from the collection of station observations, and they provide a useful baseline that expresses the range of variations that can occur for southwest Pacific rainfall related to ENSO. The set of reference ENSO rainfall patterns were then categorized (well coupled, ocean dominated, atmospheric dominated, or neutral), which allowed comparisons to be made for a pre–satellite era case study. We consider that similarities between specific case study years described in this study to different varieties of ENSO, in terms of the spatial pattern observed for rainfall anomalies across the southwest Pacific, can also assist in the identification of the SPCZ position in the presatellite era.
List of stations used to generate the southwest Pacific regional rainfall anomaly map for the warm season (November–April) during the 1955/56 La Niña. Stations north of the equator have an ‘N’ listed alongside the latitude, while all other stations are south of the equator.
It was reasonable to expect that an SPCZ signature, as well as distinct low rainfall regions, would be highly pronounced during the austral warm season during an extreme ENSO event. As an initial test of this approach, the 1950s decade was the focus for this study because it directly precedes and partly includes the onset of the classic reanalysis period (1958 onward). That decade is also one of the most data-rich periods for rainfall measurements in the southwest Pacific region (Lorrey et al. 2008, 2009). We selected the most extreme ENSO event of the 1950s for this case study, which was a well-coupled and protracted La Niña that began in 1955 and extended through austral autumn 1956. This very strong ENSO event had significant ocean–atmosphere coupling (as identified in Gergis and Fowler 2005), and additional proxy evidence in ENSO-affected regions confirm the status of this event as having been regionally pervasive (Gergis and Fowler 2009). Atmospheric pressure anomalies for the 1955/56 La Niña indicates there were sustained positive Southern Oscillation index (SOI) values for 16 consecutive months during this time [all but 2 exceeded +10 (one standard deviation) on the Australia Bureau of Meteorology scale].
The coupled ENSO index (CEI; Gergis and Fowler 2005)1 was used to identify ENSO event types that occurred during the austral warm season, and it provides the basis for selecting analog years used to generate composite rainfall reference patterns for the case study seasons that are discussed below. Once the pre–satellite era case study season was selected, the departure from the climatic average (1961–90) was calculated for each station and mapped as an anomaly (% of average), and contours were projected via nearest-neighbor interpolation for the southwest Pacific based on the rainfall anomalies at each station.
The occurrence and position of TCs were considered as complementary information so the SPCZ could potentially be identified using the regional in situ rainfall measurements. Tropical cyclones, and their relationship to the SPCZ, play an important role in generating anomalously high rainfall for many stations in the southwest Pacific. Using the rainfall data from TC-affected stations (and associated tracks) was useful in terms of contextualizing seasonal rainfall patterns related to ENSO behavior. The TC data were provided from the International Best Tracks Archive for Climate Stewardship (IBTrACS) database (http://www.ncdc.noaa.gov/oa/ibtracs/). IBTrACS compiles tropical cyclone best-track data from several tropical cyclone monitoring and forecast centers around the globe, producing a unified global best-track dataset (Knapp et al. 2010). The data from IBTrACS was used to produce an enhanced tropical cyclone database tailored to the southwest Pacific basin from 120°W to 135°E. That database subset, known as the South Pacific Enhanced Archive of Tropical Cyclones (SPEArTC; Diamond et al. 2012; http://apdrc.soest.hawaii.edu/projects/speartc/), was used for this paper.
In addition, the signatures of several well-coupled La Niñas during the satellite era are examined for comparison. A comparatively strong well-coupled La Niña similar to the 1955/56 event occurred most recently during 2010/11. Four other well-coupled La Niña events of lesser strength also occurred within the last 30 years according to the CEI (1988/89, 1998/99, 1999/2000, and 2007/08). The rainfall anomaly patterns for those sets of events, and a demonstration of using their rainfall patterns to identify the SPCZ location, are also discussed below.
A test of the veracity of using an in situ rainfall anomaly approach to identify the SPCZ position was to apply the method to modern examples for the satellite era. We interpreted the reconstructed SPCZ position for 1955/56 season relative to examples in the satellite era using the NOAA–Cooperative Institute for Research in Environmental Sciences (CIRES) twentieth-century reanalysis data (Compo et al. 2011, 2006; Whitaker et al. 2004). The NOAA–CIRES twentieth-century reanalysis (20CR) dataset and data products were provided by the NOAA/Office of Oceanic and Atmospheric Research (OAR)/Earth System Research Laboratory (ESRL) Physical Sciences Division (PSD), Boulder, Colorado (http://www.esrl.noaa.gov/psd/data/gridded/data.20thC_Rean.html). Monthly and daily resolved reanalysis data were analyzed for November 1955–April 1956 and several satellite era La Niñas (listed above) using the online visualization routines provided by NOAA. Mean cloud water content and omega at 500 hPa were leading candidates for a direct comparison of our land-based rainfall SPCZ approximation. Folland et al. (2002) previously noted that SPCZ definitions based on OLR, maximum SSTs, and low-level convergence may not exactly align to surface precipitation maxima, but of several metrics they tested, omega at 500 hPa most closely aligned to rainfall maxima (for IPO examples). Hence, omega at 500 hPa was chosen in this study to compare to the station-based rainfall anomalies along with mean cloud water content, which has an obvious relationship to surface precipitation. In addition, a suite of other reanalysis data, including meridional and zonal flow, atmospheric pressure, and wind streamlines were used to describe the spatial characteristics of the regional circulation supporting the SPCZ formation during the pre–satellite era case study.
3. Results
a. Rainfall patterns
The regional rainfall averages (November–April 1961–90, corresponding with the austral warm season; Fig. 2) and rainfall anomalies associated with different ENSO types (Fig. 3) were established using monthly rainfall data from the study area. The criterion used to determine the ENSO types are explained below.
Austral warm season (November–April rainfall climatology based on South Pacific island station data, 1961–90). Contours represent total rainfall for the warm season in millimeters. The lines superposed on the rainfall contours indicate the location of the SPCZ during austral summer for conjoint IPO and SOI phases based on the divergence of the 10-m wind (after Folland et al. 2002). Solid white line: −SOI/+IPO; dashed white line: −SOI/−IPO; solid black line: +SOI/−IPO; dashed black line: +SOI/+IPO.
Citation: Monthly Weather Review 140, 11; 10.1175/MWR-D-11-00228.1
Southwest Pacific rainfall anomalies associated with different ENSO types as defined by the CEI for the 1961–90 period. (middle) CEI-neutral conditions, with the black and purple solid lines indicating the “neutral” and “negative” positions, respectively, of the SPCZ for austral summer (DJF) based on the analysis of Vincent et al. (2011) which included two of the selected CEI analog seasons contributing to that anomaly map. (top left) CEI-NIÑA (well-coupled La Niña) conditions, with the purple solid line indicating the “negative” position of the SPCZ for austral summer (DJF) based on the analysis of Vincent et al. (2011), which included the 1988/89 season that contributes to that anomaly map. (bottom right) CEI-NIÑO (well-coupled El Niño) conditions, with the red and blue solid lines indicating the “asymmetric” and “positive” positions, respectively, of the SPCZ for austral summer (DJF) based on the analysis of Vincent et al. (2011), which included two of the selected CEI analog seasons contributing to that anomaly map. (top middle) CEI-SOI NIÑA (atmospheric-dominant La Niña) conditions. (bottom middle) CEI-SOI NIÑO (atmospheric-dominant El Niño) conditions, with the black solid line indicating the neutral position of the SPCZ for austral summer (DJF) based on the analysis of Vincent et al. (2011) which included one of the selected CEI analog seasons contributing to that anomaly map. (left middle) CEI-N3.4 NIÑA (ocean-dominant La Niña) conditions, the black solid line indicates the neutral position of the SPCZ for austral summer (DJF) based on the analysis of Vincent et al. (2011), which included two of the three selected CEI analog seasons contributing to that anomaly map.
Citation: Monthly Weather Review 140, 11; 10.1175/MWR-D-11-00228.1
The CEI classification of Gergis and Fowler (2005) can be used to identify seven different ENSO types based on a dual threshold analysis of the Southern Oscillation index and Niño-3.4 index (see Gergis and Fowler 2005 for details). The ENSO types consist of well-coupled ocean and atmospheric conditions (termed CEI-NIÑO and CEI-NIÑA events). Neutral ENSO conditions (CEI-NEUTRAL) was also identified as a type, and poor coupling between the ocean and atmosphere, where either the oceanic anomalies were dominant (CEI-N3.4 NIÑA or N3.4 NIÑO) or the atmosphere was dominant (CEI-SOI NIÑO and CEI-SOI NIÑA) were also identified. Note that no CEI-N3.4 NIÑO event type was identified in the 1961–90 period.
The 1955/56 La Niña had a spatial pattern (Fig. 4) characterized by anomalously low rainfall in the northeast half of the southwest Pacific sector, including Tuvalu, Tokelau, the Northern Cook Islands, the Society Islands, and the Marquesas. Outside of the northeast sector, Raoul Island also had below-normal rainfall. Normal rainfall was observed for northern Fiji, Samoa, the northern half of the Southern Cook Islands, the eastern Austral Islands, and Pitcairn Island. Above normal rainfall was recorded in southeast Papua New Guinea, Vanuatu, Tonga, northern New Zealand, northern Queensland, and Norfolk Island. Well above-normal rainfall occurred in the Solomon Islands, Fiji, southern New Caledonia, Willis Island, and Sydney.
Regional rainfall anomaly pattern for the 1955/56 austral warm season (November–April) for the southwest Pacific, depicting a large region of suppressed rainfall (NE corner of map). Contours linking normal and above-normal rainfall in the SW portion of the map are intentionally not shown (see Fig. 5 and section 3c for more details).
Citation: Monthly Weather Review 140, 11; 10.1175/MWR-D-11-00228.1
b. Tropical cyclone tracks
A 30-yr (1981–2010) climatology for the annual number of TCs in the southwest Pacific basin (the area that is defined in section 2) indicates an average of 12.33 storms occur per year (Diamond et al. 2012). During the November 1955–April 1956 tropical cyclone season 16 TCs were initially counted east of 135°E,2 indicated by the IBTrACS database. However, a further analysis of the database indicated an actual occurrence of only 15 TCs (Diamond et al. 2012), which indicated one previously included track was erroneous. Only two TCs traveled east of the international date line and only one originated east of the international date line. For most TCs that traveled south of 25°S, the extratropical transition was characterized by a south to southeast exit pathway into the midlatitudes either near northern New Zealand or into the north Tasman Sea. West of the date line, six storms passed close (within 555 km) to New Caledonia, three passed west of Fiji, and three passed northern New Zealand. To the east of the date line only one storm occurred near Niue and the Southern Cook Islands.
c. SPCZ position from rainfall anomalies
The rainfall anomalies and TC tracks for 1955/56 are shown in Fig. 5. The position of the SPCZ was derived by selecting the island groups with anomalously high (>180% of normal) rainfall observations as a first cut, and a priori knowledge of the average SPCZ position during La Niña years coincident with a negative IPO phase (Folland et al. 2002). The a priori knowledge suggests the SPCZ position would be southwest of normal, flanking a dry region in the northeast of the southwest Pacific. It was also considered that the northern edge of the SPCZ would align approximately to the southwest flank of the near-normal rainfall contour. As a second cut, stations with >150% normal rainfall that were not directly crossed by TCs were selected as potential locations where the SPCZ could have been close by. With these qualitative guidelines in mind, extreme rainfall anomalies seen for land-based stations could then be connected for the 1955/56 case study to determine a possible SPCZ location.
Regional rainfall anomaly pattern for the 1955/56 austral warm season (November–April) for the southwest Pacific, depicting a large region of suppressed rainfall (NE corner of map). Tropical cyclone tracks are derived from the SPEArTC database (Diamond et al. 2012), and the solid red line depicts the “best guess” position of the SPCZ based on station anomalies that were not influenced by proximal TC activity. The blue dashed line is the climatological position of the SPCZ for the November–April period during 1961–90 (using data from Folland et al. 2002).
Citation: Monthly Weather Review 140, 11; 10.1175/MWR-D-11-00228.1
Only eight stations recorded more than 180% of normal rainfall [three in Fiji, two in New Caledonia, one Australian offshore island (Willis Island), one in Solomon Islands, and one in Sydney, Australia] during 1955/56, and these stations were chosen as potential candidates for the SPCZ seasonal mean position. Of the eight stations with anomalously high rainfall, those in New Caledonia, Willis Island, and several stations in Fiji west of the date line were rejected because of direct interactions with TC tracks during the 1955/56 season. In addition, Sydney was rejected because it is too far south and west to be realistically influenced directly by the SPCZ. This left only two primary sites that could have been affected by SPCZ rainfall, which included one Fiji station east of 180° (Lakeba) and one in the Solomon Islands (Honiara). In Fiji, Lakeba had 196% of normal rainfall, while 181% of normal rainfall occurred at Honiara in the Solomon Islands. Secondary sites affected by the SPCZ, based on stations unaffected by TCs that received >150% of normal rainfall, were located in the northwest (tropical) and southwest (subtropical) region of the southwest Pacific and included Port Morseby (Papua New Guinea) and Nuku’alofa (Tonga). In a north-to-south transect through Tonga, the rainfall gradient implies a maximum to the south of Nuku’alofa, which suggests this is a realistic position for the maximum northern boundary of the SPCZ location. Similarly, the gradient between Cairns (northeast Australia) and Port Morseby shows a change from near-normal rainfall to above-normal rainfall (170% of normal at Port Morseby), which suggests an estimated SPCZ position can be extended from Solomon Islands northeast toward Papua New Guinea. Extrapolating the trajectory through central-northern Papua New Guinea slightly to the north of the Port Moresby station is consistent with SPCZ joining up with the ITCZ in the equatorial tropics.
Well-coupled La Niña characteristics for events during the satellite era are seen in Fig. 6. The composite of average daily rainfall for four seasons (1988/89, 1998/99, 1999/2000, and 2007/08) was generated from the Global Precipitation Climatology Project (GPCP) dataset to illustrate a typical spatial pattern during well-coupled La Niñas. This pattern is paired with omega at 500 hPa for the same seasons. A region of very low rainfall is centered over the equator and spans the international date line. A low rainfall area typically coincides with colder-than-normal equatorial sea surface temperatures that exist during La Niña years. Some of the highest daily rainfall values are centered along a diagonally orientated band (the SPCZ) that extends from Papua New Guinea to the southeast across the Solomon Islands, north of Vanuatu, and Fiji to name a few islands. The orientation of this band and spatial position in the tropics and subtropics aligns to omega at 500 hPa for the same analogs. The GPCP dataset does not cover the 2010/11 period, but that seasonal example has a similar spatial pattern for rainfall when compared to the 1955/56 event. It was also typified by well-coupled ocean and atmosphere conditions, and generated one of the most severe droughts in recent times (Lorrey and Renwick 2011). The rainfall pattern for the 2010/11 La Niña also corresponds well to the spatial position of maximum values for omega at 500 hPa for the same time period.
(top left) Composite omega at 500-hPa geopotential height (vertical velocity in Pa s−1) patterns and (bottom left) GPCP rainfall for four La Niña events during the satellite era. (right) As in (left), but for the 2010/11 La Niña (rainfall contour map is from Lorrey and Renwick 2011). Black dots represent stations used to determine contours, and daily rainfall amounts (in mm) are noted next to stations. The dashed white line on each of the omega at 500-hPa diagrams indicates the core location of vertical motion, and the dashed black line superposed on the regional rainfall patterns is identical to the dashed white line.
Citation: Monthly Weather Review 140, 11; 10.1175/MWR-D-11-00228.1
d. Use of the twentieth-century reanalysis
Mean sea level pressures averaged over November 1955–April 1956 from the 20CR are shown in Fig. 7. The lowest pressures occurred about the equator from 140°E east to the date line, with a trough extending east of New Zealand, and another trough extending over eastern Australia. The subtropical high pressure belt is evident between 25° and 35°S, across southern Australia and east of the North Island of New Zealand, but with a trough over eastern Australia and the Tasman Sea. The pressure pattern seen during this season would have been associated with surface winds characterized by northeasterly and easterly flow across the southwest Pacific islands in the subtropical latitudes, while southeasterly flow would have occurred over eastern Australia. The circulation pattern would have also seen northerly and northwesterly flow across northern New Zealand, which is typical of blocking regimes that coincide with La Niña during the warm season (Kidson 2000).
November 1955–April 1956 average daily surface pressure (pressure and anomalies relative to the zonal mean are in Pa) over the southwest Pacific region derived from the twentieth-century reanalysis.
Citation: Monthly Weather Review 140, 11; 10.1175/MWR-D-11-00228.1
A vertical profile from 500 to 1000 hPa of the meridional (υ) wind component from the 20CR shows a southerly signature (positive υ wind) over central Australia and in the region of northern Queensland and the Coral Sea, most prominently at the lower levels, for the 1955/56 case study (Fig. 8). This is consistent with the location of the return flow from the descending branch of the Australian monsoon. The northerly meridional component (negative υ wind) was strongest just above the surface to the northern side of the Solomon Islands, Vanuatu, and Fiji during this La Niña event. Overall, the core of the northerly υ-wind strength shifted to the southeast, had a steeper angle of incidence into the subtropics, and weakened with progression upward through the atmosphere from 900 to 500 hPa during the 1955/56 La Niña austral warm season.
November 1955–April 1956 average daily υ-wind composite (from top to bottom) at 500-, 700-, 800-, 900-, and 1000-hPa geopotential heights (in m s−1). Blue and purple shades are negative, green shades are close to zero, orange and red shades are positive. The spatial coverage of each panel is 10°N–40°S, 130°E–130°W.
Citation: Monthly Weather Review 140, 11; 10.1175/MWR-D-11-00228.1
A similar vertical transect for the zonal wind component (u wind) illustrates westerly flow across and north of Papua New Guinea and the Solomon Islands and easterly flow farther south (from near New Caledonia to north of New Zealand) during the 1955/56 austral warm season (Fig. 9). The easterly flow is strongest at the surface and decreases with height as the general westerly circulation comes to dominate (e.g., at 500 hPa). The equatorward edge of the Southern Hemisphere westerlies can be seen extending to about 20°S at 500 hPa. This pattern is consistent with a cyclonic vortex over the Coral Sea (as seen in Fig. 11).
November 1955–April 1956 average daily u-wind composite (from top to bottom) at 500-, 700-, 800-, 900-, and 1000-hPa geopotential heights (in m s−1). Blue and purple shades are negative, green shades are close to zero, orange and red shades are positive. The spatial coverage of each panel is 10°N–40°S, 130°E–130°W.
Citation: Monthly Weather Review 140, 11; 10.1175/MWR-D-11-00228.1
Omega at 500 hPa from the 20CR was characterized by enhanced updraft in a diagonal line that extended from eastern Papua New Guinea southeast to Fiji during the austral warm season in 1955/56 (Fig. 10). This metric has previously been employed as a way to identify SPCZ location by Folland et al. (2002). The mean cloud water content (also a proxy for SPCZ location) pattern confirms that the maximum vertical velocity encompassed a region of moist convection and rainfall (Fig. 10). Subsidence is observed on both sides of the diagonally orientated convection zone. To the northeast, the subsidence appears to have spanned the region from central Kiribati to beyond 130°W, including northeast French Polynesia, while to the southwest it existed over central Australia, across the north Tasman Sea, and to the east of the North Island of New Zealand. The wind streamlines diagrams (Fig. 11) show how the subsidence described above fits into the larger hemispheric pattern, with low-level anticyclonic circulations over southern Australia and to the northeast of New Zealand. A cyclonic center is evident over the Coral Sea southeast of Papua New Guinea, with the SPCZ trough extending southeast. As noted by Takahashi and Battisti (2007), the strength and location of the subtropical high over the eastern Pacific defines the eastern boundary of SPCZ-related convection. In the 1955/56 case study the streamlines diagrams illustrate that the subtropical high itself appears to be strongly constrained and influenced by the Andes mountain range.
(top) Vertical velocity (omega) at the 500-hPa geopotential height and (bottom) mean cloud H2O content for the entire atmosphere for November 1955–April 1956 period. Green, blue, and purple shades are negative (upward motion), yellow shades are close to zero, and orange and red shades are positive (downward motion) for omega (in Pa s−1). Blue and purple shades represent low H2O and yellow to red shades represent high H2O concentrations for cloud H2O content. The spatial coverage of each panel is 10°N–40°S, 130°E–130°W. The red solid line is the estimated SPCZ position based on Fig. 5, while the blue solid line represents the core location of vertical motion.
Citation: Monthly Weather Review 140, 11; 10.1175/MWR-D-11-00228.1
November 1955–April 1956 wind streamlines (from top to bottom) at 700-, 800-, 900-, and 1000-hPa geopotential heights from the twentieth-century reanalysis.
Citation: Monthly Weather Review 140, 11; 10.1175/MWR-D-11-00228.1
4. Discussion
Regionally dispersed surface rainfall measurements have allowed a large-scale precipitation anomaly pattern for the austral warm season to be reconstructed for 1955/56. The pattern observed for this case study season was very similar to a well-coupled La Niña reference pattern that was established for 1961–90 (a CEI-NIÑA, and to some extent similar to a CEI-SOI NIÑA). This indicates that ENSO characteristics, including the strength of its atmospheric and oceanic components, might be usefully discerned from large-scale precipitation patterns across the southwest Pacific. This case study also indicates that use of station-based precipitation data back in time into the mid- to late 1800s could complement other reconstructions of ENSO activities that are solely based on SST and atmospheric circulation indices.
The integration of in situ rainfall anomalies and tropical cyclone track data highlighted a possible SPCZ location during 1955/56. Our estimate of the SPCZ position was compared with several metrics derived from the NOAA–CIRES twentieth-century reanalysis. Both vertical velocity and cloud water content indicate great similarity to the SPCZ position that was ascribed on anomalously high rainfall anomalies (Fig. 10). As a result, we suggest that the atmospheric characteristics that were supplied by the twentieth-century reanalysis have faithfully captured the location of a key climate system component in the southwest Pacific (the SPCZ). While more case studies are required to definitively prove this point, our findings indicate that the SPCZ could be reconstructed back in time in the presatellite era using surface rainfall data. Moreover, in the absence of surface rainfall measurements and detailed tropical cyclone information, an extension of our understanding of the SPCZ could be undertaken by using the twentieth-century reanalysis (Compo et al. 2006) via the method and metrics described above. In addition, extension of the analysis even farther into the 1800s could be afforded by supplementing the reanalysis dataset with spatially dense surface pressure measurements derived from colonial-era ship logs and land-based outposts, which is a major aim of the Atmospheric Circulation Reconstructions over the Earth (ACRE; Allan et al. 2011) and Recovery of Logbooks and International Marine data (RECLAIM; Wilkinson et al. 2011) projects.
5. Conclusions
The observations from 64 land-based meteorological stations spread across the southwest Pacific islands were used to develop regional rainfall averages and reference spatial patterns for different types of ENSO events (differentiated by coupling of the ocean and atmosphere) for the austral warm season. Historical rainfall data for a strong historical La Niña were subsequently compared to the ENSO reference patterns, and it is indicated that the 1955/56 event had a pattern typical of well-coupled, protracted La Niña. Many stations in the southwest sector of the study area had well above-normal rainfall (>150%) during this event, and there were 15 TCs that could be accounted for during the November–April period (above-normal activity for the season). Several stations along the northern edge of the tropical cyclone swarm were selected as candidates where the SPCZ could have resided. Connecting these selected stations illustrated a way to highlight past SPCZ position in the presatellite era. Complementary analyses of several modern La Niña events during the satellite era, depicted both as seasonal composites and individual events, and subsequent comparisons between gridded and station-based rainfall to omega at 500 hPa indicates a close correspondence between station-based rainfall amounts and SPCZ position. Our reconstructed SPCZ location based on in situ rainfall anomalies during the 1955/56 La Niña indicated a position southwest of normal, and this compared well with the spatial orientation of several NCEP–NCAR twentieth-century reanalysis metrics (total atmosphere cloud water content and omega at 500 hPa). Limitations of the in situ rainfall-based SPCZ reconstruction approach are anticipated for periods when data becomes more sparse (prior to the early 1900s). In addition, the subtropical and midlatitude portions of the SPCZ are not anticipated to be represented well by this approach because of a lack of station data east of the international date line and south of 25°S.
The spatiotemporal limitations of this approach suggest that the emerging reanalysis without radiosondes effort (Compo et al. 2006) could be usefully employed to reconstruct southwest Pacific atmospheric circulation features, and shed light on dynamics of the SPCZ and ENSO behavior prior to the satellite and radiosonde era in the absence of land-based observations of rainfall. Such an approach would provide a powerful tool to calibrate/corroborate signals observed in paleoclimate archives and place those signals in a wide-scale atmospheric circulation context. As such, continued support for ACRE and the 20CR is well warranted from a number of climate research perspectives because the recovery and digitization of surface pressure data from land-based or ship-board instruments can improve the quality of the reanalysis data. The case study we have presented also indicates that additional high-quality rainfall maps are possible, and that a linkage with the reanalysis can help better understand the driving processes behind rainfall in the southwest Pacific. As a result, the research approach employed this study will be extended in the future by continuing to produce ENSO case study maps for the early twentieth century. These maps will constitute a reference set that will be housed in the forthcoming South Pacific Rainfall Atlas (SPRAT). We consider that this resource will be useful for Pacific Island Meteorological Services as a reference tool, provide verifications of circulation characteristics seen in the twentieth-century reanalysis, and provide information for other parties that are interested understanding medium- to large-scale historical climate patterns of the southwest Pacific.
Acknowledgments
This work was made possible by support from all of the meteorological services that participate in the Island Climate Update (ICU) forum. Access to rainfall data was graciously granted by the Pacific Island Meteorological Service Directors in WMO Regional Association V at the 13th Regional Meteorological Service Directors Meeting held in Nadi, Fiji. The opportunity to undertake this study was afforded by a lengthy commitment from the Pacific Islands Meteorological Services and other regional meteorological organizations to observing and cataloguing rainfall information. In realizing how important these surface observations are, and recognizing that at present surface observations are on the decline, every effort should be made to support the continued presence of rainfall data collection in the region so a comprehensive documentation of climate variability for a key region that is influenced by ENSO, tropical cyclones, and the SPCZ is retained.
Funding for this study was received through bilateral agreement from the NOAA’s U.S. GCOS Program based at the NCDC (in cooperation with the Meteorological Service of New Zealand) to the South Pacific Rainfall Atlas (SPRAT) project, the New Zealand Ministry of Science and Innovation [Core funding to NIWA’s Climate Present and Past project (formerly Adaptation to Climate Variability and Change CO1 × 0701)], and peripherally through the New Zealand Ministry of Foreign Affairs and Trade funding for the Island Climate Update. Support for the Twentieth Century Reanalysis Project dataset is provided by the U.S. Department of Energy, Office of Science Innovative and Novel Computational Impact on Theory and Experiment (DOE INCITE) program, and the Office of Biological and Environmental Research (BER), and by NOAA’s Climate Program Office. Simon McGree (formerly at Fiji Meteorological Service) and Salesa Kaniaha (Acting Director of the Vanuatu Meteorological Service) are thanked for relaying accounts of intense SPCZ activity and rainfall to the authors. Georgina Griffiths is thanked for reviewing this manuscript.
REFERENCES
Allan, R., P. Brohan, G. P. Compo, R. Stone, J. Luterbacher, and S. Brönnimann, 2011: The international Atmospheric Circulation Reconstructions over the Earth (ACRE) Initiative. Bull. Amer. Meteor. Soc., 92, 1421–1425.
Basher, R. E., and X. Zheng, 1998: Mapping rainfall fields and their ENSO variation in data-sparse tropical south-west Pacific Ocean region. Int. J. Climatol., 18, 237–251.
Compo, G. P., J. S. Whitaker, and P. D. Sardeshmukh, 2006: Feasibility of a 100-year reanalysis using only surface pressure data. Bull. Amer. Meteor. Soc., 87, 175–190.
Compo, G. P., and Coauthors, 2011: The Twentieth Century Reanalysis Project. Quart. J. Roy. Meteor. Soc., 137, 1–28, doi:10.1002/qj.776.
Diamond, H. J., A. M. Lorrey, K. R. Knapp, and D. H. Levinson, 2012: Development of an enhanced tropical cyclone tracks database for the southwest Pacific from 1840 to 2010. Int. J. Climatol., doi:10.1002/joc.2412, in press.
Folland, C. K., J. A. Renwick, M. J. Salinger, and A. B. Mullan, 2002: Relative influences of the Interdecadal Pacific Oscillation and ENSO on the South Pacific Convergence Zone. Geophys. Res. Lett., 29, 1643, doi:10.1029/2001GL014201.
Gergis, J. L., and A. M. Fowler, 2005: Classification of synchronous oceanic and atmospheric El Niño-Southern Oscillation (ENSO) events for palaeoclimate reconstruction. Int. J. Climatol., 25, 1541–1565.
Gergis, J. L., and A. M. Fowler, 2009: A history of ENSO events since A.D. 1525: Implications for future climate change. Climatic Change, 92, 343–387.
Kidson, J. W., 2000: An analysis of New Zealand synoptic types and their use in defining weather regimes. Int. J. Climatol., 20, 299–316.
Kiladis, G. N., H. von Storch, and H. van Loon, 1989: Origin of the South Pacific convergence zone. J. Climate, 2, 1185–1195.
Knapp, K. R., M. C. Kruk, D. H. Levinson, H. J. Diamond, and C. J. Neumann, 2010: The International Best Track Archive for Climate Stewardship (IBTrACS). Bull. Amer. Meteor. Soc., 91, 363–376.
Kuleshov, Y., 2003: Tropical cyclones in the Southern Hemisphere: Influence of the El Niño/Southern Oscillation. Preprints, Seventh Int. Conf. on Southern Hemisphere Meteorology and Oceanography, Wellington, New Zealand, Amer. Meteor. Soc., 202–203.
L’Ecuyer, T. S., H. Masunaga, and C. D. Kummerow, 2006: Variability in the characteristics of precipitation systems in the tropical Pacific. Part II: Implications for atmospheric heating. J. Climate, 19, 1388–1406.
Lorrey, A., and J. A. Renwick, 2011: Assessment of the 2010-11 Southwest Pacific drought. NIWA Client Rep. AKL2011-036, New Zealand Ministry of Foreign Affairs and Trade, 20 pp.
Lorrey, A., H. J. Diamond, J. Renwick, J. Salinger, J. Gergis, and G. Dalu, 2008: A map-based South Pacific rainfall climatology. Eos, Trans. Amer. Geophys. Union, 89 (Fall Meeting Suppl.), Abstract A13A-0230.
Lorrey, A., and Coauthors, 2009: Historical rainfall patterns and South Pacific Convergence Zone variations, 1951-2007. Ninth Int. Conf. on Southern Hemisphere Meteorology and Oceanography, Melbourne, Australia, American Meteorological Society and the Australian Meteorlogical and Oceanographic Society, Poster Presentation, Short Abstracts, 103 pp.
Madden, R. A., and P. R. Julian, 1971: Description of a 40-50 day oscillation in the zonal wind in the tropical Pacific. J. Atmos. Sci., 28, 702–708.
Madden, R. A., and P. R. Julian, 1994: Observations of the 40–50-day tropical oscillation—A review. Mon. Wea. Rev., 122, 814–837.
Masunaga, H., T. S. L’Ecuyer, and C. D. Kummerow, 2005: Variability in the characteristics of precipitation systems in the tropical Pacific. Part I: Spatial structure. J. Climate, 18, 823–840.
Matthews, A. J., 2012: A multiscale framework for the origin and variability of the South Pacific Convergence Zone. Quart. J. Roy. Meteor. Soc., 138, 1165–1178, doi:10:1002/qj.1870.
Nesbitt, S. W., R. Cifelli, and S. A. Rutledge, 2006: Storm morphology and rainfall characteristics of TRMM precipitation features. Mon. Wea. Rev., 134, 2702–2721.
Salinger, M., J. Renwick, and A. Mullan, 2001: Interdecadal Pacific Oscillation and South Pacific Climate. Int. J. Climatol., 21, 1705–1721.
Streten, N. A., and J. W. Zillman, 1984: Climate of the South Pacific Ocean. World Survey of Climatology, H. van Loon, Ed., Vol. 15, The Oceans, Elsevier, 263–429.
Takahashi, K., and D. S. Battisti, 2007: Processes controlling the mean tropical Pacific precipitation pattern. Part II: The SPCZ and the Southeast Pacific dry zone. J. Climate, 20, 5696–5706.
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.
Trenberth, K. E., 1991: General characteristics of the El Niño–Southern Oscillation. Teleconnection Linking Worldwide Climate Anomalies, M. Glantz, R.W. Katz, and N. Nicholls, Eds., Cambridge University Press, 13–41.
van Loon, H., and D. J. Shea, 1987: The Southern Oscillation. Part VI: Anomalies of sea level pressure on the Southern Hemisphere and of Pacific sea surface temperature during the development of a warm event. Mon. Wea. Rev., 115, 370–379.
Vincent, D. G., 1994: The South Pacific convergence zone (SPCZ): A review. Mon. Wea. Rev., 122, 1949–1970.
Vincent, E. M., M. Lengaigne, C. E. Menkes, N. C. Jourdain, P. Marchesiello, and G. Madec, 2011: Interannual variability of the South Pacific Convergence Zone and implications for tropical cyclone genesis. Climate Dyn., 36, 1881–1896, doi:10.1007/s0038200907163.
Webster, P. J., 1982: Seasonality in the local and remote atmospheric response to sea surface temperature anomalies. J. Atmos. Sci., 39, 29–40.
Whitaker, J. W., G. P. Compo, X. Wei, and T. M. Hamill, 2004: Reanalysis without radiosondes using ensemble data assimilation. Mon. Wea. Rev., 132, 1190–1200.
Widlansky, M., P. Webster, and C. Hoyos, 2011: On the location and orientation of the South Pacific Convergence Zone. Climate Dyn., 36, 561–578, doi:10.1007/s00382-010-0871-6.
Wilkinson, C., and Coauthors, 2011: Recovery of logbooks and international marine data: The RECLAIM project. Int. J. Climatol., 31, 968–979, doi:10.1002/joc.2102.
The CEI is an index used to jointly indicate ocean and atmosphere conditions associated with ENSO (Gergis and Fowler 2005). The primary components of the CEI are the Southern Oscillation index, which is the normalized pressure difference between Tahiti in French Polynesia and Darwin, Australia, and the Niño-3.4 (N3.4) index, which is a measure of central-western equatorial Pacific sea surface temperature anomalies. Each index is smoothed (as explained in Gergis and Fowler 2005) and minimum threshold values indicate whether the atmosphere (SOI) or ocean (N3.4) indicates if an ENSO phase is present. When both indices indicate the same ENSO state, the ocean–atmosphere system is considered to be well coupled (either a NIÑO or NIÑA style event). When only the N3.4 index or SOI indicates ENSO is present, a N3.4 NIÑO/N3.4NIÑA or SOI NIÑO/SOI NIÑA style of event is suggested, respectively.
The IBTrACS database defines the westernmost longitude of the southwest Pacific basin as beginning at 135°E from a climatological standpoint (Kuleshov 2003).
