• CPC, 1998: Climate Diagnostics Bulletin. 78 pp.

  • Doherty, R., M. Hulme, and C. Jones, 1999: A gridded reconstruction of land and ocean precipitation for the extended Tropics from 1974 to 1994. Int. J. Climatol.,19, 119–142.

  • Huffman, G. J., and Coauthors, 1997: The Global Precipitation Climatology Project (GPCP) combined precipitation dataset. Bull. Amer. Meteor. Soc.,78, 5–20.

  • Kummerow, C., and L. Giglio, 1994a: A passive microwave technique for estimating rainfall and vertical structure information from space. Part I: Algorithm description. J. Appl. Meteor.,33, 3–18.

  • ——, and ——, 1994b: A passive microwave technique for estimating rainfall and vertical structure information from space. Part II: Applications to SSM/I data. J. Appl. Meteor.,33, 19–37.

  • Rasmusson, E. M., and T. H. Carpenter, 1982: Variations in tropical sea surface temperature and surface wind fields associated with the Southern Oscillation/El Niño. Mon. Wea. Rev.,110, 354–384.

  • Reynolds, R. W., and T. M. Smith, 1995: A high-resolution global sea surface temperature climatology. J. Climate,8, 1571–1583.

  • Ropelewski, C. F., and P. D. Jones, 1987: An extension of the Tahiti–Darwin Southern Oscillation index. Mon. Wea. Rev.,115, 2161–2165.

  • Salinger, M. J., R. E. Basher, B. B. Fitzharris, J. E. Hay, P. D. Jones, J. P. MacVeigh, and I. Schmidely-Leleu, 1995: Climate trends in the Southwest Pacific. Int. J. Climatol.,15, 285–302.

  • Wolter, K., and M. S. Timlin, 1998: Measuring the strength of ENSO events: How does 1997/98 rank? Weather,53, 315–324.

  • Wright, P. B., 1989: Homogenized long-period Southern Oscillation indices. Int. J. Climatol.,9, 33–54.

  • Xie, P., and P. A. Arkin, 1998: Global monthly precipitation estimates from satellite-observed outgoing longwave radiation. J. Climate,11, 137–164.

  • View in gallery

    Maps of correlation coefficient between GPCP monthly precipitation anomalies and (a) Niño-3.4 and (b) SOI. Period is 1986–97. Thick solid lines denote zero, thin solid lines are positive, and thin dashed lines denote negative values. Spacing of isolines is by 0.2. The large heavy-solid-lined and heavy-dashed-lined boxes indicate the Pacific (P) and Maritime Continent (MC) regions, respectively.

  • View in gallery

    GPCP-derived rainfall anomalies (mm day−1). Boxed regions are as in Fig. 1. The solid-lined box within P is the area with the maximum-averaged precipitation anomaly (Ap+), and the dashed-lined box within MC is the area with the minimum-averaged precipitation anomaly (Amc−). (a) Aug 1997, (b) Oct 1997, (c) Dec 1997, and (d) Feb 1998.

  • View in gallery

    Flow diagram illustrating the procedure in calculating the El Niño and La Niña indices (EI and LI). Boxed regions are as in Fig. 1. Shaded boxes within these blocks represent the areas over which the maximum (Amc+, Ap+) and minimum (Amc−, Ap−) average precipitation anomalies were computed.

  • View in gallery

    The maximum (solid lines) and minimum (dashed lines) average precipitation anomalies (mm day−1) for (a, top) Pacific (Ap+ and Ap−) and (b, bottom) Maritime Continent (Amc+ and Amc−). Thick lines denote GPCP data and adjusted GPROF (Table 1). Period is Jan 1979–May 1999. Thin lines denote raw GPROF data. Period is Jan 1991–May 1999. See Fig. 3 for location of Pacific (P) and Maritime Continent (MC) regions.

  • View in gallery

    El Niño index (EI), La Niña index (LI), and ENSO precipitation index (ESPI) for the periods 1979–84, 1985–90, 1991–96, and 1997–99 (note change of timescale). The EI is denoted by the dashed line, LI by the solid line, and ESPI by the heavy solid line.

  • View in gallery

    Comparison of sea surface temperature, sea level pressure, and rainfall-based ENSO indexes for the 20-yr period 1979–98. Dashed line denotes Niño-3.4, dotted line −(SOI/2), and solid line ESPI.

  • View in gallery

    Time-lag correlations between (a) Niño-3.4 or (b) SOI and Ap+, Ap−, Amc+, Amc−, EI, LI, and ESPI. The absolute magnitude correlation percentages are represented by thick solid line for ESPI, thick long dashed line for LI, thick short dashed line for EI, thin dash–dot line for Ap+, thin solid line for Ap−, thin long dashed line for Amc+, and thin short dashed line for Amc−. Positive numbers on the x axis indicate the number of months the rainfall indices lead the Niño-3.4 and SOI [(a) and (b), respectively]; negative numbers indicate the lag time.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 327 327 43
PDF Downloads 147 147 20

ENSO Indices Based on Patterns of Satellite-Derived Precipitation

View More View Less
  • 1 Laboratory for Atmospheres, Goddard Space Flight Center, Greenbelt, Maryland, and Joint Center for Earth Systems Technology, University of Maryland, Baltimore County, Baltimore, Maryland
  • | 2 Laboratory for Atmospheres, Goddard Space Flight Center, Greenbelt, Maryland
© Get Permissions
Full access

Abstract

In this study, gridded observed precipitation datasets are used to construct rainfall-based ENSO indices. The monthly El Niño and La Niña indices (EI and LI) measure the steepest zonal gradient of precipitation anomalies between the equatorial Pacific and the Maritime Continent. This is accomplished by spatially averaging precipitation anomalies using a spatial boxcar filter, finding the maximum and minimum averages within a Pacific and Maritime Continent domain for each month, and taking differences. The EI and LI can be examined separately or combined to produce one El Niño–Southern Oscillation (ENSO) precipitation index (ESPI). ESPI is well correlated with traditional sea surface temperature (e.g., Niño-3.4) and pressure indices [e.g., Southern Oscillation index (SOI)], leading Niño-3.4 by a month. ESPI has a tendency to produce stronger La Niñas than does Niño-3.4 and SOI. One advantage satellite-derived precipitation indices have over more conventional indices is describing the strength and position of the Walker circulation. Examples are given of tracking the impact of recent ENSO events on the tropical precipitation fields. The 1982/83 and 1997/98 events were unique in that, during the transition from the warm to the cold phase, precipitation patterns associated with El Niño and La Niña were simultaneously strong. According to EI and ESPI, the 1997/98 El Niño was the strongest event over the past 20 years.

Corresponding author address: Dr. Scott Curtis, JCET/UMBC, Laboratory for Atmospheres, NASA GSFC, Code 912, Greenbelt, MD 20771.

Email: curtis@agnes.gsfc.nasa.gov

Abstract

In this study, gridded observed precipitation datasets are used to construct rainfall-based ENSO indices. The monthly El Niño and La Niña indices (EI and LI) measure the steepest zonal gradient of precipitation anomalies between the equatorial Pacific and the Maritime Continent. This is accomplished by spatially averaging precipitation anomalies using a spatial boxcar filter, finding the maximum and minimum averages within a Pacific and Maritime Continent domain for each month, and taking differences. The EI and LI can be examined separately or combined to produce one El Niño–Southern Oscillation (ENSO) precipitation index (ESPI). ESPI is well correlated with traditional sea surface temperature (e.g., Niño-3.4) and pressure indices [e.g., Southern Oscillation index (SOI)], leading Niño-3.4 by a month. ESPI has a tendency to produce stronger La Niñas than does Niño-3.4 and SOI. One advantage satellite-derived precipitation indices have over more conventional indices is describing the strength and position of the Walker circulation. Examples are given of tracking the impact of recent ENSO events on the tropical precipitation fields. The 1982/83 and 1997/98 events were unique in that, during the transition from the warm to the cold phase, precipitation patterns associated with El Niño and La Niña were simultaneously strong. According to EI and ESPI, the 1997/98 El Niño was the strongest event over the past 20 years.

Corresponding author address: Dr. Scott Curtis, JCET/UMBC, Laboratory for Atmospheres, NASA GSFC, Code 912, Greenbelt, MD 20771.

Email: curtis@agnes.gsfc.nasa.gov

1. Introduction

The Global Precipitation Climatology Project’s (GPCP) community precipitation dataset (Huffman et al. 1997) merges monthly gauge observations and satellite analyses and is thus useful for monitoring extreme climatic events around the globe. One interannual phenomenon that deserves particular attention is the El Niño–Southern Oscillation (ENSO). The warm phase (El Niño) of this event is characterized in the tropical Pacific Ocean basin by anomalously warm waters in the east, weak trades, and anomalously low pressure and heavy rains in the east and high pressure and dry conditions in the west. Anomalies of the opposite sign tend to accompany the cold phase (La Niña). With this background, GPCP monthly estimates are used to construct novel indices that quantify the ENSO precipitation signal in the equatorial Pacific during the past 20 years.

Many atmospheric and oceanic indices are monitoring ENSO in near–real time (CPC 1998). Some are a combination of station data such as the Southern Oscillation index (SOI), which is the difference between Tahiti and Darwin surface pressure anomalies (see review in Ropelewski and Jones 1987). Others are spatial averages of fields of data such as the “NINO” indices, measures of SST in fixed regions from the central to eastern Pacific (Rasmusson and Carpenter 1982; Reynolds and Smith 1995). SST and SOI indices have the advantage of much longer records than satellite-based indices. Averages of 850-mb wind, 200-mb wind, and outgoing longwave radiation (OLR) are also used to monitor ENSO. There has been a recent effort to combine several atmospheric–oceanic variables into a single index. The multivariate ENSO index (MEI; Wolter and Timlin 1998) is the unrotated first principal component for the set of Pacific SST, sea level pressure (SLP), zonal and meridional components of the wind, air temperature, and cloudiness.

Precipitation in the Pacific is an especially important variable in characterizing the ENSO. Precipitation releases latent heating into the atmosphere that drives the large-scale circulation. During El Niño there is a deficit of rainfall over the Maritime Continent and abundant rain in the central to eastern Pacific, accompanied by a weakened and sometimes reversed Walker circulation. However, creating a suitable index of the interannual variability of large-scale precipitation patterns is a challenge. Combining rain rates at individual stations in the Pacific, for example Tahiti and Darwin, would not be satisfactory given the highly variable nature of precipitation. Although rain gauge records in the Pacific have been analyzed extensively (Salinger et al. 1995) and have even been used to create precipitation indices (Wright 1989), GPCP’s global dataset allows for the construction of mean precipitation anomalies over large areas of open ocean, giving insight to variations in the general circulation.

2. Data

The primary data source for this study is an experimental version of GPCP’s monthly global precipitation dataset (Huffman et al. 1997) from January 1979 to March 1999. The OLR precipitation index (Xie and Arkin 1998), trained on the original GPCP record, was used to extend back to 1979. Tropical precipitation estimates based on OLR are now available back to 1974 (Doherty et al. 1999). Goddard’s Profiling Algorithm (GPROF) applied to low-orbit SSM/I microwave data (Kummerow and Giglio 1994a,b) was used for the months April and May 1999. GPCP uses SSM/I data, but not the GPROF algorithm. All data sources have 2.5° spatial resolution. The disadvantage of GPROF is that it does not use gauge data to determine precipitation, but it allows for the extension of the index into near-real time. In future releases of this index our strategy is to replace GPROF when the more accurate GPCP analyses become available. The entire dataset was made homogeneous by the process described in the following section.

SST and SOI indices were taken from the Climate Prediction Center Web site (http://www.cpc.ncep.noaa.gov/data/indices).

3. Methods

The goal in constructing a precipitation-based measure of ENSO was to estimate the gradient of rainfall anomalies across the Pacific basin and ensure a good relationship with SST- and pressure-based indices (Fig. 1). Areas were selected (Fig. 1) that represent the “Maritime Continent” (MC) (10°N–10°S, 90°–150°E) and central to eastern Pacific (P) (10°N–10°S, 160°E–100°W). These regions capture the largest precipitation anomalies associated with the interannual variations of the Walker circulation. As an example, Fig. 2 shows precipitation anomalies based on the 1987–96 mean in the Tropics (20°N–20°S, 70°E–90°W) for four months during the 1997/98 ENSO. The precipitation gradient consists of positive anomalies over P and negative anomalies over MC. These areas also contain the largest correlations between GPCP and Niño-3.4 (Fig. 1a) and SOI (Fig. 1b). Within P and MC the absolute magnitude of the largest correlation is over +0.6.

The procedure for constructing the ENSO rainfall indices is represented by the flow diagram in Fig. 3. Because of the spatially varying nature of rainfall, it was decided to use a moving block average, described below, that would capture the strongest zonal gradients within the equatorial Pacific (Figs. 2, 3). This procedure is unlike many fixed area average indices and allows for a realistic meridional component of the precipitation gradient (Figs. 2, 3) and migration of the ascending and descending branches of the Walker circulation.

In the moving block average procedure, 10° lat × 50° long boxes are moved in 2.5° increments (grid block by grid block) throughout the P and MC domains (Fig. 3). The maximum and minimum average precipitation anomalies are found for P (Ap+ and Ap−, respectively) and MC (Amc+ and Amc−, respectively). To create a homogeneous record, monthly averages during 1991–97 of Ap+, Ap−, Amc+, and Amc− were computed for GPCP and GPROF. Then the GPROF values were subtracted from the GPCP values to create adjustments that were applied to the GPROF part of the record (April and May 1999; Table 1). Overall, GPROF gives higher maximums and lower minimums than GPCP over the Maritime Continent and Pacific basin (Table 1), with the largest differences occurring during ENSO events (Fig. 4). The reasons for the GPCP–GPROF differences are being studied. Significant adjustments are made to GPROF during the winter and spring months in constructing Ap+; otherwise the adjustments are small (Table 1). It is planned to replace GPROF values with the standard GPCP analyses as they become available, about three months after the observation time. The final series of Ap+ and Ap− are shown in Fig. 4a and Amc+ and Amc− in Fig. 4b. The interannual variability of Ap+ and Ap− is more pronounced than Amc+ and Amc− (Fig. 4). The ENSO precipitation anomalies are stronger and more coherent over the Pacific as compared to the Maritime Continent. Also, Amc+ and Amc− exhibit a high-frequency signal consistent with the Madden–Julian 30–60-day Oscillation. Also, Amc+ and Amc− are more in phase than Ap+ and Ap− (Fig. 4). This is primarily caused by the smaller size of the MC domain as compared with the P domain. In fact, there is often some overlap between the boxes used to calculate Amc+ and Amc−. The Amc− is subtracted from Ap+ and normalized to create the El Niño index (EI). The Ap− is subtracted from Amc+ and normalized to create the La Niña index (LI). Normalization is achieved for each month by subtracting off the 1979–98 means and dividing by the standard deviations as shown in Eq. (1):
i1520-0442-13-15-2786-e1
where i = (January, February, . . . , December). Positive EI (LI) values would indicate that the ENSO cycle was in its warm (cold) phase. There are advantages to quantifying the evolution of the warm and cold phases of ENSO through separate indices. However, there is also an advantage to having one index describe the ENSO cycle. Thus, before normalizing the indices, LI is subtracted from EI to create the ENSO precipitation index (ESPI), which is itself normalized, consistent with Eq. (1). Positive (negative) ESPI values indicate the warm (cold) phase of the ENSO cycle.

4. Results and discussion

The indices are plotted in Fig. 5 for three 6-yr time series from January 1979 to December 1996 and for 1997–99. The indices have substantial month to month variability, but also show a distinct interannual cycle. El Niño years 1982/83, 1986/87, 1991/92, and 1997/98 have EI values in excess of 2, representing the number of standard deviations away from the mean. The LI is larger than 2 standard deviations in 1984, 1988, and 1999 when ENSO was in the cold phase. No discontinuities are apparent between March and April 1999, the transition from GPCP to GPROF precipitation estimates. The maximum EI value of 3.0 in October 1997 and ESPI value of 2.8 in June 1997 indicate that the 1997/98 warm event was the strongest over the past 20 years within the limitations of the homogeneous datasets. However, the MEI, based on six oceanic–atmospheric variables other than precipitation, suggests that the 1997/98 event is second to the 1982/83 ENSO (Wolter and Timlin 1998). There is agreement between LI and ESPI that the 1983/84 cold event was the strongest on record. However, according to longer SST and SOI records, this was a weak La Niña. The 1997/98 event is detailed in the bottom panels of Fig. 5, where the EI indicates onset of the El Niño from March 1997 to a maximum in October 1997. A rapid decrease occurs from April to June 1998. The LI increases quickly from December 1997 to April 1998 just before EI drops.

The 1997/98 event is also presented in Fig. 2. From August to December 1997 there was an eastward movement and intensification of the largest positive precipitation anomaly in the P area. During this same time the driest area within MC was found off the coast of Indonesia. In February 1998 EI was strong but LI was positive bringing down the ESPI value (Fig. 5). LI increased from January to April 1998 because the negative anomaly moved out of the MC region and into the P domain, and the Maritime Continent was no longer dominated by negative anomalies as in previous months (Fig. 2). This precipitation anomaly pattern in February signaled the beginning of the end of the El Niño. A similar evolution in the precipitation anomalies occurred during the 1982/83 event, but not the 1986/87 event (Fig. 5). The time of the maximum precipitation anomalies during the annual cycle may have contributed to the way the El Niños transitioned into La Niñas. In the 1982/83 and 1997/98 events the EI was high in winter and dropped rapidly in late spring. However, for the 1986/87 event the EI was high in summer and gradually decayed into winter (Fig. 5).

Figure 6 shows a comparison of temperature, pressure, and rainfall indices for the period 1979–98. At the peaks of the largest El Niños ESPI tends to lead Niño-3.4. In the 1997/98 event in particular, ESPI peaks in June 1997 followed by Niño-3.4 in November and SOI in March 1998. ESPI also indicates La Niña–type precipitation patterns in 1984 and 1994 when Niño-3.4 and SOI are near zero. ESPI, Niño-3.4, and SOI all indicate, by the crossing of the anomaly zero line, that the 1997/98 event initiated in March 1997 and decayed between May and June 1998 (Fig. 6).

ESPI, EI, LI, Ap+, Ap−, Amc+, and Amc− were correlated with other measures of ENSO for the period January 1979 to December 1997 (Fig. 7). Niño-3.4 and SOI are most strongly correlated with ESPI (+0.81 and −0.73, respectively), indicating that the combination of east–west and west–east gradients of anomalous rainfall is the more comparable measure of ENSO. ESPI is also highly correlated with Niño-1+2 (+0.66), Niño-3 (+0.79), and Niño-4 (+0.71). ESPI, Ap−, and Amc− lead Niño-3.4 by a month or less, Ap+ and EI lag Niño-3.4 by a month or less, and LI and Amc+ lead Niño-3.4 by one to two months (Fig. 7a). Thus, the precipitation anomaly in the Maritime Continent precedes the evolution of Pacific SST, whereas the precipitation in the equatorial Pacific is closely tied to the surface temperature. The correlations in Fig. 7b are not as normally distributed as in Fig. 7a, but are skewed towards SOI leading the precipitation indices. However, there is a definite peak at zero lag time for all indices except Ap−.

5. Conclusions

The ENSO phenomenon is accompanied by substantial changes in the global climate system. Prediction and monitoring of ENSO events are key to mitigating the impacts of these changes. Historically, SST and SLP indices within the Pacific basin have been used to gauge the strength and duration of the warm and cold phases of ENSO. In this study, indices based on the zonal gradient of precipitation in the equatorial Pacific were constructed to monitor ENSO. Data were obtained from GPCP’s precipitation product. GPCP is spatially complete, so there is no need to rely on a few gauge observations in the Pacific Ocean to construct a gradient;unlike most global meteorological fields, the dataset is observed and uses no model output. Precipitation has an advantage over SST in that it does not have data voids over land areas. Last, “moving blocks” were used instead of fixed area averages to capture the value and location of the largest rainfall anomalies. Thus, this method accounts for the variable nature of ENSO events. Applying the moving block method to SST and other atmospheric–oceanic variables may be beneficial to the timeliness of ENSO monitoring.

EI and LI give insight to physical phenomena in the ENSO precipitation signature. For example, they indicate that equatorial Pacific rainfall anomalies during the 1982/83 and 1997/98 ENSOs evolved in a similar fashion. The decay of the events during the spring season was characterized by an El Niño signal (high EI) occurring simultaneously with a La Niña signal (high LI). Unlike the 1982/83 and 1997/98 ENSOs that peaked in winter, the 1986/87 event peaked during the summer, which may have contributed to a different pattern of decay. The EI and ESPI, as measures of the gradient of anomalous precipitation in the Pacific, provide further evidence that the 1997/98 El Niño was stronger than the 1982/83 event. ESPI tends to lead the Niño indices and suggests La Niña conditions when SOI and Niño-3.4 would indicate otherwise.

Here EI, LI, and ESPI are not only tools for monitoring ENSO, but can be used to analyze and to compare historic and future events. Ensembles of months with low and high ESPI values can be used to represent the phases of the ENSO cycle. Last, precipitation and accompanying latent heat release are the engine that drives the atmosphere. Therefore, the evolution of observed rainfall anomalies in the tropical Pacific can be related to changes in the Walker circulation system: SST, pressure gradient, wind, and divergence.

The latest EI, LI, and ESPI values and information pertaining to these indexes can be found at http://rsd.gsfc.nasa.gov/912/gpcp.

Acknowledgments

The authors wish to thank George Huffman for reviewing the manuscript and offering insightful advice. The authors also appreciate discussions with David Bolvin and Eric Nelkin. Research is supported through NASA’s Tropical Rainfall Measuring Mission (TRMM) Science Program and NASA’s Atmospheric Dynamics and Thermodynamics Program.

REFERENCES

  • CPC, 1998: Climate Diagnostics Bulletin. 78 pp.

  • Doherty, R., M. Hulme, and C. Jones, 1999: A gridded reconstruction of land and ocean precipitation for the extended Tropics from 1974 to 1994. Int. J. Climatol.,19, 119–142.

  • Huffman, G. J., and Coauthors, 1997: The Global Precipitation Climatology Project (GPCP) combined precipitation dataset. Bull. Amer. Meteor. Soc.,78, 5–20.

  • Kummerow, C., and L. Giglio, 1994a: A passive microwave technique for estimating rainfall and vertical structure information from space. Part I: Algorithm description. J. Appl. Meteor.,33, 3–18.

  • ——, and ——, 1994b: A passive microwave technique for estimating rainfall and vertical structure information from space. Part II: Applications to SSM/I data. J. Appl. Meteor.,33, 19–37.

  • Rasmusson, E. M., and T. H. Carpenter, 1982: Variations in tropical sea surface temperature and surface wind fields associated with the Southern Oscillation/El Niño. Mon. Wea. Rev.,110, 354–384.

  • Reynolds, R. W., and T. M. Smith, 1995: A high-resolution global sea surface temperature climatology. J. Climate,8, 1571–1583.

  • Ropelewski, C. F., and P. D. Jones, 1987: An extension of the Tahiti–Darwin Southern Oscillation index. Mon. Wea. Rev.,115, 2161–2165.

  • Salinger, M. J., R. E. Basher, B. B. Fitzharris, J. E. Hay, P. D. Jones, J. P. MacVeigh, and I. Schmidely-Leleu, 1995: Climate trends in the Southwest Pacific. Int. J. Climatol.,15, 285–302.

  • Wolter, K., and M. S. Timlin, 1998: Measuring the strength of ENSO events: How does 1997/98 rank? Weather,53, 315–324.

  • Wright, P. B., 1989: Homogenized long-period Southern Oscillation indices. Int. J. Climatol.,9, 33–54.

  • Xie, P., and P. A. Arkin, 1998: Global monthly precipitation estimates from satellite-observed outgoing longwave radiation. J. Climate,11, 137–164.

Fig. 1.
Fig. 1.

Maps of correlation coefficient between GPCP monthly precipitation anomalies and (a) Niño-3.4 and (b) SOI. Period is 1986–97. Thick solid lines denote zero, thin solid lines are positive, and thin dashed lines denote negative values. Spacing of isolines is by 0.2. The large heavy-solid-lined and heavy-dashed-lined boxes indicate the Pacific (P) and Maritime Continent (MC) regions, respectively.

Citation: Journal of Climate 13, 15; 10.1175/1520-0442(2000)013<2786:EIBOPO>2.0.CO;2

Fig. 2.
Fig. 2.

GPCP-derived rainfall anomalies (mm day−1). Boxed regions are as in Fig. 1. The solid-lined box within P is the area with the maximum-averaged precipitation anomaly (Ap+), and the dashed-lined box within MC is the area with the minimum-averaged precipitation anomaly (Amc−). (a) Aug 1997, (b) Oct 1997, (c) Dec 1997, and (d) Feb 1998.

Citation: Journal of Climate 13, 15; 10.1175/1520-0442(2000)013<2786:EIBOPO>2.0.CO;2

Fig. 3.
Fig. 3.

Flow diagram illustrating the procedure in calculating the El Niño and La Niña indices (EI and LI). Boxed regions are as in Fig. 1. Shaded boxes within these blocks represent the areas over which the maximum (Amc+, Ap+) and minimum (Amc−, Ap−) average precipitation anomalies were computed.

Citation: Journal of Climate 13, 15; 10.1175/1520-0442(2000)013<2786:EIBOPO>2.0.CO;2

Fig. 4.
Fig. 4.

The maximum (solid lines) and minimum (dashed lines) average precipitation anomalies (mm day−1) for (a, top) Pacific (Ap+ and Ap−) and (b, bottom) Maritime Continent (Amc+ and Amc−). Thick lines denote GPCP data and adjusted GPROF (Table 1). Period is Jan 1979–May 1999. Thin lines denote raw GPROF data. Period is Jan 1991–May 1999. See Fig. 3 for location of Pacific (P) and Maritime Continent (MC) regions.

Citation: Journal of Climate 13, 15; 10.1175/1520-0442(2000)013<2786:EIBOPO>2.0.CO;2

Fig. 5.
Fig. 5.

El Niño index (EI), La Niña index (LI), and ENSO precipitation index (ESPI) for the periods 1979–84, 1985–90, 1991–96, and 1997–99 (note change of timescale). The EI is denoted by the dashed line, LI by the solid line, and ESPI by the heavy solid line.

Citation: Journal of Climate 13, 15; 10.1175/1520-0442(2000)013<2786:EIBOPO>2.0.CO;2

Fig. 6.
Fig. 6.

Comparison of sea surface temperature, sea level pressure, and rainfall-based ENSO indexes for the 20-yr period 1979–98. Dashed line denotes Niño-3.4, dotted line −(SOI/2), and solid line ESPI.

Citation: Journal of Climate 13, 15; 10.1175/1520-0442(2000)013<2786:EIBOPO>2.0.CO;2

Fig. 7.
Fig. 7.

Time-lag correlations between (a) Niño-3.4 or (b) SOI and Ap+, Ap−, Amc+, Amc−, EI, LI, and ESPI. The absolute magnitude correlation percentages are represented by thick solid line for ESPI, thick long dashed line for LI, thick short dashed line for EI, thin dash–dot line for Ap+, thin solid line for Ap−, thin long dashed line for Amc+, and thin short dashed line for Amc−. Positive numbers on the x axis indicate the number of months the rainfall indices lead the Niño-3.4 and SOI [(a) and (b), respectively]; negative numbers indicate the lag time.

Citation: Journal of Climate 13, 15; 10.1175/1520-0442(2000)013<2786:EIBOPO>2.0.CO;2

Table 1.

Monthly adjustments made to maximum and minimum average precipitation anomalies for the part of the record for which GPROF data are used.

Table 1.
Save