NCEP/Climate Prediction Center ATLAS No. 5

A Precipitation Climatology for Stations in the Tropical Basin; Effects of ENSO


Appendix

Procedure for Categorization of Boreal Winters as "Warm" or "Cold" ENSO Cases

In order for a boreal winter to qualify as a warm (El Nino) or cold (La Nina) case, the SST in the Nino 3.4 region must show an anomaly of the appropriate sign, and the Tahiti-minus-Darwin sea level pressure-based SOI must show a mean anomaly of the opposite sign. The SST anomaly must be of at least a critical magnitude and for a minimum duration prior to and including the boreal winter season. Only the months of October through February are included in the evaluation, because the aim is to classify boreal winters rather than entire 12-month periods centered at some specified time of the calendar year. The months of January and February period must pass a separate requirement to ensure that the episode does not largely dissipate prior to the winter season. While SST is considered more critical than SOI, the SOI cannot show a conflicting anomaly, and, unless the SST anomaly is strong, the SOI must show an agreeing anomaly of nonnegligible magnitude.

The procedure begins by forming 3-month means of the SOI and the Nino 3.4 SST. These are then standardized with respect to the mean and standard deviation of the individual 3-month period, based on 1950-1996. Use of the standard deviation of the individual 3-month period in question rather than of all twelve 3-month periods pooled does not follow the signal-to-noise enhancement consideration discussed in Trenberth (1984), but does ensure uniform interannual variances across the seasonal cycle. (Note that the five years of 1950-54 are used to get a more complete sampling of ENSO behavior despite their absence from the analyses in this atlas.) The standardized anomalies for the five 3-month periods centered on October (i.e. Sep-Oct-Nov), November (Oct-Nov-Dec), December, January and February are the data of interest. For ease of communication, these 3-month periods will be named by their center month in this discussion. The standardized anomalies for each of the five periods are shown in Tables A1 and A2 for SST and SOI, respectively, Anomalies for 1997 are shown for completeness, although only data for the 1955-96 period are directly applied to the analyses in this atlas. The SOI values are sign-reversed so that both SST and SOI are expected to be positive for a warm ENSO episode and negative for the cold ENSO case. To be classified as a non-neutral winter, the data must satisfy the following requirements:

SST Requirements:

1) Of the five periods, four-in-a-row must be >0.6, and Feb must be >0.4. Exemption #1: The winter can also qualify if 0.00.6 or (Dec+Jan+Feb)/3>0.6. Exemption #2: The winter can also qualify with only three-in-a row periods >0.6 when Jan and Feb <0.6 but Feb qualifies (possibly through exemption #1) and Jan is >0.5 when the SOI is >1.00 for at least four of the five periods.

2) At least one of the five periods must be >0.8, or three of the five must be >0.7.

SOI Requirements:

1) The average of the five periods must be >0.3, and either Jan or Feb must be >0.0. However, the winter can qualify if the average of the five periods is below 0.3 but at least 0.0, when the SST >1.0 for Feb, Jan or Dec, or for (Oct+Nov)/2.

In determining the sets of warm and cold winters as listed in Table 4, the following decisions are noted with respect to borderline winters, or non-borderline winters that have been classified differently by others. Note that 3-month periods are always being discussed, referred to by their middle month.

1956-57 (neutral): While the SOI qualifies as a cold winter, the SST fails, with Jan very weak and Feb actually slightly in the wrong (positive) direction.

1963-64 (warm): While the SOI qualifies as a warm winter, February SST is only 0.37. However, when averaged with January and December, 0.6 is exceeded, qualifying the winter. The (Dec+Jan+Feb)/3 SST is 0.70.

1964-65 (cold): While the SST qualifies as a cold winter, the SOI fails, with a five-period mean of -0.25 (recall that the SOI has been sign-reversed). However, the SST exceeds -1.00 when averaged over Oct and Nov, qualifying the winter.

1971-72 (neutral): While the SOI qualifies as a cold winter, the SST fails, with Jan weaker than -0.6 and Feb weaker than -0.4. The SST in Dec, Jan and Feb is not cold enough to override the Feb SST failure, and the SOI is not strong enough to override the Jan SST failure.

1976-77 (neutral): While the SST qualifies for a warm episode, the SOI shows no tendency toward a warm episode, with average values slightly in the opposite (cold) direction.

1977-78 (warm): The SOI strongly qualifies as a warm winter, but Jan SST slightly misses the 0.6 requirement. However, it does exceed 0.5, and at least four of the five periods (in fact, all five) have SOI>1.0, allowing for qualification.

1983-84 (neutral): While the SST qualifies as a cold winter, the SOI fails, with a five-period mean of -0.24 (recall that the SOI has been sign-reversed). The strongest SST over the five periods is -0.85, falling short of the -1.00 required to override the SOI failure.

1984-85 (cold): While the SST qualifies as a cold winter, the SOI fails, with a five-period mean of -0.07 (recall that the SOI has been sign-reversed). However, the SST exceeds -1.00 from Nov through Feb, qualifying the winter.

1987-88 (warm): The SST easily qualifies as a warm winter, and the SOI also narrowly passes with a five-period mean of 0.34 (recall that the SOI has been sign-reversed) and positive values for all five periods. Even if the mean SOI failed the 0.3 requirement (but was positive), the winter would qualify because the SST exceeds 1.00 for the Oct, Nov and Dec periods. Note that this winter represents the final stage of the strong 1986-87 event, and has not traditionally been classified as a warm episode case. (The 1986-87 event actually peaked in the middle of 1987 rather than boreal winter 1986-87.)

1992-93 (neutral): The SOI meets requirements for a warm winter. However, the SST falls short, never exceeding 0.4 until February. While the SST from March through August 1993 definitely depicts a warm episode (with continued unequivocal support from the SOI), it occurs during the portion of the annual cycle that prevents it from qualifying for either 1992-93 or 1993-94 winters. This is the only clear case within the 1955-96 period of an ENSO episode that does not lap over at least one boreal winter season. A similar situation occurred in boreal spring-summer of 1987, except that the warm event was so strong that both 1986-87 and 1987-88 qualified for moderately strong events, though mid-1987 had the outstandingly strong El Nino.

1994-95 (warm): Both the SOI and SST qualify for a warm winter.

1995-96 (cold): While the SST unambiguously qualifies as a cold winter, the SOI fails the 5-period mean, falling between 0.0 and the -0.3 requirement (recall that the SOI has been sign-reversed). However, the SST in Nov, Dec and Jan periods exceeds -1.00, creating an exemption for the SOI failure.

As it turns out, the same list of warm and cold winters as that shown in Table 4 would emerge if the criterion for qualification were simply to average the standardized SOI values for the five months of Oct to Feb, do the same for SST, and then form a weighted average the two results with a 0.85 weight for SST and 0.15 weight for the SOI. A winter qualifies if its weighted average exceeds 0.7. However, this simpler criterion may not always result in the same list in the future or before 1955, and risks accepting cases that are strong in boreal fall but dissipate too rapidly as winter begins. The weighted average provides a relative strength score for all winters. The following table lists the strength score for the winters analyzed here, with qualifying winters shown in bold. Qualifying winters with score magnitudes of 0.70 to .99 can be considered standard, 1.00 to 1.49 moderately strong, and 1.50 strong.

      
1954-55  -.89       1960-61  -.23       1970-71 -1.49       1980-81  -.03    1990-91  .29
1955-56 -1.46       1961-62  -.43       1971-72  -.68       1981-82   .01    1991-92 1.64
1956-57  -.38       1962-63  -.54       1972-73  1.64       1982-83  2.61    1992-93  .25
1957-58  1.37       1963-64   .82       1973-74 -1.76       1983-84  -.64    1993-94  .20
1958-59   .37       1964-65  -.72       1974-75  -.58       1984-85  -.91    1994-95  .98 
1959-60  -.20       1965-66  1.43       1975-76 -1.52       1985-86  -.45    1995-96 -.77
                    1966-67  -.31       1976-77   .66       1986-87  1.22     1996-97 -.41
                    1967-68  -.49       1977-78   .74       1987-88  1.00 
                    1968-69   .80       1978-79  -.01       1988-89 -1.88
                    1969-70   .77       1979-80   .47       1989-90  -.03


Table A1. Standardized anomaly of 3-month mean SST in region Nino 3.4

YEAR      DJF   JFM   FMA   MAM   AMJ   MJJ   JJA   JAS   ASO   SON   OND   NDJ
1950    -1.73 -1.55 -1.65 -1.74 -1.79 -1.50 -0.99 -0.98 -0.92 -0.97 -0.84 -0.97
1951    -0.64 -0.72 -0.36 -0.42 -0.06  0.14  0.56  0.78  0.87  0.79  0.78  0.59
1952     0.47  0.12  0.26  0.07 -0.08 -0.48 -0.36 -0.11  0.08  0.00 -0.15 -0.04
1953     0.17  0.45  0.72  0.70  0.74  0.64  0.55  0.77  0.59  0.68  0.25  0.30
1954     0.27  0.39  0.03 -0.20 -0.59 -0.83 -1.20 -1.35 -1.23 -0.99 -0.85 -1.01
1955    -0.89 -1.03 -1.00 -1.43 -1.47 -1.45 -1.23 -1.42 -1.71 -2.01 -1.78 -1.52
1956    -1.01 -0.97 -1.07 -0.90 -0.72 -0.68 -0.99 -1.09 -0.78 -0.70 -0.47 -0.48
1957    -0.12  0.07  0.47  0.66  0.75  0.90  1.21  1.20  1.11  0.98  1.25  1.55
1958     1.70  1.82  1.56  1.20  0.88  0.72  0.69  0.20  0.11 -0.05  0.28  0.40
1959     0.58  0.46  0.46  0.35  0.20 -0.22 -0.54 -0.65 -0.42 -0.26 -0.08 -0.07
1960    -0.16 -0.13 -0.03  0.09 -0.04 -0.13 -0.03  0.12  0.02 -0.18 -0.29 -0.27
1961    -0.17 -0.18 -0.05  0.08  0.42  0.29  0.06 -0.49 -0.60 -0.57 -0.37 -0.28
1962    -0.29 -0.38 -0.52 -0.64 -0.58 -0.34 -0.05 -0.16 -0.30 -0.50 -0.52 -0.58
1963    -0.59 -0.40 -0.12  0.01  0.05  0.50  0.98  1.33  1.21  1.08  1.02  0.92
1964     0.80  0.37 -0.28 -0.96 -1.42 -1.36 -1.19 -1.09 -1.07 -1.09 -0.99 -0.94
1965    -0.66 -0.33 -0.07  0.28  0.69  1.32  1.68  1.88  1.85  1.74  1.67  1.49
1966     1.36  1.24  1.27  0.88  0.69  0.49  0.51  0.23  0.00 -0.09 -0.19 -0.28
1967    -0.31 -0.51 -0.88 -0.89 -0.48 -0.07 -0.09 -0.41 -0.51 -0.49 -0.39 -0.42
1968    -0.67 -0.87 -0.92 -0.83 -0.49 -0.01  0.45  0.48  0.40  0.48  0.64  0.89
1969     1.05  1.24  1.17  1.25  1.11  0.95  0.71  0.73  0.95  0.89  0.90  0.86
1970     0.79  0.58  0.57  0.43  0.02 -0.74 -1.16 -1.31 -1.28 -1.31 -1.45 -1.51
1971    -1.59 -1.56 -1.73 -1.47 -1.29 -0.90 -0.82 -0.75 -0.76 -0.82 -0.81 -0.78
1972    -0.55 -0.33  0.12  0.59  1.08  1.48  1.73  1.85  1.95  1.95  2.00  1.88
1973     1.70  1.40  0.78  0.02 -0.83 -1.32 -1.56 -1.55 -1.49 -1.53 -1.63 -1.76
1974    -1.76 -1.75 -1.63 -1.54 -1.18 -0.91 -0.61 -0.55 -0.62 -0.70 -0.77 -0.65
1975    -0.49 -0.56 -0.75 -1.08 -1.45 -1.66 -1.72 -1.61 -1.67 -1.50 -1.48 -1.51
1976    -1.51 -1.35 -1.19 -1.03 -0.77 -0.27  0.15  0.53  0.85  0.99  0.91  0.81
1977     0.63  0.61  0.28  0.34  0.31  0.56  0.41  0.42  0.53  0.72  0.73  0.72
1978     0.56  0.46 -0.07 -0.33 -0.70 -0.64 -0.75 -0.64 -0.49 -0.23 -0.08 -0.03
1979     0.01  0.22  0.35  0.43  0.31  0.07 -0.02  0.28  0.53  0.66  0.40  0.48
1980     0.56  0.45  0.37  0.34  0.59  0.64  0.39  0.10 -0.05  0.05  0.14  0.03
1981    -0.20 -0.45 -0.57 -0.49 -0.45 -0.44 -0.65 -0.55 -0.23  0.08  0.08  0.06
1982     0.09  0.07  0.18  0.66  1.33  1.75  1.83  1.87  2.13  2.28  2.36  2.46
1983     2.71  2.86  2.75  2.47  1.89  1.16  0.35 -0.12 -0.39 -0.58 -0.83 -0.85
1984    -0.70 -0.59 -0.47 -0.51 -0.74 -0.77 -0.79 -0.46 -0.51 -0.65 -1.08 -1.14
1985    -1.22 -1.23 -1.41 -1.29 -1.16 -0.88 -0.67 -0.47 -0.43 -0.41 -0.37 -0.45
1986    -0.65 -0.83 -0.74 -0.52 -0.19  0.12  0.43  0.74  0.97  1.12  1.10  1.16
1987     1.31  1.61  1.85  1.95  2.10  2.44  2.67  2.64  2.28  1.85  1.35  1.05
1988     0.78  0.55  0.08 -0.80 -1.74 -2.36 -2.36 -2.03 -1.94 -1.93 -2.08 -2.01
1989    -1.82 -1.83 -1.81 -1.62 -1.26 -0.85 -0.67 -0.48 -0.41 -0.32 -0.25 -0.14
1990     0.03  0.21  0.43  0.59  0.47  0.40  0.32  0.38  0.41  0.32  0.32  0.32
1991     0.35  0.28  0.30  0.61  1.04  1.37  1.25  0.99  0.95  1.09  1.41  1.58
1992     1.89  2.16  2.50  2.51  1.94  1.33  0.55  0.25 -0.06 -0.05  0.01  0.12
1993     0.23  0.38  0.87  1.45  1.62  1.36  0.75  0.51  0.39  0.43  0.32  0.18
1994     0.02 -0.05  0.02  0.28  0.48  0.54  0.62  0.62  0.80  0.94  1.13  1.14
1995     1.02  0.80  0.63  0.33  0.10 -0.05 -0.21 -0.46 -0.77 -0.89 -0.91 -0.84
1996    -0.88 -0.87 -0.80 -0.65 -0.51 -0.31 -0.19 -0.16 -0.30 -0.31 -0.39 -0.45
1997    -0.53 -0.38 -0.04  0.74  1.67  2.49  3.02  3.02  3.15  3.00  2.73  2.53
 



Table A2 Standardized anomaly of 3-month mean of SOI - multiplied by -1

YEAR      DJF   JFM   FMA   MAM   AMJ   MJJ   JJA   JAS   ASO   SON   OND   NDJ
1950    -1.11 -1.61 -1.98 -1.65 -1.92 -2.01 -2.15 -1.44 -1.33 -1.29 -1.91 -1.74
1951    -1.60 -0.53  0.24  0.97  0.84  1.05  0.80  1.11  1.07  1.20  1.05  0.88
1952     0.84  0.46  0.29 -0.28 -0.46 -0.71 -0.32 -0.01  0.05 -0.01  0.35  0.39
1953     0.60  0.33  0.44  1.23  1.14  1.08  0.79  1.14  1.11  0.60  0.28  0.06
1954     0.11 -0.05  0.05 -0.22 -0.15 -0.12 -0.40 -0.55 -0.50 -0.25 -0.66 -0.41
1955    -0.86 -0.47 -0.46 -0.27 -0.77 -1.58 -1.71 -1.67 -1.62 -1.68 -1.55 -1.36
1956    -1.28 -1.33 -1.27 -1.44 -1.50 -1.43 -1.25 -0.84 -1.13 -0.84 -1.18 -0.62
1957    -0.46 -0.07  0.04  0.37  0.41  0.41  0.35  0.62  0.66  0.76  0.54  1.15
1958     0.98  0.87  0.16  0.35  0.37  0.30 -0.40 -0.27 -0.12  0.35  0.45  0.71
1959     1.04  0.43  0.00 -0.69 -0.05  0.27  0.53  0.30 -0.03 -0.58 -0.88 -0.75
1960    -0.31 -0.29 -0.58 -0.78 -0.38 -0.20 -0.34 -0.67 -0.50 -0.53 -0.54 -0.46
1961    -0.45  0.66  0.19  0.40 -0.43 -0.03  0.08 -0.05  0.19 -0.12 -0.61 -1.44
1962    -1.04 -0.47  0.10 -0.53 -0.85 -0.63 -0.36 -0.33 -0.71 -0.69 -0.59 -0.53
1963    -0.57 -0.83 -0.81 -0.74  0.05  0.47  0.60  0.42  0.85  1.07  1.31  0.89
1964     0.62 -0.16 -0.87 -0.95 -0.90 -0.42 -1.00 -1.24 -1.53 -1.08 -0.47  0.20
1965     0.20 -0.08 -0.03  0.05  0.70  1.35  1.66  1.67  1.26  1.51  1.00  1.03
1966     0.55  1.10  0.80  0.94  0.37  0.24 -0.19 -0.08 -0.02  0.14  0.21 -0.42
1967    -0.91 -1.46 -0.96 -0.34 -0.15 -0.10 -0.45 -0.46 -0.42 -0.03  0.42  0.25
1968    -0.30 -0.46 -0.28 -0.43 -1.01 -1.18 -0.66 -0.15  0.15  0.26  0.12  0.54
1969     0.71  0.70  0.33  0.32  0.45  0.56  0.46  0.79  0.95  0.83  0.33  0.24
1970     0.67  0.71  0.40 -0.12 -0.44 -0.21 -0.29 -0.41 -0.96 -1.53 -1.75 -1.49
1971    -1.39 -1.52 -2.28 -2.14 -1.41 -0.47 -0.70 -1.16 -1.76 -1.51 -1.00 -0.44
1972    -0.48 -0.53 -0.37  0.81  1.44  2.02  1.46  1.50  1.21  1.03  0.95  0.66
1973     1.08  0.49  0.47 -0.21 -0.54 -0.76 -1.09 -1.15 -1.20 -1.88 -2.07 -2.58
1974    -2.08 -2.25 -1.93 -1.88 -1.10 -1.02 -0.80 -1.13 -1.01 -0.73 -0.32  0.21
1975    -0.08 -0.54 -1.21 -1.29 -1.39 -1.60 -2.11 -2.35 -2.22 -2.05 -1.96 -1.73
1976    -1.70 -1.49 -1.22 -0.95 -0.39  0.29  0.89  1.27  0.73  0.01 -0.29 -0.02
1977    -0.04  0.14  0.33  1.16  1.51  1.63  1.63  1.26  1.19  1.30  1.41  1.04
1978     1.47  1.22  1.32 -0.25 -0.61 -0.91 -0.44 -0.28  0.10  0.16  0.24  0.16
1979    -0.09 -0.06 -0.08  0.02 -0.30 -0.94 -0.53 -0.38  0.20  0.20  0.53  0.32
1980     0.11  0.17  0.74  0.94  0.74  0.33  0.16  0.19  0.16  0.35  0.22  0.07
1981     0.07  0.61  0.84  0.40 -0.82 -1.16 -1.00 -0.63 -0.13 -0.02 -0.04 -0.64
1982    -0.62 -0.55 -0.20  0.07  0.98  1.72  2.28  2.20  2.23  2.54  2.73  3.10
1983     3.28  3.54  2.98  1.45  0.42  0.16  0.45 -0.05 -0.50 -0.53 -0.19 -0.06
1984    -0.28 -0.06 -0.22 -0.02  0.09  0.29  0.20 -0.16  0.04 -0.03  0.13  0.06
1985    -0.12 -0.37 -0.93 -0.62 -0.10  0.45  0.10 -0.24 -0.11  0.27  0.21 -0.30
1986     0.04  0.01  0.24 -0.01 -0.31 -0.24 -0.12  0.39  0.21  0.46  0.77  1.27
1987     1.24  1.29  1.78  2.13  2.40  2.16  1.87  1.50  1.07  0.59  0.37  0.19
1988     0.42  0.14  0.01 -0.65 -0.46 -0.76 -0.92 -1.67 -1.79 -2.03 -1.75 -1.68
1989    -1.31 -1.21 -1.44 -1.71 -1.70 -1.15 -0.35 -0.31 -0.26 -0.40 -0.02  0.30
1990     0.85  0.98  0.89 -0.32 -0.70 -0.75 -0.03  0.25  0.35  0.38  0.22  0.12
1991    -0.12  0.12  0.73  1.45  1.33  0.85  0.52  0.90  1.33  1.34  1.40  1.83
1992     1.97  2.21  1.80  1.48  1.05  0.76  0.69  0.20  0.55  0.89  1.18  0.80
1993     0.86  0.86  1.15  1.08  1.51  1.22  1.49  1.14  1.23  0.77  0.47  0.01
1994    -0.02  0.38  1.04  1.60  1.68  1.51  1.67  1.84  1.72  1.35  1.18  0.78
1995     0.67  0.07  0.39  0.64  0.91  0.20 -0.12 -0.30 -0.11 -0.11  0.22 -0.15
1996    -0.13 -0.69 -0.74 -0.96 -1.16 -0.92 -0.94 -0.69 -0.65 -0.42 -0.46 -0.43
1997    -0.98 -0.40  0.11  1.59  2.07  2.06  1.93  1.61  1.81  1.61  1.49  1.68
 

Note: 1997 data (shown in bold), expressed relative to the 1955-96 base period statistics, are not analyzed in this atlas. They are included here for completeness and to show the powerful 1997-98 ENSO episode which may or may not end up being stronger than the famous 1982-83 episode.