TABLE 1. Specific values of nine predictors for each year from 1950 through 1990 used to make the 1 August forecasts of subsequent seasonal TC activity. Data in columns 4, 5, and 8 are expressed in terms of standardized deviations. Note that data in columns 6, 7, and 9 are departures from the long-term (1950-1990) means.
 

	September (estimated) QBO zonal wind at 10°N-(ms -1)			Observerd rainfall		Observed Carribean basin		Observed ENSO
Year	1  50 mb	2 30 mb	3 Absolute shear |50-30 mb|	4 Western sahelian (std. dev.) June-July	5 Gulf of Guinea Previous yr (std. dev.) Aug-Nov.	6 SLPA (mb) June-July	7 ZWA (ms -1) June-July	8 SOI (std. dev.) June-July	9 SSTA (C) June-July
1950	0	-1	1	0.35	1.07	-0.3	-2.5	2.1	-0.74
1951	-4	-16	12	-0.16	-0.66	-0.6	0.0	-0.7	0.94
1952	-25	-14	11	0.61	0.65	0.0	-2.5	0.5	-0.18
1953	-3	-14	11	0.87	0.41	-0.2	0.0	-0.2	0.56
1954	-22	-32	10	0.63	-0.16	-0.8	-3.0	0.0	-1.14
1955	-2	-2	0	1.44	0.64	-0.7	-5.2	1.5	-1.00
1956	-17	-28	11	0.06	0.41	0.2	-0.7	1.1	-0.66
1957	-6	-6	0	0.13	-0.36	-0.4	3.0	0.0	1.14
1958	-9	-25	16	0.34	1.03	-0.4	-4.5	0.2	0.53
1959	-12	-2	10	-0.32	-0.74	0.1	2.5	-0.5	-0.10
1960	-12	-29	17	0.73	0.12	-0.3	-3.8	0.1	-0.10
1961	-3	0	3	0.89	1.05	0.3	-1.0	-0.2	0.02
1962	-22	-26	4	-0.12	-0.74	0.2	-3.5	0.2	-0.20
1963	-22	-10	12	-0.01	0.73	-0.1	-1.5	-0.6	0.51
1964	0	-11	11	0.92	1.18	-0.9	2.2	0.5	-0.70
1965	-11	-29	18	-0.11	-0.68	0.2	2.5	-1.5	1.06
1966	-8	-7	1	-0.45	-0.17	-1.0	-3.5	0.0	0.24
1967	-5	-18	13	0.41	-0.14	-0.5	-1.4	0.3	0.05
1968	-16	-32	16	-0.29	-0.51	1.0	-1.0	0.8	0.25
1969	-3	-4	1	0.56	1.28	-0.8	-3.5	-0.4	0.83
1970	-14	-24	10	-0.29	-0.31	-0.1	-2.1	0.1	-0.78
1971	-6	-5	1	-0.33	-0.23	0.5	1.4	0.2	-0.41
1972	-13	-25	12	-0.75	-0.40	0.0	5.1	-1.4	1.21
1973	-5	-7	2	-0.50	-0.88	-0.2	2.2	0.8	-0.88
1974	-13	-27	14	-0.42	0.43	0.4	2.8	0.6	-0.29
1975	-10	-6	4	0.63	-0.08	0.5	-0.7	1.6	-0.81
1976	-3	-18	15	0.45	-0.55	1.0	2.4	-06	0.60
1977	-22	-27	5	-0.77	-0.59	0.5	-0.7	-1.5	0.20
1978	-2	-7	5	0.16	-0.50	0.3	-0.7	0.5	-0.29
1979	-7	-27	20	0.40	-0.73	0.4	-3.6	0.9	0.44
1980	-5	-6	1	-0.90	0.55	0.3	1.2	-0.3	0.65
1981	-7	-24	17	-0.07	0.36	0.1	0.2	1.0	-0.03
1982	-22	-19	3	-0.37	-0.93	0.5	2.8	-1.8	0.96
1983	-5	-20	15	0.77	-0.61	-0.5	1.7	-0.5	1.39
1984	-20	-31	11	-0.47	-1.32	-0.3	-1.1	-0.4	-0.75
1985	-3	-2	1	-0.27	0.04	0.6	2.0	-0.6	-0.70
1986	-5	-23	18	-0.86	0.13	1.2	3.6	0.5	-0.10
1987	-25	-26	1	-0.51	-0.48	-0.5	4.4	-1.8	1.43
1988	-3	-19	16	-0.17	1.37	0.1	-10.3	0.4	-1.95
1989	-14	-27	13	0.87	0.35	0.6	0.5	0.7	-0.18
1990	-6	-5	1	-0.53	0/19	0.0	1.0	0.3	0.18

TABLE 2. Ratios for each of seven indexes of TC activity for those years comprising the 10 highest yearly values versus the 10 lowest yearly values for each of the nine predictor indices occurring during the 41 -year period from 1950 through 1990. For example, the ratio of HDP for the 10 most westerly QBO years to the 10 most easterly QBO years is 2.48, as shown in the upper right of the array.
 
 

	Corresponding ratio of TC activity
Predictor index	NS	NSD	H	HD	IH	IHD	HDP
U 50-mb QBO
10 low / 10 high	1.57	1.97	1.60	2.10	3.29	3.51	2.48
U 30-mb QBO
10 low / 10 high	1.48	1.96	1.66	1.69	2.53	1.63	1.83
U 50-30mb QBO-shear
10 low / 10 high	1.46	1.64	1.50	1.62	1.68	1.36	1.59
Western Sahelian rain-june-July
10 high / 10 low	1.45	1.65	1.47	1.93	3.33	5.57	2.39
Gulf of Guinea rain-Aug-Nov.
10 high / 10 low	1.52	1.86	1.78	2.42	2.94	4.17	2.85
SLPA-June-July
10 low / 10 high	1.21	1.43	1.27	1.68	2.27	3.10	1.97
ZWA-June-July
10 low / 10 high	1.54	1.75	1.78	2.2	3.08	3.76	2.62
SOI-June-July
10 high / 10 low	1.39	1.52	1.29	1.49	1.76	2.26	1.68
SSTA-June-July
10 low / 10 high	1.35	1.32	1.25	1.35	1.75	1.95	1.46

TABLE 3. Listing of the 10 highest and the 10 lowest years for each predictor used to compute the ratios in Table 2.
 

Predictive Parameter	10 high years	10 low years
1. Extrapolated	1952, 54, 56, 62, 63	1950, 53, 55, 61, 64
QBO U50	1968, 77, 82, 84, 87	1969, 76, 78, 85, 88
2. Extrapolated	1954, 56, 60, 65, 68	1950, 55, 59, 61, 69
QBO U30	1974, 77, 79, 84, 89	1971, 75, 80, 85, 90
3. Extrapolated	1958, 60, 63, 65, 68	1950, 55, 57, 66, 69
QBO |U50-U30|	1979, 81, 83, 86, 88	1971, 80, 85, 87, 90
4.) Western-Sahelian rain	1952, 53, 54, 55, 60	1966, 72, 73, 76, 77
June-July	1961, 64, 69, 75, 89	1980, 83, 84, 86, 87
5. Gulf of Guinea rain	1950, 52, 55, 58, 61	1951, 59, 62, 65, 73
Aug-Nov. (previous year)	1963, 64, 69, 80, 88	1979, 82, 83, 84, 87
6. Caribbean SLPA	1968, 71, 74 75, 76	1951, 54, 55, 57, 58
June-July	1977, 82, 85, 86, 89	1964, 67, 69, 83, 87
7. ZWA-Caribbean	1957, 59, 65, 72, 73	1950, 54, 55, 58, 60
June-July	1974, 76, 82, 86, 87	1962, 66, 69, 79, 88
SOI	1950, 55, 56, 68, 73	1951, 63, 65, 72, 76
June-July	1975, 79, 81, 88, 89	1977, 82, 83, 85, 87
9. SSTA (eastern Pacific)	1951, 57, 65, 69, 72	1950, 54, 55, 64, 70
June-July	1976, 80, 82, 83, 87	1973, 75, 84, 85, 88

TABLE 4. Pearson product-moment correlation coefficients (r) for each of the seven measures of seasonal TC activity
with each of the nine seasonal predictors.
 

	NS	NSD	H	HD	IH	IHD	HDP
1. QBO 50 mb	-0.46	-0.54	-0.41	-0.46	-0.48	-0.36	-0.45
2. QBO 30 mb	-0.51	-0.57	-0.54	-0.53	-0.50	-0.36	-0.49
3. shear	-0.20	-0.21	-0.29	-0.24	-0.18	-0.12	-0.19
4. Western Sahelian	0.39	0.53	0.39	0.52	0.64	0.71	0.60
5. Gulf of Guinea	0.51	0.60	0.57	0.59	0.67	0.63	0.63
6. SLPA	-0.33	-0.38	-0.31	-0.39	-0.41	-0.37	-0.40
7. ZWA	-0.38	-0.33	-0.35	-0.40	-0.45	-0.45	-0.44
8. SOI	0.29	0.38	0.31	0.33	0.36	0.30	0.35
9. SSTA	-0.34	-0.19	-0.20	-0.18	-0.22	-0.21	-0.19

TABLE 5. Empirical weights for composite functions of QBO, African rainfall, and Caribbean basin-ENSO data for the jackknifed LAD regression procedure prediction equations.
 

	QBO			Rainfall		Caribbean basin-ENSO
	a1	a2	a3	a4	a5	a6	a7	a8	a9
NS	1,000	0.711	-0.082	1,000	-1.260	1,000	0.106	-0.275	0.046
NSD	1,000	0.560	-0.358	1,000	1.109	1,000	0.096	-0.864	-1.489
H	1,000	1.928	0.284	1,000	17.876	1,000	-1.172	-1.966	-0.025
HD	1,000	0.894	0.266	1,000	0.613	1,000	0.182	0.050	-0.075
IH	1,000	0.614	-0.467	1,000	0.591	1,000	0.051	0.474	0.630
IHD	1,000	0.730	2.777	1,000	0.248	1,000	-0.019	1.272	0.567
HDP	1,000	0.608	0.629	1,000	0.256	1,000	0.169	0.705	0.312

TABLE 6. Agreement coefficient , probability P, and r² values from a jackknifed LAD regression procedure.
 

		P	r2
NS	0.447	0.33 x 10-6	0.433
NSD	0.608	0.14 x 10-8	0.584
H	0.472	0.48 x 10-6	0.455
HD	0.505	0.23 x 10-6	0.558
IH	0.618	0.25 x 10-8	0.677
IHD	0.617	0.50 x 10-9	0.598
HDP	0.556	0.12 x 10-7	0.584

TABLE 7. Measures of agreement p between observed and predicted values associated with nonjackknifed and both grouped (G) and nongrouped (N_G) jackknifed LAD and OLS regression procedures.
 
 

	NS	NSD	H	HD	IH	IHD	HDP
LAD regression model:
1 December forecast results
Nonjackknifed	.440	.514	.447	.493	.498	.451	.457
Jackknified-G	.440	.514	.447	.491	.498	.451	.447
Jackknified-NG	.411	.386	.279	.305	.376	.374	.235
1 August forecast results
Nonjackknifed	.447	.608	.472	.516	.618	.617	.557
Jackknified-G	.447	.608	.472	.505	.618	.617	.556
Jackknified-NG	.186	.435	.256	.210	.544	.499	.325
OLS regression model:
1 December forecast results
Nonjackknifed	.359	.407	.388	.400	.491	.450	.430
Jackknified-G	.308	.359	.339	.355	.448	.403	.385
Jackknified-NG	.244	.306	.280	.300	.408	.367	.338
1 August forecast results
Nonjackknifed	.435	.528	.428	.489	.605	.579	.531
Jackknified-G	.374	.477	.367	.432	.562	.529	.476
Jackknified-NG	.270	.388	.247	.330	.490	.450	.379

TABLE 8. Squared Pearson product-moment correlation coefficients r² between observed and predicted values associated with nonjackknifed and both grouped (G) and nongrouped (NG) jackknifed LAD and OLS regression procedures.
 
 

	NS	NSD	H	HD	IH	IHD	HDP
LAD regression model:
1 December forecast results
Nonjackknifed	.395	.488	.466	.514	.581	.517	.544
Jackknified-G	.395	.488	.466	.511	.581	.517	.527
Jackknified-NG	.368	.358	.280	.320	.415	.431	.276
1 August forecast results
Nonjackknifed	.433	.584	.455	.573	.677	.598	.585
Jackknified-G	.433	.584	.455	.558	.677	.598	.584
Jackknified-NG	.164	.436	.198	.198	.593	.509	.302
OLS regression model:
1 December forecast results
Nonjackknifed	.416	.523	.482	.538	.585	.545	.570
.Jackknified-G	.313	.434	.391	.464	.502	.454	.495
Jackknified-NG	.202	.347	.299	.378	.433	.399	.423
1 August forecast results
Nonjackknifed	.527	.649	.511	.597	.711	.668	.643
Jackknified-G	.406	.564	.399	.504	.646	.577	.556
Jackknified-NG	.255	.430	.154	.321	.528	.449	.394

TABLE 9. Regression weights for the nonjackknifed (LAD) regression procedure prediction equations.
 
 

	0	1	2	3
NS	11.228	0.100	-0.757	-0.931
NSD	68.890	1.029	5.106	-5.309
H	7.410	0.046	0.050	0.239
HD	33.509	0.493	4.503	-4.318
IH	3.459	0.041	1.142	-0.677
IHD	5.023	0.055	6.370	-1.217
HDP	93.364	1.756	31.814	-14.372