Geoscience Reference
In-Depth Information
Fig. 3: As for Fig. 2, but for the bi-monthly indices.
Fig. 4: As for Fig. 2, but for the tri-monthly indices.
Table 1: The highest correlation coefficients between the annual number of TCs in
the Australian region and the monthly, bi-monthly and tri-monthly 5VAR, NIÑO3.4
and SOI indices
One month
Two months
Three months
5VAR
-0.668 S
-0.665 AS
-0.655 JAS
NINO3.4
-0.613 S
-0.628 AS
-0.623 JAS
SOI
+0.622 A
+0.640 AS
+0.643 JAS
Letters following the numerical values denote which months or
combinations of months are used.
As can be seen from Table 1, the 5VAR performs better than other two
indices examined demonstrating the strongest monthly, bi-monthly and tri-
monthly correlations. Inspecting the correlations of each index, we found that
the strongest correlation for the 5VAR is one month pre-season September,
while it is the two-month average of August and September for NIÑO3.4 and
the three-month average of July, August and September for the SOI (shown in
bold font in Table 1). Based on these findings, models were built for each of
the indices using the appropriate m -month averages of the data.
 
Search WWH ::




Custom Search