2
Storm Selection
This chapter describes typhoons selected for modeling and the procedures
used for selection. The objective was to develop a set of approximately
30 typhoons which are representative of storms affecting flooding at Apra
Harbor, Guam. Coincidently, a similar study of flooding around the island of
Rota, 50 km (31 miles) north northeast of Guam, was ongoing at the time of this
study (Thompson and Scheffner, in preparation). The two islands are sufficiently
close together that they are affected by the same storms. One historical storm
data set served the needs of both studies.
The database of historical typhoons in the western Pacific is available on the
Internet through the U.S. Navy's Joint Typhoon Warning Center (JTWC),
track data covering the years 1945-97 was used. Track data is given at 6-hr inter-
vals, including latitude and longitude of the storm eye (with 0.1-deg precision)
and maximum sustained 1-min mean surface wind, in knots. Western Pacific
storms are identified with prefix BWP followed by a four-digit number: the first
two digits give sequential storm number for the year (01 is the first western
Pacific storm for the year, etc.) and the second two digits give the year. For
example, Typhoon Pamela (BWP0676) was the sixth western Pacific typhoon
during the year 1976. These four-digit numbers are used as storm identifiers in
the present study. Tropical storms originating outside the western Pacific which
may affect the study area have other prefix identifiers. These other possibilities
were considered, but the only such storm important in this study was Paka
(BCP0597), indicating it originated in the central Pacific.
Available information about storm impacts on Guam was also gathered and
reviewed to insure the storm selection process included all important historical
storm events. Sources included JTWC (1991) and JTWC annual and special
storm reports. This review resulted in elimination of one typhoon (Querida 1246)
from consideration because the best track data differed significantly from the
published description of storm track relative to Guam and Rota. Subsequent
discussions with JTWC indicated that storms from before 1959 in the present
data set should be considered less reliable. Other than Querida, pre-1959 storms
were retained for modeling consideration because there were no inconsistencies
evident in the best track files, and it was desirable to preserve the full 53-year
historical database.
6
Chapter 2
Storm Selection
Previous Page
