1 Introduction 



During the winter of 1991-92, two intense storms, one large and of long 

 duration, the other small and of short duration, caused extraordinary damage 

 along the eastern seaboard of the United States. Specifically, the storms 

 occurred on 29 October to 2 November 1991, and 4-5 January 1992. The 

 most notable was that of 29 October to 2 November 1991, commonly referred 

 to as the "Halloween Storm." In terms of its long duration and great extent of 

 affected coastline, the Halloween Stonn was indeed very unusual. Using his- 

 torical water level data as a basis, at some locations it was the event of record 

 (e.g., Nantucket, MA). At others it was calculated as relatively frequent, a 

 l-in-5 or 1-in-lO year event. Yet, there is general agreement in the geoscien- 

 ces community that storms such as the Halloween Storm occur less than once 

 in 20 or perhaps 25 years. Qearly, as estimated from coastal water level ele- 

 vations, the recurrence interval assigned to a storm depends upon the location 

 of the measurement site in relation to the storm and therefore is not unique. 



This report will explore factors other than coastal water levels which can be 

 used to assign "frequency of occurrence" or "return interval" to events such as 

 the Halloween Storm. The approach used focuses on characteristics rather than 

 effects of the event. These characteristics include the duration, intensity, size, 

 area of origin, and track. In addition, meteorological records for approxi- 

 mately the past 30 years were examined for systems which exhibited track 

 characteristics similar to those of the Halloween Stom. These systems were 

 selected based upon genesis location within a specific geographic region and 

 track characteristics similar to that of the Halloween Storm. Correlation of 

 other characteristics will require further study and wiU be explained later. 



Chapter 1 Introduction 



