SCOOP STORM DATABASE
The bulk of the information in this database is derived from the storm data provided annually by the National Hurricane Center (NHC) in HURDAT format. The HURDAT file contains storms from 1851 to the previous storm season, but is not updated during the current storm season.

We have augmented these data with information about tropical depressions that were recorded during the season, but not (for whatever reason) considered "significant" enough to include in the HURDAT file. For years 1995 through 2005 we have full track information for these events which are listed with names such as "TD01", "TD12", etc. For years 1991 through 1994, only the starting and ending dates of the events were available, so there is no track information and the time values were arbitrarily set at 1200Z. There are no additional tropical depression data prior to 1991.

We encountered a problem with storm numbering: The HURDAT file contains one numbering scheme; however, the HURDAT data have been augmented by NHC in post-season analysis to include previously-undetected storms. These storms are inserted into the existing order with names such as "SUBTROP 1", "SUBTROP 2", etc. The addition of these storms upsets the numbering scheme that is popularly used. Our addition of the unlisted tropical depressions upsets the numbering scheme even more.

To solve this problem we have insituted three parallel numbering schemes:

  • The "stormid" is a consecutive number beginning with one that counts all storms for the year. Its format is yyyynn, where yyyy is the year.

  • The "hnumber" is the storm number assigned by the NHC in the HURDAT file. The tropical depressions added by us do not have an "hnumber" value, so "hnumber" will typically lag behind the "stormid" values. Its format, however, is the same.

  • The "pnumber" is the "popular" storm number that was assigned to the storm during the season. This number does count the added tropical depressions, but does not count the extra "SUBTROP" storms added by the NHC in post-season analysis. "pnumber" will also lag behind the "stormid" values, and it uses the same format.

We have elected not to display any of these numbers on these web pages, nor is it possible to search for a storm by number. This is to avoid some confusion. You may, however, search for a storm by year and name (e.g., "2005" and "KATRINA"). The underlying search service does return all three sequence numbers which may be used by applications that desire this information.

We plan to update the storm database in near-real-time during future and current storm seasons.