JISAO data
Tropical cyclone positions

Analyses | Data
Distribution of hurricanes
big GIF | JPEG | PS

Daily hurricane positions during 1970-89: red and turquoise dots for the first and successive days with hurricane force winds (>64knots), respectively.
ENSO influence on Atlantic hurricane landfall, 1886-2000
big GIF | JPEG | PS

Hurricane positions on the last day that they exhibit hurricane-force winds during the (left panel) 25 warmest and (right panel) 25 coldest years in terms of sea surface temperature in the equatorial cold tongue region (6°N-6°S, 180-90°W) based on the period of record 1886-2000.
ENSO influence on tropical cyclones and hurricanes, 1949-2000
big GIF | JPEG | PostScript
Daily hurricane and tropical cyclone positions during the (top panel) 10 warmest and (bottom panel) 10 coldest years in terms of sea surface temperature in the equatorial cold tongue regions based on the period of record 1949-2000.

values range from 0 to 8
big GIF | PS | JPEG .
This analysis is derived from the Atlantic Hurricane Re-analysis Project, which includes previously unavailable observations for 1851-85, and improvements in characterization of storms in 1886-present.
Four-times daily cyclone position data is available from the NOAA National Hurricane Center (Atlantic and eastern Pacific only) and from the National Center for Atmospheric Research (all oceans). I have taken the original four-times daily data in various formats and written a table in a single format with one record each day for each storm.

The notes on the data set at the NCAR WWW site are excellent, and you should read them if you want to use this data set.

Data:
1) Monthly number of Atlantic hurricanes and named tropical storms for 1948-present. Calculated by Cathy Smith of NOAA Climate Diagnostics Center.

2) Daily cyclone positions as ASCII tables

  • Atlantic Ocean, Caribbean Sea, and Gulf of Mexico 1851-2000
    This is the hurricane Atlantic Hurrican Re-analysis Project data set, which seems to have the same number of hurricanes during the original analysis period (1886-present), but has added observations for 1851-1885. My impression is that there are improvements to the characterization of storms during the period 1886-2000.
  • eastern north Pacific Ocean 1949-2000
  • western north Pacific Ocean 1945-1994
  • Australia region (western south Pacific Ocean) 1958-1994
  • north Indian Ocean 1877-1994
  • south Indian Ocean 1877-1994

    
 The columns in each table are:
   1     2   3     4   5   6       7       8    9  10
 storm year month day lat lon storm_type speed slp SS

where
  storm is the number of the storm for that basin.  For the Atlantic
    this number ranges from 1 to 959.
  year and month are the year and month that the storm was first
    identified in.
  day is the number of days that it has been a storm.  "Day" is not
    the day of the month.
  lat and lon are the latitude and longitude of the storm.
  storm_type 1 hurricane (maximum wind > 64 knots)
             2 tropical storm (34 knots < maximum winds < 64 knots)
             3 unknown speed (early records) or storm < 34 knots maximum winds
  speed slp are the wind speed (knots) and sea-level pressure (mb).
    NaN for missing.
  SS, the Saffir-Simpson hurricane scale.
     value    maximum winds 
       0       speed < 64 knots
       1        64 <= speed < 82
       2        82 <=       < 96
       3        96 <=       < 113
       4       113 <=       < 135
       5            speed > 135
    I have only retained one observation per day for each storm in the data set to make it more compact. For each storm, I determined the last observation with the highest wind speed, and used observations for that time of the day as an indicator of the storm position. This choice was so I could could determine landfall with the greatest precision. Frankly I don't think that this choice is a big deal. For storms without wind speed estimates, I used the time of the day of the first observation.

    October 2005
    Todd Mitchell (mitchell@atmos.washington.edu)
    JISAO data