Global Cyclone Climatology
from the NCEP Reanalysis

Data Details FTP Site Trends

A 40-year climatology of cyclonic systems has been developed based on the 1000 mb and 500 mb geopotential height data from the National Center for Environmental Prediction (NCEP) and NCAR "reanalysis" data set. Cyclones at the two levels were identified automatically and tracked from one 12-hour grid to the next. The data have been used for the analysis of trends (Key and Chan, 1999), but could also be used to formulate climatologies of cyclone frequencies, cyclogenesis, and tracks.

The data set available here consists of (1) the times and locations of individual cyclone centers at the two levels, and (2) individual cyclone tracks. See the Data Details page for more information.

Algorithm

Twice daily, global 1000 mb and 500 mb geopotential height data from the National Center for Environmental Prediction (NCEP)/National Center for Atmospheric Research (NCAR) 40-year Reanalysis Project are used in the identification and analysis of cyclones. The data cover the period 1958-1997 with a grid resolution of 2.5 x 5 degree latitude-longitude.

Closed cyclones are defined by having at least one closed 30 m contour around a central minimum value. If the geopotential height along lines extending from a grid cell increases 30 m or more before falling or reaching a distance limit, it is recorded as a closed cyclone. Cyclones are tracked individually over their lifespan. Each cyclone is found in subsequent grids until no further continuation is detected. The grid cell (cyclone center) of the first cyclone in a track is a position of cyclogenesis, and that of the last cyclone in a track is one of cyclolysis. Figure 1 gives an example of the identification results, showing 1000 mb cyclone centers in the northern hemisphere on June 30, 1958 at 0Z. In all analyses "frequency" is the total number of closed cyclone centers in all 12-hour grids. For example, if a region had 10 cyclone centers on one 12-hour grid and 12 on the next (some of which will be the same cyclones), then the frequency for that day in that region is 22.

High-latitude and midlatitude cyclones identified in this study are wave cyclones. Those detected in the tropical regions are primarily the large-scale thermal lows over India (summer monsoon) and Argentina, and might be more accurately labelled as "subtropical". Due to the relatively coarse resolution of the sampled data, hurricane frequencies are not well represented.

Examples

Figure 1 shows cyclone centers identified by the algorithm described above for a single day over the northern hemisphere. Figure 2 shows cyclone frequencies for summer of the 40 year period.

Figure 1. An example of cyclones detected by the automated method for a single day in the Northern Hemisphere. The black diamonds represent the centers of the cyclones.

Figure 2. Total number of twice-daily grids in which a cyclone center was identified in the 500 mb geopotential height data during the northern hemisphere summer (June, July, August), 1958-97.

References

Key, J. and A. Chan, 1999. Multidecadal global and regional trends in 1000 mb and 500 mb cyclone frequencies, Geophys. Res. Lett., 26(14), 2053-2056.