Climate Audit

I’ve re-collated the hurricane track data from http://www.weather.unisys.com/hurricane/atlantic/2006/index.html and similar tables, updating my earlier collations. I’ve saved these collations in ascii form in the directory http://data.climateaudit.org/data/hurricane/unisys/*.txt using the following nomenclature

hurricane.ATL.txt and Track.ATL.txt with other basins being denoted WPAC,EPAC,SH, NIO and SPAC.

The unisys data contains collated “Best Tracks” information up to 2003-2005 depending on the basin and uncollated track data up to date. I’ve interpolated the newer track data to 6-hour intervals consistent with the Best Tracks data. These are not official interpolations, but are probably decent enough. The “hurricane” tables contain the following information:
id year day_start month_start M name XING SSS class state level peakwind count

Most of the information is collated from unisys; peakwind and count are my calculations from the Track data. The “Track “tables contain the following information (most of the pressure information doesn’t exist).
id year month day qtr lat long wind press

The SH data differs from the Georgia Tech versions, which I’ve also collated in a similar format and will post up.

There has been quite a bit of checking – for example, some hurricane names in the summaries differ slightly from hurricane names in the track directories and I’ve occasionally had to manually edit names e.g. EIGHTEE to EIGHTEEN_E. I haven’t documented the collation but will try to do so at some point.

I use the Track tables in R, where the tapply function trivially yields individual Webster calculations in a single line. I’ll try to look at the new H&W paper, but I’m really trying to do some other things right now; so I’ve posted the collations up for others to use.