Climatology of Caribbean Hurricanes
	| home | tools | pleas for help | QHWRN | guide | climatology | archive |

First Storm of the Season

Suppose, it's already August, and still no hurricane on the horizon. Is this normal? Is this a good sign? Is there any relationship when the first storm forms and the number of storms for that season? Read on for the answers!

The graph above shows when the first tropical system has developed in the complete Atlantic and for storms that traveled through the Eastern Caribbean (see orange box). The number of years are counted (since 1851 through 2005) for each week of the season that the first storm appeared. For example, for the complete Atlantic the first storm appeared in the week of June 18 for a number of 17 years since 1851. Quite often the first named storm forms in the middle of June or beginning of August (week starting August 6). The first tropical system (subtropical storms are ignored) ever to appear in the Atlantic was on February 2 in the year 1952, other early storms appeared on March 6 in 1908 and April 18 in 2003 (Ana) [storms before April 1 not shown in graph]. On the other hand, in 1914 the first storm didn't appear until September 15! See the table below for all storms that formed since 1851 before June 1, the official start of Atlantic Hurricane Season.

Storms passing through the eastern Caribbean have normally formed in the central Atlantic. Early in the season that area is not conducive to tropical cyclone formation, since water temperatures are too low. Therefore it is normal to not see a storm in that area until the beginning of July. During 27 years since 1851 no storms at all passed through this region. In most years the first storm forms in the week of August 13. In 1908 however, the Islands saw the first storm form in March! That was the only storm for the Eastern Caribbean that formed before June 1, which eventually became a Category 2 hurricane (yes, March!). This one passed over St.Kitts, just east of St.Barts. While in 1923 the first storm which passed through the eastern Caribbean didn't form until October 24. Another notable exception was 1999. It looked like the season was over without any storms until October 17. Then Jose passed by. And even later Lenny. So a late start seems no guarantee for a 'quiet' season. You only need one bad storm to spoil a whole season. More below.

The plot above shows the cumulative distribution of the date of the first tropical system formed. It shows that for the whole Atlantic the first storm has formed before June 12 for 25% of the years since 1851. For half of the years (50%) this was July 1, and 75% August 6. Since the last recorded date for the first hurricane to form is September 15 (see above) the 100% level is reached at that date. For the Eastern Caribbean only, for 25% of the years the first storm has formed by August 4, the 50% level is August 19, and the 75% level September 3.

The first plot for the Atlantic looks bi-modal (ie., it has two peaks). Tropical systems basically form in two areas; the Gulf of Mexico/Western Caribbean Sea and Central Atlantic. Early in the season most storms form in the Gulf or W. Caribbean, later on, when the central Atlantic warms up, more storms form there (for more details see a separate section). If for a number of years there is no activity in the Gulf or W.Caribbean then the first storm will have originated at a later date in the central Atlantic (second peak). Storms travelling through the eastern Caribbean are usually formed in the central Atlanitc. Therefore it is probably no coincidence that the second peak overlays more or less with the eastern Caribbean first time activity. The really interesting thing would be to find an explanation why sometimes the Gulf of Mexico and W.Caribbean Sea are so quiet...

And does a late start of the season forbode a 'quiet' season? The answer is no! Above are the number of storms plotted corresponding to the date the first storm formed for the complete Atlantic. For example, the first storm ever to form was on February 2 (in 1952, see above). In that year only 6 more storms developed (so a total of 7). This is below the long term average of about 10. As can be seen on the above graph even when the first storm of the season doesn't form until the end of August, the total number of storms during that year can also be over 10. There is a very weak (statistically not significant) correlation with number of storms and time the first one formed. The later in the season the first storm develops, the less storms form during that year. So despite a shorter 'hurricane season' left when the first storm forms 'late', this does not really translate in less storms for that season.

Does it make any difference when we only focus on the eastern Caribbean? No, not really, as can be seen on the graph below. The correlation is a little bit stronger (but still statistically not significant). There is a potential for more storms when the first storm appears early, but even in years when the first storm was very early, e.g. the March storm in 1908, the total number of storms can be well below normal. Also, when the first storm forms after September 6 or so, it will probably be the only one for that year, or at most one more can pass through the island. But again, you only need one storm to spoil a whole season.