On the incompleteness of the historical record of North Atlantic tropical cyclones

• Published in: Geophysical Research Letters
• Main Authors: Solow, Andrew R., Beet, Andrew R.
• Format: Article in Journal/Newspaper
• Language: English
• Published: American Geophysical Union 2008
• Subjects: Poisson regression model; Observation probability; North Atlantic tropical cyclones; North Atlantic
• Online Access: https://hdl.handle.net/1912/3361

Author Posting. © American Geophysical Union, 2008. This article is posted here by permission of American Geophysical Union for personal use, not for redistribution. The definitive version was published in Geophysical Research Letters 35 (2008): L11803, doi:10.1029/2008GL033546. There is some question as to whether the historical record of observed North Atlantic tropical cyclones prior to the advent of satellite coverage is complete. This question is central to understanding the historical trend in tropical cyclone activity and the effect of environmental factors on it. To address this question, a statistical model of the relationship between annual cyclone counts between 1870 and 2004 and sea surface temperature and the state of the Southern Oscillation is extended to allow for non-decreasing observation probability prior to 1966. The estimated observation probabilities increase from 0.72 in 1870 to 1 in 1964. Allowing for record incompleteness reduces the estimated effect of sea surface temperature on annual tropical cyclone activity. This work was supported by NOAA Grant NA17RJ1223.