Exploring possibly increasing trend of hurricane activity by a SiZer approach

Statistical studies of hurricane activity in the North Atlantic basin are of current interest, not the least after the year of 2005 when many records were broken. An exploratory analysis of possibly increasing trend of the yearly number of hurricanes can be carried out by employing a graphic device,...

Full description

Bibliographic Details
Main Author: Jesper Rydén
Other Authors: The Pennsylvania State University CiteSeerX Archives
Format: Text
Language:English
Subjects:
Key words
Kernel estimation
Local regression
Smoothing
Poisson regression
Cli
North Atlantic
Online Access:http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.66.9237
http://www.math.uu.se/research/pub/Ryden1.pdf
Description
Summary:Statistical studies of hurricane activity in the North Atlantic basin are of current interest, not the least after the year of 2005 when many records were broken. An exploratory analysis of possibly increasing trend of the yearly number of hurricanes can be carried out by employing a graphic device, the SiZer map, for visualization of features with respect to both location and scale. This is a smoothing method based on scale-space ideas. Use of this method provides a useful complement to other statistical techniques for detection of trends or shifts in time series.

Similar Items