Tracking convective features using the Baddeley image metric
Abstract
Establishing an understanding of whether numerical weather prediction models run with high spatial resolution produce weather systems that have lifetimes similar to observed systems requires an automated method to track and characterize these systems. Such a technique is developed here and uses the binary Baddeley image metric to determine the similarity of various systems of interest across time. This approach implicitly examines the similarity in size, shape, location, and spatial distribution and permits the splitting and merging of systems throughout their lifetime. Ten years of observed and simulated radar reflectivity factor data from April through June are analyzed using this method. Characteristics such as the duration and area covered by the tracks are represented fairly well in the simulation but a disparity between the intensity in the simulations and observations is seen. It is also demonstrated that the duration has a marked impact on several other characteristics of the track. Additionally, the spatial distribution of long-duration tracks indicates that these systems are generally produced in the same locations that they are observed. An evaluation of the speed and direction of movement of the identified systems indicates that, in general, the model produces a similar distribution for direction of movement but with a decreased magnitude.
Degree
M.S.
Advisors
Trapp, Purdue University.
Subject Area
Meteorology|Atmospheric sciences
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.