Premium
Automatic estimation of tropical cyclone intensity using multi‐channel TMI data: A genetic algorithm approach
Author(s) -
Kishtawal C. M.,
Patadia Falguni,
Singh Randhir,
Basu Sujit,
Narayanan M. S.,
Joshi P. C.
Publication year - 2005
Publication title -
geophysical research letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.007
H-Index - 273
eISSN - 1944-8007
pISSN - 0094-8276
DOI - 10.1029/2004gl022045
Subject(s) - tropical cyclone , intensity (physics) , algorithm , channel (broadcasting) , storm , genetic algorithm , sensitivity (control systems) , mean squared error , meteorology , computer science , environmental science , remote sensing , statistics , mathematics , geology , machine learning , geography , engineering , physics , optics , computer network , electronic engineering
An automatic method for intensity estimation of tropical cyclones using multi‐channel observations from TRMM Microwave Imager (TMI) is developed using a non‐linear data fitting approach called Genetic Algorithm. The intensity estimation technique SIEGA (Storm Intensity Estimation using Genetic Algorithm) uses only 9 simple statistical variables based on TMI observations and does not require any subjective input except the center of the cyclone. SIEGA was trained using 91 randomly arranged TMI scenes corresponding to several tropical storms (1998–2002), and produced a root‐mean square error of 13.88 kt, and average absolute error of 10.71 kt when tested on an independent test sample of 230 TMI scenes. Sensitivity of SIEGA's performance to the uncertainties in the location of TC center has also been carried out.