ISSN 0869-6632 (Print)
ISSN 2542-1905 (Online)


Cite this article as:

Вакуленко Н. В., Seryh I. V., Sonechkin D. M. Chaos and order in atmosheric dynamics part 3. Predictability of el nino. Izvestiya VUZ. Applied Nonlinear Dynamics, 2018, vol. 26, iss. 4, pp. 75-94. DOI: https://doi.org/10.18500/0869-6632-2018-26-4-75-94

Published online: 
31.08.2018
Language: 
Russian
UDC: 
551.465

Chaos and order in atmosheric dynamics part 3. Predictability of el nino

Abstract: 

Topic. Based on the assumption that short-term climatic variations are nonchaotic, and, therefore, the paradigm of the limited predictability of weather formulated by E.N. Lorenz is not applicable to these variations, a question is posed about the unlimited predictability of the short-term climatic variations. It differs from the opinion generally accepted in climatology now that atmospheric motions of all time scales, beginning from daily weather variations, and including interannual, centennial and even millennial variations of climate are unstable. Aim. Specifically, the interannual scales are considered in this paper, and the predictability of the well-known phenomenon of El Ni ?no is investigated. For this purpose, the so-called Global Atmospheric Oscillation (GAO) is considered which has been recently recognized by climatologists. GAO represents a synchronized integrity of the well-known processes in tropics connected with El Ni ?no, and some extratropical processes. Method. Assuming GAO to be the main mode of the short-term climatic variations, some indices are defined which characterize the dynamics of GAO itself as well as the interrelations between the extratropical and tropical components of GAO with each other. It turns out that crosscorrelations exist between these indices which are so high that they may be considered as evidences of some one-to-one relationships between the tropical and the extratropical components of GAO. Results. It allows give a positive answer to the question posed on nonchaoticity of the short-term climatic variations. Among the indices characterizing GAO there is one by means of which it is possible to predict El Ni  ?no with the lead time of 14 months. Then, by means of a specially designed technique of the crosswavelet analysis of pairs of time series, a range of time scales is found in which the closest crosscorrelations exist of the index-predictor with an index characterizing El Ni ?no itself. This time scale range includes within itself all known El Ni ?no rhythms, i.e. the time periods from 2 to about 16 years. Discussion. As a result, it is indicated a possibility of a further increase in the lead time of the of El Nino prediction up to several ? years. It is much more, than the lead times of all present-day hydrodynamical and statistical forecasts of El Ni ?no.  

DOI: 
10.18500/0869-6632-2018-26-4-75-94
Short text (in English):
(downloads: 13)
Полный текст в формате PDF(En):
(downloads: 31)