Climate classification in the Northern hemisphere using phases of temperature signals

The results of structuring surface temperatures in the Northern hemisphere for the period of modern climate changes are presented. The main idea of the classification suggested is the geographic conditionality of the phase modulation of the temperature signal. The consistency, namely, phasing of the temperature oscillations in certain geographic areas, serves the criterion. We believe that changes in the synchronization modes of climatic processes under changing climate lead to transformations of the spatial structure of the temperature field due to the system transition to a new state. The temperature series are represented as phase-modulated oscillations. External and internal factors that disturb the climate system form a complicated phase modulation, which partly corresponds to these disturbances. The initial space of 818 temperature series is structured into 17 regional clusters, where the temperature changes synchronously. Properties of the resulting clusters and their compliance with the known climate classifications are discussed. The classifying algorithm allows the researchers to choose the degree of differentiation of the field under study depending on the task. The phase modulation indices were estimated to identify manifestations of the external forcing in the surface temperature. Inconsistency of the indices with those in the case of the harmonic phase modulation allows the role of the regional climate factors to be assessed for each class. Modulation, which is the closest to the harmonic one, was found in the North Atlantic thermohaline conveyor. During the study of the climate change, the approach suggested can be used as an analytical framework on any spatial scale, based on only data on the surface temperature, and with predetermined level of quality. The search for synchronization in nonlinear chaotic systems may be one of the promising ways to optimize the predictive models.