A Study of Stochastic Resonance in a Climate model

Historically, the earth’s fluctuation between interglacial and glacial climates has been observedto have a period of 105 years [1]. However, simulations of the global average temperature didn’tmanage to reproduce this cycle period until 1982, when Benzi et al. [2] introduced the combinationof long-term variations in incoming solar radiation and stochastic noise in an energy balancemodel. Using an energy balance model means that the change in global average temperature isset as proportional to the difference in ingoing and outgoing energy. The result of the simulationsdemonstrated so-called stochastic resonance, where small stochastic perturbations amplified thepattern of the variation in insolation, causing a pattern of large changes in the global averagetemperature, i.e. changes in the climate. The stochastic perturbations model unpredictable shorttime scale phenomena like the weather. Our study aimed to reproduce the result of Benzi et al.[2] and to investigate the model and its parameters. The presence of a 105-year climatic cycle insimulated data was found. The combination of both noise and varying incoming solar radiationwas necessary to observe the 105-year cycle. The characteristics of the climate cycle pattern did,however, vary greatly depending on the values of constants in the model, illustrating how themodel and constants were imprecise. Therefore, no conclusions can be drawn from this studyabout the earth’s current or future climate. However, the study still confirms that stochasticnoise is an important part of modeling the climate, and manages to simulate the earth’s observed105-year climate cycle