The climate system is a chaotic, hierarchical open nonlinear system. The system of higher level dominates the behavior of the lower one and determines the basic configuration of climate change. Generally, what the observational data record is the output of lower layer system which necessarily contains the information of interaction between different levels. How to extract information of higher level system from observational data has been a hotspot in the research of nonlinear spatio-temporal distribution theory and methods of observational data processing. A unified form ulation of the climate system has not yet been established. For the convenience of study, in this paper we use a two-level model comprising the Logistic model and Lorenz model, and analyze the output of lower level system through power spectrum and wavelet analysis under different control parameter and obtain the scaling rules for extracting information of the higher level system.