In this paper, we develop a series of exponential probability density functions for modeling different distributions of elevation angle (EA) of arrival signals in different tree-dimensional (3D) coverage area scattering environments, and implement the modeling of channel characteristics. First, by assuming that the distribution of azimuth angle is uniform, in this paper the closed-form expressions of power spectrum density (PSD) for both symmetric and asymmetric situations of EA are derived. It can be observed from the analysis results that the PSD is closely correlated to EA function and the boundary angles βmin and βmax of the arrival signals and also to the Doppler shift. Then the spatial fading correlation (SFC) of MIMO multi-antenna signals in 3D environment is derived and simulated. The results show that the SFC between MIMO multi-antenna elements is closely related to βmin and βmax, and the parameter of EA function has little effect on SFC. The exponential EA probability function which is introduced in this paper can be applied to channel parameter estimation of multiple wireless communication environments. Compared with traditional models, this model presents the parameter estimation that satisfies theoretical and empirical values, and this model also expands the modeling of statistical channel in 3D environment.