The ocean ambient noise field experiences a stochastic process of many such noise sources and the respective interactions of their wave fields with the waveguide boundaries. At frequencies of about 1 kHz and higher, forward scattering from surface wave can strongly affect shallow water sound propagation. However, most of the available ambient forecasting models do not consider the effects of multiple forward scattering from surface wave. Therefore, there is a need for an accurate method of predicting ambient noises at middle and high-frequency which can account for surface scatterings. Aiming at such a requirement, a propagation model based on transport theory method is described which yields the second-order moment of the acoustic field. Monte Carlo simulations of acoustic propagation loss are employed to validate the transport theory method. The mode number dependence of mode coupling phenomenon is demonstrated at 1000 Hz via the competing effects of mode coupling and attenuation ranges. Low and middle propagating modes are seen to have a smaller coupling range than the attenuation range, allowing mode coupling effects to take precedence over attenuation effects. The mode energies and the coherences are also examined, and it is found that the mode coupling rate for surface wave is significant, but strongly dependent on mode number. Mode phase randomization by surface waves is found to be dominated by coupling effects. On the basis of transport theory propagation model, connecting with the properties of ambient noise sources, a spatial characteristic model for ambient noise under surface wave is presented. Further, the effects of surface wave on ambient noise intensity, vertical correlation and vertical directionality are analyzed. Simulation results show that the surface wave may result in energy transfer from medium modes to low modes and high modes, the rate of energy transfer depends on the mode energy difference. Since the medium mode plays an important role in noise intensity, the noise intensity decreases with the increase of surface wave. In addition to noise intensity, the vertical correlation of ambient noise also decreases due to mode phase randomization by surface wave. Besides, mode coupling can also lead to a change of vertical beam intensity distribution, positive high-angle beams associated with direct, surface, and bottom-surface-bounced rays become weaker, while negative high-angle beams associated with bottom bounced rays become stronger. Since the vertical directionality is sensitive to surface wave, the model can be applied to ocean surface parameter inversion. In summary, the model provided in this paper is closer to actual ocean waveguide and has future prospect in ocean acoustic engineering application.