Present approaches for site effect evaluation are either theoretical or empirical. The ratio of microtremor horizontal to vertical spectra of one station (H/V method or H/V) is empirical. The highest amplitude of H/V is considered to be site amplification and the lowest frequency corresponding to the largest H/V amplitude is considered to be the site predominant frequency. In practical

the H/V method of microtremor measurement provides accurate site amplification and site predominant frequency. However

many engineering scientists considered that the H/V method lacks sufficient theoretical background

because numerous assumptions were introduced in formulating the method. Some assumptions are discussed in this paper. The microtremor spectral ratio method is assumed that the spectra ratio of the horizontal component to the vertical component of microtremors on the rock surface is equal to unit. Is this correct? There is no satisfying theoretical explanation. Microtremors are been divided into body microtremors and surface microtremor. In order to reflect the random of microtremors

body microtremor is considered by superposing inclined P-SV waves under different incident angles. Surface microtremor is assumed to be Rayleigh wave on elastic half space. Expressions of the spectra ratio of the horizontal component to the vertical component of body microtremors on the elastic half rock surface are deduced in this paper. The characteristics of the spectral ratio of the horizontal to vertical spectra on the Poisson rock surface are discussed. It is shown that the H/V is close to unit for body microtremors and approximately 0.68 for surface wave microtremors on the Poisson rock surface.