The shear-wave velocity (Vs) of soils has been acknowledged among the most critical factors affecting the nature of seismic surface motions, known as seismic site effects, in the course of probable earthquakes. Together with the one-dimensional (1D) Vs profiles of soils, the lateral heterogeneity of velocity contrasts may strongly impact seismic ground motions within the surface. Such conditions seem to occur in the city of Tehran, Iran, mainly in the eastern part, Hakimiyeh Valley, wherein alluvial deposits have been settled between the Alborz and Sepayeh (the Anti-Alborz) mountain ranges. In this resepct, the significant differences observed between previous analytical and empirical studies doubles the necessity of estimating real bedrock depth and Vs structure in this area. Given that the improper analysis of seismic site effects, using unrealistic dynamic characteristics, gives rise to the low-accuracy prediction of seismic hazards as well as significant human and financial losses; this study is an attempt to extract the two- dimensional (2D) Vs structure of the subsurface layers in the form of two north-south and east-west profiles, through the single-station and array measurements of ambient noises (viz., microtremors), as the first deep alluvial models in this area. For this purpose, the ambient noises are recorded with reference to 4 and 11 array and single-station measurements, respectively. A circular layout, comprised of two concentric circles, is further considered for the array measurements with 11 three-component narrow-band seismometers, Lennartz LE-3D/20S, within each array. The recorded array seismograms are then analyzed via the spatial autocorrelation (SPAC) method and the frequency-wavenumber (F-K) analysis, and all seismograms recorded in this line are subsequently evaluated by the time-frequency analysis (TFA) of ellipticity in the Geopsy software. Afterward, the results are utilized as the objective functions (OFs) in the inversion operation for calculating the Vs profiles at the selected stations. The neighborhood algorithm is further applied for the inversion of the OFs and the extraction of the Vs profiles. The study results as it comes to the extraction of the 2D profiles and the recognition of the Vs structure in this area, along with the velocity of 2200 m/s, correspond to the seismic bedrock located in the deepest part of the sedimentary basin, 800 m. Moreover, the underground topographies, such as faults, can be identified in the Vs profiles by means ofusing the ambient noises. As a final point, the profiles indicate that the variations in the Vs and dynamic properties of the soil are significant in this area, as one with a complicated subsurface structure. The dominant frequencies of the sediments in Hakimiyeh Valley, reported in previous analytical research, are also significantly different from that those extracted in this empirical study. As an illustration, no sharp and clear peaks can be spotted in the analytical amplification functions in Hakimiyeh Valley, while there are distinct peaks below 2 Hz at most stations in the present study. Furthermore, the study results are in good agreement with the geology of this area, and pave the ground for analytical site effect studies in the future. [ABSTRACT FROM AUTHOR]