We present calculations of the altitude at which the micrometeor phenomenon begins, that is, the point where the interaction between micrometeoroids and the Earth's atmosphere becomes important. At these altitudes, physical processes such as light emission, heating, electron production, etc., begin to occur. The calculations are performed using four very different initial height models including (1) solving the full integration of the single-body meteor equations, (2) using a balance between the loss of momentum and the component of the acceleration due to gravity along the meteor trajectory, (3) using a solution that emanates from a linearized form of the meteor energy equation but without including either atmosphere or meteoroid radiation emission effects, and finally (4) utilizing a solution of the meteor energy equation that is specifically approximated for small particles. We compare our evaluated theoretical results with direct micrometeor observations detected using the 430 MHz Arecibo Observatory (AO) radar system. The goal of these calculations is to provide reliable initial conditions in order to completely model the AO micrometeor observations, most of which have nearly constant decelerations. The nature of this study, although performed with already existing theoretical formulations, is of unprecedented value because it is the first study where these models are directly compared against very highly resolved micrometeor velocity and altitude distributions that are derived directly from the radar observations. We found that the meteor energy equation approximated for small particles agrees very well with the radar observations, in particular for meteor melting temperatures of the order of 2100 K and entry angles lower than 30¿ with respect to the radar beam normal direction. Unfortunately, from this model the composition characteristics of the particles detected by the AO radar cannot conclusively be drawn. However, comparison with the calculation of the penetration height of meteoroids reported by ReVelle (2005a) suggests that chondritic material seem to be the best candidate to explain the observed penetration of these particles in the mesosphere. Calculation of the light emission and electron density production of the meteor are also presented and discussed.