One of the greatest uncertainties in modeling the ionospheric densities and temperatures lies in the neutral winds that affect the vertical ion drift. Wind measurements are difficult to make, and current wind models are not adequate. An alternative is to deduce an equivalent meridional neutral wind from meausrments of the height of the F2 peak electron density. This method has been effective in allowing ionospheric modelers to reproduce the observed height with reasonable precison when the peak height does not vary rapidly. However, when there are rapid movements of the peaks height, this method underestimates the magnitude of necessary wind changes, and it is also complicated and computer intensive to do accurately. This paper introduces an improved algorithm for use in time dependent ionospheric models that reduces the complexity of the procedure while more accurately reproducing the observed heights. The improvement is achieved by estimating the wind that is needed to reproduce the observed height at the next time step using the calculated height and wind at the current time step. Thus the winds are continuously being adjusted to bring the calculated height into better agreement with the observed height during the time dependent simulation. The winds from the new algorithm agree well with both optical and radar measurments made at Arecibo on August 17--18, 1982. ¿American Geophysical Union 1991