A correction to the definition of the atmospheric coherence diameter is suggested here. The basis of this new approach is the existence of an aerosol MTF which is often the dominant ingredient of the atmospheric MTF. As defined by Fried about 25 years ago, atmospheric MTF was related to turbulence MTF only. In the case of a Gaussian approximation of the aerosol MTF, an analytical expression is derived for the aerosol-caused coherence diameter. This parameter is related to the aerosol MTF's cutoff frequency, and to its asymptote at high spatial frequencies which was recently shown to be higher than the atmospheric transmittance. Qualitative validation of the theory is presented, based on measured MTFs in the open atmosphere. Overall atmospheric coherence diameter is generally the smaller between the turbulence and aerosol coherence diameters.