Autonomous underwater vehicles do not have sufficient communications bandwidth over long ranges to send back real time images even for monitoring purposes. Autonomous imaging from underwater vehicles will therefore, require realtime imaging system performance prediction in order to ensure that the vehicle can position itself at a range that will allow it to take an image of the scene or target of interest at the required resolution and contrast level. Ideally the inherent optical properties of the surrounding waters should be measured onboard. This may not be feasible or only a restricted set may be measurable. In order to improve the prediction of the imaging performance, a physics-based analytic phase function that could effectively exploit any a priori or in-situ measured parameters would be extremely helpful. Such a new physics-based analytic phase function has been derived and tested against exact scattering codes. Among other features it is sufficiently precise to allow an accurate determination of the backscatter ratio based on an estimate of the mean index of refraction. The new formulation shows clearly why the backscatter ratio, which is the dominant factor in determining imaging range, is insensitive to the inverse power of the size distribution and almost entirely controlled by the mean index of refraction. This new formulation also has a direct application to improve inverse radiative transfer equation (RTE) modeling for estimating inherent optical properties (total absorption and total backscattering) from measured apparent optical properties (ocean color).