This paper presents a non-parametric method for 3-D imaging of natural volumes using Synthetic Aperture Radar tomography. This array processing-based technique aims at characterizing a spatially distributed density of incoherent sources, whose shape is imprecisely known. The proposed technique estimates the moments of the reflectivity density using a low-complexity covariance matching approach, and retrieves the mean location, dispersion, and power of the distributed source. Numerical simulations of realistic tomographic scenarios show that the proposed model-free scheme achieves better accuracy than slightly misspecified maximum likelihood estimators, derived from approximately known distribution shapes.