Limb sounders measure atmospheric radiation that is dependent on a radial and angular distribution of temperature and atmospheric constituents. Based on this dependency, we perform and characterize both two dimensional (radial and angular) and one dimensional (vertical) limb profile retrievals of a distribution of carbon monoxide (CO), which represents a simulated plume off the coast of south-east Asia. Both the one dimensional (1-D) and two dimensional (2-D) limb retrievals are characterized by evaluating their averaging kernels and error covariances on a radial and angular grid that spans the plume. This 2-D characterization allows the radial and angular resolution as well as the degrees of freedom and information content to be computed for these limb retrievals. We apply this 2-D characterization of a limb retrieval to a comparison of the 2D retrieval with the 1-D (vertical) retrieval. We find the information content for the 2-D retrieval is an order of magnitude larger than that of the 1-D retrieval and the degrees of freedom is about a factor of two larger. We also use the 2-D averaging kernel to develop a strategy for validation of a limb retrieval with an in-situ measurement.