In the field of hydrogeophysics, images are often used to communicate information about subsurface properties and processes. Such images are derived from measurements, modeling, or a combination of the two and inherently contain a varying degree of uncertainty. This poses a challenge, since not only should the interpreted results be conveyed in such images, but also the degree of uncertainty in the representation. The question we seek to address is: How do we generate a single image that communicates the uncertainty in the interpretation? In an attempt to answer this question, we focus, as an example, on a hydrostratigraphic model obtained from airborne electromagnetic data calibrated with lithologic data. The uncertainty in this model stems (mainly) from the uncertainty of accurately assigning the electromagnetic data to unique lithologic units. Two representations of the uncertainty, based on color schemes, were tested. The first approach uses an additive color model, the RGB model, and consists of directly mapping the data to the different primary colors. The second approach utilizes basic concepts from color theory, in particular, the use of harmonic color schemes, and is based on a subtractive color model. We find that the uncertainty in describing the lithologic units is difficult to distinguish using the direct RGB mapping, but, by using the harmonic mapping, the uncertainties are recognizable. We conclude that the use of harmonic color schemes is a viable way of communicating uncertainties in the subsurface.