Identifying various impairments is critically important in optical channel monitoring. We present a new approach for analyzing and characterizing the amplitude histogram of optical signals. This new approach is based on the X2 distance, which was originally used in digital image processing. To effectively distinguish various impairments, we introduce two different metrics: (1) the difference between the X2 distance of the mark histogram and the X2 distance of the space histogram, and (2) the asymmetry property of the mark and space histograms. Using these two metrics together, we show by simulation that three major impairments, namely, inband crosstalk, amplified spontaneous emission (ASE) noise, and chromatic dispersion can be effectively identified.