The NASA Icing Remote Sensing System is a prototype that fuses together the output from a laser ceilometer, a multi-channel radiometer and a Ka-band Doppler radar to create an in-flight icing hazard index and has been in development by NASA and NCAR since 2004. A temperature profile and integrated liquid water value are derived from the radiometer, and the vertical cloud boundaries are defined from the radar and ceilometer. The algorithm distributes the detected liquid into a profile of liquid water contents and creates an in-flight icing severity product at altitudes where liquid water exists and the temperature is below freezing. The current algorithm derives a liquid profile through a combination of fuzzy logic input weights based on liquid's relation to reflectivity, temperature, a wedge-shaped profile and a uniform liquid distribution with height. Previous research has shown that the existance of dual fall velocity modes in the Doppler spectra could be useful in discerning the existance of supercooled small drops (< 50 ìm). The authors present case studies to show how supercooled liquid can be detected and ranged in a mixed-phase profile, and thus the icing severity product can be improved by incorporating a Ka-band Doppler fall velocity fuzzy logic weighting functions.