Dual-polarization Radar Icing Algorithm (RadIA): Verification/Validation with Research Flights and Application at Military Test Ranges

The 'Seeded and Natural Orographic Wintertime clouds: the Idaho Experiment' (SNOWIE) field campaign was conducted during the winter of 2017 in the Payette Mountains of Idaho. The University of Wyoming's King Air research aircraft carried a full suite of microphysical instrumentation that sampled 24 large-drop, small-drop and mixed-phase in-flight icing cases over 70 flight hours.

In this work, the prototype 'Radar Icing Algorithm' (RadIA) is compared to in-situ microphysical data from the research aircraft. RadIA ingests polarized moment fields from operational S-band weather radars plus a numerical weather prediction model temperature profile as inputs. The height of the freezing level is determined from the moment data and is utilized to make adjustments to the model temperature profile. Non-meteorological targets and radar return at non-freezing heights in the volume are removed. Fuzzy logic membership functions define the presence of freezing drizzle, small-drop supercooled liquid, and mixed-phase conditions. The resulting interest values of each of these four calculations are combined through rule-based thresholding to identify areas of the radar volume where RadIA has a high, medium or low interest in the presence of these various forms of in-flight icing. Particle imagery from the 2DS probe are run through a shape/phase characterization algorithm. Shape/phase data, probe liquid water contents, and parameters of state are averaged over 30-seconds of flight time and matched spatially and temporally to the polar-coordinate RadIA interest values.

From the SNOWIE analysis, RadIA's probability of detecting in-flight icing conditions was about 0.85 out of 1.00 when not blocked by terrain. Median small-drop interest in known small-drop conditions (N=40) was 1.00, which is the maximum possible interest value. Median large-drop interest in known large-drop conditions (N=176) was 0.74. Median mixed-phase interest in known mixed-phase conditions (N=293) was 0.64. Area under the curve receiver operating characteristic values were 0.75 for homogeneous large-drop conditions, 0.73 for homogeneous small-drop conditions, and 0.84 for mixed-phase conditions. These results show that RadIA has good performance at detection of in-flight icing conditions. Further correlation statistics are presented and discussed related to hydrometeor phase at discrete points along the aircraft flight track compared to RadIA's icing classification.

RadIA is next demonstrated on operational S-band and X-band radars on several military ranges located across diverse geographical locations and climates in the Continental United States and Alaska. RadIA will assist range forecasters in providing near-surface and aloft icing information to base personnel and to support range missions. A prototype icing hazard profile product for use at these ranges is presented.