A Multi-Parameter Predictor for Improved Convective Winds Nowcasting at Cape Canaveral Using C-band Dual-Polarization Radar and Environmental Observations

This study presents an overview and preliminary results of the Convective Winds project, a joint collaboration between the 45^th Weather Squadron (45WS) U.S. Air Force Unit at Patrick Air Force Base, NASA Marshall Space Flight Center (MSFC) Earth Science Office, and the University of Alabama in Huntsville (UAH). The 45WS provides weather support to the space program at Cape Canaveral Air Force Station (CCAFS) and NASA Kennedy Space Center. This includes issuing weather warning for hazardous weather to protect personnel, payloads, space launch vehicles, and infrastructure. Weather warnings for convective winds are issued for ≥ 35 kt with a desired lead time of 30 minutes and for ≥ 50 kt with a desired lead time of 60 minutes. The 45WS wants to improve their convective wind warnings, especially by lowering their false alarms and increasing their lead times. Since 32% of convective winds at CCAFS/KSC are ≥ 35 kt, but 50% are ≥ 30 kt, distinguishing between which convective cells will produce winds < 35 kt and ≥ 35 kt is the key. While convective winds ≥ 50 kt are more important to operations at CCAFS/KSC, they occur on only 6% of the events. This project is developing dual-polarimetric radar-based applications in order to improve the capability to predict these hazardous winds. A 5.62 GHz Radtec Titan Doppler Radar (TDR 43-250) and 29 instrumented weather towers, which have an average station density of one tower per 29 km², are used in this study. Storms analyzed are from the 2015 warm season, which includes both convective wind events as well as null cases in order to do an inter-comparison between these storm types. Several parameters that are based on traditional radar reflectivity (Z_h), dual polarization variables such as differential reflectivity (Z_dr), and environmental variables are tested. Some of these parameters that are analyzed in this study include: (1) the vertical extent of the 1 dB Z_dr contour in a Z_dr column at temperatures colder than 0°C, (2) the vertical extent of co-located values of Z_h ≥ 30 dBZ and Z_dr ~0 dB at temperatures colder than 0°C, (3) the peak Z_h at temperatures colder than 0°C, (4) vertically integrated liquid (VIL), (5) vertically integrated ice (VII), (6) the correlation between cloud-top altitude variation and convective wind events, (7) microburst-day potential index (MDPI), and many more. These parameters will be integrated into a predictor for the probability of a threshold level (e.g., 35+ kt) wind event using a simple logistic regression model. Optimal selection of input parameters to the model will be evaluated. The best performing input parameters could then be implemented by 45WS forecasters in order to improve the predictability of threshold level convective wind events. A description of the model, as well as associated methodologies of the various parameters, together with the physical implications of the results, will be explored in this study. Future analysis will include expanding the number of cases and exploring other statistical approaches to find an optimal algorithm for nowcasting convective winds using dual polarization radar and environmental parameters.