During the warm season, nocturnal bow echoes, bowing segments of convection, frequently generate some of the most damaging straight-line surface winds, even in situations where elevated convection occurs and the threat of damaging winds would normally be thought to be lower. The prediction of the strength of bow echoes is of crucial importance to operational meteorologists and the public. Moreover, correct forecasts of wind intensity of nocturnal bow echoes tend to be more challenging for operational meteorologists than daytime bow echoes. To address this forecast problem, this work analyzes multiple meteorological variables obtained from the Storm Prediction Center mesoanalysis system to investigate differences in warm-season nocturnal bow echo environments that produce severe winds versus those that do not. A sample of 132 cases from 2010 to 2018, was classified into high intensity (>70 knots) severe winds (HS), low intensity (50-55 knots) severe winds (LS), and non-severe winds (NS). A total of 43 forecast parameters were analyzed over a 120 x 120 km region centered on the highest storm wind report or most pronounced bowing convective segment.

Results suggest that shear-related parameters discriminate only between severe and non-severe cases, while CAPE-related parameters discriminate only between HS and LS/NS bow echoes. Severe composite parameters are among the best discriminators between all severity types, especially Evans Derecho Composite Parameter (DCP) and Significant Tornado Parameter (STP); on the other hand, CIN-related parameters are among the worst discriminators for all severity types. The most skillful forecast parameters for HS bow echoes are STP and most unstable CAPE, while for LS bow echoes are V wind component at best CAPE (VMXP) level, STP and Supercell Composite Parameter. Results also indicate that combining surface-based CAPE with 0 – 6 km U shear component, and DCP with VMXP, resulted in the most skillful HS and LS forecasts, respectively. These findings may help forecasters identify long-lived, high intensity bow echoes from low intensity severe and non-severe ones.

An additional preliminary work was performed to better understand the reasons behind bow echo cases that are simulated well and those that are not and whether any systematic environmental differences were present between good and bad simulations. The Weather Research and Forecasting model (WRF) was used to simulate ten bow echo cases, and environmental parameters in a 200x200 km box just ahead of the system were averaged and compared with observations for both good and bad simulated bow echo cases. Preliminary results were inconclusive as no systematic differences between well simulated and poorly simulated bow echo environments were observed. Most parameters were highly case-dependent and unlikely to help forecasters anticipate errors in WRF-depicted systems for bow echo events.