Search

Decades of research have investigated processes that contribute to downburst development, as well as identified precursor radar signatures that can accompany these events. These advancements have increased downburst predictability, but downbursts still pose a significant forecast challenge, especially in low-shear environments that produce short-lived single and multicell thunderstorms. Additional information provided by dual-polarization radar data may prove useful in anticipating downburst development. One such radar signature is the KDP core (where KDP is the specific differential phase), which can indicate processes such as melting and precipitation loading that increase negative buoyancy and can result in downburst development. Therefore, KDP cores associated with 81 different downbursts across 10 states are examined to explore if this signature could provide forecasters with a reliable and useable downburst precursor signature. The KDP core characteristics near the environmental melting layer, vertical gradients of KDP, and environmental conditions were quantified to identify any differences between downbursts of varying intensities. The analysis shows that 1) KDP cores near the environmental melting layer are a reliable signal that a downburst will develop; 2) while using KDP cores to anticipate an impending downburst's intensity is challenging, larger KDP near the melting layer and larger vertical gradients of KDP are more commonly associated with strong downbursts than weak ones; 3) downbursts occurring in environments with less favorable conditions for downbursts are associated with higher magnitude KDP cores, and 4) KDP cores evolve relatively slowly (typically longer than 15 min), which makes them easily observable with the 5-min volumetric updates currently provided by operational radars. [ABSTRACT FROM AUTHOR]

Research has shown that dual-polarization (dual-pol) data currently available to National Weather Service forecasters could provide important information about changes in a storm’s structure and intensity. Despite these new data being used gradually by forecasters more over time, they are still not used extensively to inform warning decisions because it is unclear how to apply dual-pol radar data to specific warning decisions. To address this knowledge gap, rapid-update (i.e., volumetric update time of 2.3 min or less) radar data of 45 storms in Oklahoma are used to examine one dual-pol signature, known as the differential reflectivity (ZDR) column, to relate this signature to warning decisions. Base data (i.e., ZDR, reflectivity, velocity) are used to relate ZDR columns to storm intensity, radar signatures such as upper-level reflectivity cores, and scientific conceptual models used by forecasters during the warning decision process. Analysis shows that 1) differences exist between the ZDR columns of severe and nonsevere storms, 2) ZDR columns develop and evolve prior to upper-level reflectivity cores, 3) rapid-update radar data provide a more complete picture of ZDR column evolution than traditional-update radar data (i.e., volumetric update time of about 5 min), and 4) ZDR columns provide a clearer and earlier indication of changes in updraft strength compared to reflectivity signatures. These findings suggest that ZDR columns can be used to inform warning decisions, increase warning confidence, and potentially increase warning lead time especially when they are integrated into existing conceptual models about a storm’s updraft and intensity. [ABSTRACT FROM AUTHOR]

In an 11-y period (2000-2010) 126 tornadoes affected central Oklahoma within a 111-km (60-nm) radius of the Twin Lakes, Oklahoma (KTLX) WSR-88D. The tornadoes resulted in 265 injuries and 3 deaths. This study used archived WSR-88D data to obtain information about storm characteristics such as mode, width, height, and measures of the mesocyclone, at the time of initial tornado formation. The radar data provided information about the supercell spectrum and highlighted differences between tornado-producing supercells and tornado-producing quasi-linear convective systems (QLCSs), especially with respect to midlevel rotational velocity. Warning lead-time information also was obtained and compared with the radar characteristics. No specific radar attribute was strongly correlated with lead time, likely due to the multitude of variables involved in the tornado warning process. A strong correlation did exist between lead time and storm mode. Applying these findings in an operational environment similar to that found in central Oklahoma may enhance tornado warning performance. [ABSTRACT FROM AUTHOR]