Tuesday, 13 August 2002: 9:30 AM
The WDSS-II Supercell Identification and Assessment Algorithm
A Supercell Identification and Assessment Algorithm (SIAA) has been developed
to integrate the results of multiple signature detection algorithms, creating
a more holistic approach to storm type classification. By simultaneously
searching for the existence of several features known to be related to severe
potential in thunderstorms, and common to supercell thunderstorms, the amount
of cognitive processing required by the human is reduced, the interrogation
efficiency is increased, and warning decisions may be made faster, yet more
informed. SIAA effectively mimics the interrogation process a human
meteorologist would follow in determining storm type, and assessing the storm's
worthiness of immediate enhanced base data scrutiny by the warning
meteorologist. Because of its intuitive use of membership, fuzzy logic is
used as a proxy for the human process of classification in the determination of
the "supercell-ness" of a thunderstorm. Fuzzy membership values are computed
to characterize the storm's rotational strength, rotational longevity, bounded
weak echo region possession, and deviant motion. These values are combined in
a fuzzy classifier to calculate the storm's membership in the classes of
Supercell, Marginal Supercell, and Non-Supercell. Each storm with a three
dimensional circulation is inspected for a hook echo and inflow notch by
searching for points of high positive(hook) and negative(inflow notch)
curvature along the inner border, near the mesocyclone. The fuzzy membership
values pertaining to rotational strength, presence of hook echo, presence of
inflow notch, low-level rotational velocity strength, and presence of bounded
weak echo region are combined in a fuzzy classifier to calculate the storm's
membership in the classes Immediately Deserving, Marginally Deserving, and
Minimally Deserving of enhanced base data scrutiny. SIAA has shown impressive
skill in identifying supercell thunderstorms, and shows promise as an
interactive warning guidance tool, posting a POD of 0.82, an FAR of 0.14, a
CSI of 0.72, and a HSS of 0.66 in early testing (1,994 cases). SIAA reports
its results via a table and icon, keeping the user involved in the decision
Supplementary URL: