Rating Tornado Damage: an Exercise in Subjectivity

Tornado ratings from the 1950s into the early 1970s were based almost entirely upon remote, post-event review of media articles and photographs (Schaefer and Edwards 1998). Starting in the middle 1970s, local National Weather Service (NWS) offices have assigned F-scale ratings to most tornadoes based on a blend of f irst-hand surveys by their meteorologists and anecdotal evidence from news media, emergency management, storm observers, law enforcement personnel and the general public. Under this system variations in local rating practices from office to office become an intrinsic part of the modern Storm Data tornado database, which is used by the NWS and by most researchers studying aspects of tornado climatology. The dangers of such nonuniformity in rating procedures to scientific utility of tornado data were also well elucidated in the DB88 discussion, and remain such a concern 14 years later.