Side-by-side tree and house damage in the May 2013 Moore, OK EF-5 tornado: Lessons for the Enhanced Fujita scale

There is now growing recognition that the tree-focused damage indicators of the EF scale are in need of improvement. For example, there are large differences in windfirmness of trees across species, sizes, soil conditions, and surroundings, yet the EF scale tree damage indicators (DIs) do not consider these factors. One promising approach to improving tree-focused DIs may be to examine tree damage in instances where tornadic winds can be inferred with confidence, such as very close to buildings with expert damage assessment and classification on the EF scale. Because structural wind damage is understood in much more detail, the structural damage and inferred winds can be taken as a benchmark against which tree damage is evaluated. Here we report initial findings from such an endeavor, based on single-family homes and nearby trees in residential neighborhoods of Moore, OK, after the 20 May 2013 EF-5 tornado.

Residential (FR12) EF-scale damage categories were obtained from the Norman, Oklahoma NWS Forecast Office storm surveys as reported in the online Damage Assessment Tool. Tree damage was estimated visually using several sources of aerial post-tornado imagery. For homes with a known EF-scale rating, damage characteristics of nearby trees (i.e., within 20 m) were recorded, subject to clarity and resolution constraints of the imagery. Tree damage was clearly less in the vicinity of EF-1 or EF-2 homes, compared to EF-4 or EF-5 homes. However, there appear to be no trends to support assignment of tree uprooting versus trunk breakage to different wind speed classes as is currently done in EF-scale damage assessments. While this method could resolve differences in tree diameter only to subjective small, medium, or large classes, smaller trees clearly experienced less debarking or removal of all major branches (stubbing) than large trees when near homes with a rating of EF-3 or greater. The current EF-scale tree DIs assign a degree of damage to debarked and stubbed trees that corresponds with inferred wind speeds at the EF-2 and EF-3 levels, but only a small percentage of the trees near EF-2 or EF-3 homes were stubbed. The percentage is much higher for trees near EF-4 and EF-5 homes. These results imply that stubbed and debarked trees are more indicative of EF-4 and EF-5 winds rather than EF-2 and EF-3 winds, supporting the conventional wisdom that this type of damage is indicative of extremely high wind speeds.