WALT Leaf Model Shows Consistent Severe Wind Signal Across Southeast Oklahoma 

Forecast prepared by Climate Scientist Wayne Adams

Field testing conducted across portions of Southeast Oklahoma has produced a remarkably consistent damaging-wind signal in the experimental WALT Leaf Model. Leaves were analyzed while still attached to trees, reducing the possibility that handling or removal could influence visible characteristics used by the model. Testing included Spiro, Panama, Shady Point, Poteau, Howe, Antlers, and two locations in Heavener.

The strongest Kink Index reading was recorded at Panama with 72, followed closely by Spiro, Shady Point and Howe at 71, Antlers 61,. Poteau registered a Kink Index of 59. The two Heavener samples produced readings of 62 and 41, indicating considerably greater variation within Heavener than was observed among the other locations.

The WALT tornado component produced a 50% model-indicated tornado potential at Spiro, Panama, Shady Point, Poteau and Howe. The two Heavener locations produced lower model readings of 39% and 35% and Antlers 40%. These percentages are outputs generated by the experimental WALT model and should not be interpreted as official probabilities that a tornado will strike these communities.

Perhaps the most notable result was the damaging-wind component. Every location tested produced the same WALT estimated wind range of 60 to 75 mph. The consistency of that signal across geographically separated locations suggests that damaging wind may be the dominant feature being detected by the model. Additional testing and future storm verification will be necessary to determine whether this pattern has predictive value.

One storm mode that will be examined when these results are compared with future severe-weather events is a Quasi-Linear Convective System, or QLCS. QLCS events consist of organized lines of thunderstorms capable of producing widespread damaging straight-line winds. Embedded circulations can also develop within these lines and occasionally produce brief, fast-moving tornadoes. A widespread wind signal accompanied by a less uniform tornado signal makes QLCS-type events an important category for future WALT verification.

The geographic pattern is also noteworthy. Spiro, Panama, Shady Point and Howe clustered extremely closely, with Kink Index values of 71–72, while Poteau was moderately lower at 59. Antlers ' Kink Index of 61, Heavener showed the greatest variation. Across all seven observations, the average Kink Index was approximately 67, while the median was 71. Additional observations from other portions of Le Flore County would help determine whether this represents a broader countywide pattern or localized differences between trees and sampling sites.

These results will eventually be combined with conventional climate information, including the influence of El Niño, to develop separate severe-weather outlooks for Southeast Oklahoma. The conventional climatological assessment and WALT results will remain separate so that the experimental model can be independently evaluated against future observed tornado, hail and damaging-wind events.

The WALT Leaf Model remains an experimental severe-weather research model. Its tornado percentages, Kink Index values and wind estimates are not National Weather Service forecasts and should not replace official watches or warnings. The value of the current field survey will ultimately depend upon prospective verification—preserving today's readings and later comparing them with what actually occurs.