Effects of Terrain and Land Cover on Radar-Observed 3-Dimensional Tornado Wind Fields

The overall impact that terrain and surface roughness have on tornado behavior is generally poorly understood. Although some advances have been made using both observations (e.g. Houser et al. 2020; Lyza et al. 2018, 2020) and numerical simulations (e.g. Lewellen 2012, 2014; Satrio et al. 2020), there are still many unanswered questions. These studies, and others, suggest that terrain particularly appears to have an effect on tornado characteristics such as intensity and perhaps path direction. However, the results between studies, and even between different cases within the same study, conclude inconsistent outcomes on the tornado as it encounters changes in terrain or land cover. For example, in some circumstances, tornadoes are shown to weaken moving uphill while in others, they strengthen. Observationally-based studies have only used single-Doppler data to this point, and model-based studies drastically simplify tornado vortex structure as well as terrain characteristics. Furthermore, studies investigating the effects of surface roughness on tornado characteristics are especially limited and have produced inconclusive results.

The goal of this work is to continue investigating the role terrain and land cover play on observed tornadoes by examining characteristics of three-dimensional tornado wind fields obtained from dual-Doppler radar analyses. Data from pairs of both mobile and fixed site radars (e.g. RaXPol, DOW’s, the PAR testbed, and KDDC) are used in combination with data from Digital Elevation Models and the National Land Cover Dataset to quantify statistical relationships between tornado intensity (as indicated by vertical vorticity, circulation, and vertical velocities) and i) terrain elevation and/or ii) land cover type (e.g. forest, grassland, water, shrubs, etc.), which is assigned a quantitative surface roughness length according to the EPA’s AERSURFACEs user guide. Dual-Doppler Analyses from 6 tornadoes having varying intensity (EF0-EF4) are acquired and merged, comprising a dataset of nearly 100 observations. Tornado wind field parameters (i.e., w, ζ, Γ, δ) are divided into a variety of groups associated with either terrain or land cover characteristics (e.g. w associated with a change in elevation > 0 vs < 0; δ associated with surface roughness values in the upper quartile vs. surface roughness values in the lower quartile; etc.) and bootstrapped to determine whether the wind field characteristics are different for the varying land cover or terrain conditions. Results of individual tornadoes follow suit with past studies suggesting that outcomes can vary on a case-by-case basis; but there is evidence that some broader generalizations can be made when the datasets are merged together, particularly with surface roughness.