Read on Twitter
I've been taking a closer look at the very intense tornado that moved through Hampton County, SC this morning since it passed quite close to the Charleston WSR-88D (KCLX). Hopefully, this little thread can help frame the potential magnitude of this event. #scwx
It appears the peak 0.5° velocity scan was at 1022 UTC. Because the GR2Analyst dealiasing algorithm is unable to correctly handle the couplet, I've had to do some mental dealiasing (using SOLO3 isn't really practical over VPN).
By my best estimation, the peak outbound radial velocity (Vr) was 56.5 m/s (110 kt or 126 MPH) and the peak outbound Vr was 52.5 m/s (97 kt or 112 MPH). Thus, the rotational velocity (VROT) was 54.5 m/s (106 kt or 122 MPH). Both pixels have a mean elevation of 155 m ARL.
Taking a difference between of the distance of the couplet center points on the 0.5° scans before and after 1022 UTC yields an estimated translation speed of 21 m/s (41 kt or 47 MPH).
Thus, if you assume all the translation transfers to the circulation's wind field, and if the tornado were moving perfectly down-(up-)radial from KCLX, the maximum Vr it would likely measure would be 75.5 m/s (147 kt or 169 MPH).
Is this an accurate estimation of the peak intensity of the tornado? It's unlikely. Radars do not measure air velocity. They measure a reflectivity-weighted mean velocity. In general, the larger the particles the beam is being scattered off of, the greater the deviation...
...of the Vr from the actual wind speed (underestimation). In addition, even though the peak Vr pixels are only 15.37 km from KCLX and super-res oversampling allows us to retrieve data at an effective beam width of 0.5°, we're at best only coarsely resolving the broader...
...tornado cyclone, and certainly nowhere near sampling key details of the tornado or fine-scale structures within the tornado, such as multiple vortices, which could have a major impact on the peak wind intensity. Furthermore, we know from observations gathered on the Plains...
...and numerical simulations that the wind structure of a tornado can change dramatically very close to the surface, particularly under ~50-100 m ARL. This change is largely owing to surface friction, which can help induce significant radial (inflow) velocities in what's...
...referred to as the tornado boundary layer (TBL). Given that the peak velocities on this scan are at the same range from KCLX, the flow at this height is likely very close to cyclostrophic balance, which indicates limited radial flow and that we're very likely about the TBL...
...at this elevation. While some of the observations gathered on the Plains indicate that winds in the TBL can actually be substantially HIGHER than in the vortex above the TBL, it's unclear how that profile may look in areas with much larger surface roughness values, such as...
...the forested Southeast. And of course, while this estimate is in the EF4 range of wind speed estimates, only damage can be used to apply a rating to a tornado. There are many other more complicated considerations that further muddy the interpretation of radar data in the...
...estimation of tornado intensity. If you're interested in digging into the weeds, @SRHelicity has a paper that does a good job of detailing these considerations at length. https://journals.ametsoc.org/doi/full/10.1175/WAF-D-14-00026.1
This is my long-winded and self-isolated way of saying that this was an impressive storm, and potentially a very intense tornado. The location along the SC coast and the time of day make it especially impressive climatologically.
