NMDA sensitivity issues. No anticyclonic detection either.

Posted in HWT, NMDA on by Tiffany.

This will be a great case study to follow up on with the NMDA:

As we were watching the differences between the NMDA compared to the MDA and DMD on a QLCS/squall event over east-central Illinois/northwest Indiana, we noticed some differences in circulation detection. Particularly, it was noted that the detection sensitivity was poor when there was a strong anticyclonic circulation present aloft (KIND ~100nm southeast of storm; 0.5deg; 12kft AGL). Interestingly, all 3 of the algorithms picked up on the weaker cyclonic circulation instead; all with different sizes and intensities.

See below 1847Z: MDA upper left, DMD upper right, NMDA lower left.

Image below 1854Z: As the storm progressed and started having a cyclonic rotation associated with the area of interest right next to an anticylconic area of rotation, the MDA (upper left), DMD (upper right), picked up that particular circulation first. The NMDA (lower left) was still slow to grasp onto this very strong circulation aloft that could have a Severe lead time significance as the rear inflow and convergence strengthens.

Looking several radar scans forward now below (1900Z), the NMDA finally recognizes significance of this large feature (see below; lower left panel). However, it is still does not signify a strong circulation as denoted by its tiny circle. I’m curious about LSRs during this time.

Finally, 16 min later (1916Z see below; 11kft AGL), the NMDA finally shows a significant circulation center due to a strong descending rear inflow jet/notch. I would suggest that the NMDA does not give me a lot of lead time confidence in issuing a severe thunderstorm warning in this instance (probwind showing over 90% at this time). NMDA only seems to help give confidence once it detects the feature after it has already evolved. What’s also interesting to note from the below image is that the DMD and MDA picked their denotations on (only) the cyclonic circulation, whereas the NMDA only seems to centrally focus on a weird convergence region and not the cylconic side of this lowest scan velocity data (i.e. more localized for NMDA). None of the products seem to attract to the anticyclonic side, which actually strengthens with time due to the descending notch. -shearluck