Michael Hosek

Death of a supercell at sunset yesterday near Hydro, OK. It was amazing to be able to watch its updraft contract and change motion so dramatically (it had been moving south towards us before this). Plus some nice colors and mammatus peaking out by the end!

The timelapse I had always dreamed of capturing, from 4/24 near Matador, TX. It's incredible to see the full picture of a supercell and all of its dynamics, from the rotating updraft to the low-level mesocyclone and tornado dancing underneath. I could watch this for hours

...in a way that can help us better understand and forecast dryline-initiated severe storms across short-term, seasonal, and long-term timescales! (5/5)

We're currently working on using that ML model to extend our dryline climatology back to 1940 (81 years!) in order to investigate sources of variability in dryline frequency, location, and characteristics of the boundary itself. I hope to be able to continue this work and put it all together...(4/5)

We also found that the frequency of drylines appeared to be increasing over this 16-year period, but using a front-identifying Machine Learning (ML) model from Justin et al. 2025 (doi.org/10.1175/AIES...), we determined that this trend may not be as significant as it initially appeared (3/5)

...that were not. To no surprise, drylines are most common in spring, and the likelihood of a dryline being associated with severe storms increases later in the spring/early summer. The secondary bump in October, and the ramp-up in February/March are underexplored compared to April-June. (2/5)

A 16-Year Climatology of WPC-Analyzed Drylines and Their Association with Severe Convection Abstract A 16-yr (2007–22) climatology of drylines is presented. Constructed using NOAA Weather Prediction Center (WPC) surface analyses, this climatology addresses the limitations of season, time of ...

Happy to share that the first paper of my PhD -- a 16-year climatology of drylines -- is now published! (doi.org/10.1175/JAMC...) We compiled statistics for drylines across the entire CONUS over the entire year from the WPC, comparing drylines associated with severe storms vs those (1/5)