Hurricane Milton: RAP Generated Tornado Pathlength and Duration Predictions and Test Cases

 I collected data from Hurricane Milton in order to evaluate tornado pathlength and duration algorithms I'm currently testing for publication. Those pathlength/duration predictions generally increased in magnitude from morning into the afternoon as discrete pre-cyclone storms spread north across southern and eastern Florida, and as supercellular rainbands advanced inland across the state. Shear profiles were very favorable for rotating storms, and mid/upper-level flow was more typical of what would be observed across the Great Plains, Midwest, and Southeast during a standard tornadic ETC scenario. Another important aspect that likely enhanced tornado production was a dry midlevel airmass that Milton encountered as it advanced east toward Florida. The dry midlevel airmass allowed cloud breaks and resultant augmented surface heating to take place over southern/eastern Florida, which boosted temperatures into the 80s, led to steepening low-level lapse rates, and precluded substantial CIN and hurricane-related saturated low to midlevel thermodynamic profiles. In addition, dewpoints were well into the 70s as southeast surface winds caused rich moisture to spread inland across the state, which aided in the generation of a very deep buoyant environment.

The latest storm survey information from NWS Melbourne can be found here. The longest pathlength tornado surveyed so far was 33.9 km (21.2 miles), with a duration of 31 min, an EF rating of 3, and unfortunately a death toll of 6. That tornado touched down at around 459 PM EDT (21 UTC) and impacted the communities of Fort Pierce, Spanish Lakes, and Vero Beach. I captured a RAP sounding valid at 19 UTC, and located ~30 km west of the surveyed tornado pathlength (lat/long of 27.41, -80.66). That data from the RAP sounding was used as input for the tornado pathlength/duration algorithms I'm testing, and yielded a predicted pathlength of 39.2 km, and a predicted duration of 41.9 min, which of course is an error of ~6 km and ~10 min. Another 19 UTC RAP sounding 68 km west of Vero Beach yielded a closer pathlength prediction of 31.5 km and 37.9 min (error of ~2 km and 6 min, respectively)

Algorithm output and other relevant data related to the tornado event are posted below.

RAP generated hodograph yielding a predicted tornado pathlength of 39.2 km and tornado duration of 41. 9 min. The sounding point was around 30 km west of an observed tornado with a surveyed pathlength of 33.9 km and duration of 31 min.

RAP sounding taken in proximity to an EF3 tornado. Besides the very favorable wind profile for supercells, dry air above the surface, favorably steep lapse rates, a warm surface airmass, and very moist boundary layer aided in creating a very buoyant airmass with minimal CIN that supported robust thunderstorm development contributing to the tornado threat.

RAP hodograph and sounding shown above were picked from the red point appended to the map. Distance to the tornado is shown right of the red point.

RAP generated hodograph valid at 19UTC and ~60 km from an EF3 tornado. The predicted pathlength with this hodograph was 31.5 km and predicted duration was 37.9 min, which was in error by ~2 km and ~6 min from the observed pathlength and duration.

Distance from the RAP hodograph shown above to the observed tornado.

Base reflectivity valid during the time of the EF3 tornado impacting Florida coastal communities.

Storm-relative velocity from the lowest radar scan valid during the time of an EF3 tornado impacting coastal Florida communities.

12Z 250 mb upper-air analysis showing upper-level flow over Florida that perhaps was more typical of a continental ETC tornado threat.

Midday surface observations over Florida ahead of Hurricane Milton. Note the pocket of relatively rain-free air over southern/eastern Florida that was associated with cloud breaks and enhanced surface heating. Also not the very rich moisture spreading inland via east-southeast surface winds.

O’Bannon Woods State Park - Indiana

 I hiked ~6 miles of semi-rugged southern Indiana trail during mid September. O'Bannon Woods State Park sits on the Ohio River just a few miles downstream from Louisville. The terrain is surprisingly rocky, and there are numerous sink holes and caves, several entrances which I believe I spotted. The hike occasionally made me feel as if I were back in northern California, and I swear I spotted black bear poop on the trail...which is not completely out of the question in this part of the state. Definitely a region I will be visiting again, plenty of trails to hike, rocks to scramble on, history to soak in, and possibly caves to explore. Map of the region can be found here.

Random Pictures from Alaska and California

Mendenhall Glacier viewed from McGinnis Mountain trail. Juneau, Alaska.

Gastineau Channel sits below Mount Jumbo on Douglas Island. Juneau, Alaska.

Sheep Mountain helipad and weather equipment. Juneau, Alaska.

Prairie Creek Redwoods hiking trail...one of the best in California.

Prairie Creek Redwoods State Park.

Centerville Beach near Ferndale, California.

Redwood National Park.

Redwood National Park.

San Andreas Fault gouge rock. Lone Pine Canyon, California.

San Andreas Fault displaced rock and sediment from an historic earthquake. Trench located near Valyermo, California.

Another view of the San Andreas Fault strand near Valyermo, California.

Fault gouge rock, San Andreas Fault, Palmdale, California region.

View from Hudson Ranch Road as the San Andreas Fault transitions from mountainous terrain to the Carrizo Plain.

San Andreas Fault at the Carrizo Plain.

Northwest California Pacific Coast forest.

Northwest California Pacific Coast Forest.

Northwest California Pacific Coast forest.

Prairie Creek Redwoods State Park in Northwest California.

 

Turkey Run State Park - Indiana

 I hiked through Turkey Run State Park during mid August. The park is a bit of an anomaly compared to the relatively flat lands of central and northern Indiana. There is exposed rock, a bit of terrain to climb on, and flowing rushing water through fern filled canyons. I’ll take what I can get, if only the crowds filled with screaming children weren’t so overwhelming. Still, definitely a bucket list place to hit if you are in Indiana. 

Borah Peak // The Top of Idaho

 I hiked the standard route up Borah Peak during late July 2024. Smoke emanating from Butte County California filled the sky throughout my brief visit to Idaho, so views of the surrounding Lost River Range were not to be had. Otherwise, I arrived at the Borah trailhead an hour or two before sunrise, taking note of the 1983 earthquake scarp during my drive in the dark. Sadly my trekking poles were confiscated by TSA at the Indianapolis Airport, so it was a truly unrelenting uphill slog during the first half of the hike. A rounded ridge was thankfully encountered after a couple of hours, and that led to the infamous Chicken Out Ridge, which was fairly mild...I only encountered one brief sketchy move, with most of the sketch avoided via thoughtful route finding. I also bypassed the roped downclimb onto the melted snowbridge section, then made my way to the final uphill ascent that led to the summit. It took me about five hours to reach the top, and then three from the summit back to the trailhead (assisted by downhill jogging).

Google map location of Borah Peak is here.

First Flight // Sweet Aviation // Fort Wayne, Indiana

 I participated in a discovery flight with Sweet Aviation during late April. The flight went flawless thanks to my awesome instructor Austin. We circled around Fort Wayne. I made a few turns left and right, and practiced steering the plane on the ground. Great first flight and I’ll definitely be back for more. 

Storm Chase // 16 April 2024 // Southeast Iowa Tornadic Supercell

 I chased southern Iowa on April 16th and was rewarded with an intense EF2 producing tornadic supercell west of Burlington, IA. I departed Fort Wayne at around 1AM, took a brief nap near the Illinois border, and then reached my target of Albia, IA by late morning. Storms developed near Kansas City and advanced northeast across my region during the next few hours. Other convection was ongoing north and east as well, all due to the early arrival of an upper jet streak spreading across the area. I moved east with the storms coming up from the south, and they took on a QLCS structure. This didn't bode well for tornadoes, but there was a quasi-discrete cell along the southern end of the line, which I decided to target as it approached far southeast Iowa. This storm quickly matured into a supercell as it reached my location, with classic wall cloud structure, strongly rotating rain currents, and a well exposed updraft that showed clear signs of midlevel rotation. I moved north with the storm as the wall cloud lowered to the ground and displayed increasing rotation. And then just as it went tornadic, rain wrapped around the low-level meso and my view of the tornado was cut off. The road network was also becoming complex as I approached the MS river, so I let the rain-wrapped tornadic storm go as it drifted north, and then intercepted several cells to my south. They displayed some interesting structure, but also seemed to ingest cooler more stable air from the storm to the north, and thus failed to produce tornadoes. All in all a good chase day...images and weather data posted below, including a NAMNEST sounding/hodograph that yielded a predicted tornado pathlength of 78 km and tornado longevity of 61 minutes, which were slightly larger than the observed pathlength of 44 km and duration of 39 minutes associated with the EF2 tornado. Storm chase video that I captured can be found here.

Strong updraft rotation and swirling rain curtains adjacent to a rain free base viewed from highway 218 southwest of Burlington, IA.

Developing wall cloud and rotating rain curtains.

Low-level rotation becoming intense as the wall cloud lowers toward the ground.

Another view of a strongly rotating wall cloud.

Storm is going tornadic at this point with a rain filled RFD just about to wrap around the low-level mesocyclone.

Another cell coming up from the south. It possessed a large rain free base.

Strongly diffluent flow at 250 mb located within the left exit region of a 100+ kt jet streak arrived early across the region and led to widespread convection that may have precluded a more significant tornado outbreak.

A cold core 500 mb low centered over SD/NE/IA/MN border region focused low-topped tornadic storms near the bent-back occluded surface low.

700 mb objective analysis valid 00Z 17 April 2024.

850 mb objective analysis valid 00Z 17 April 2024.

Subjectively analyzed surface map valid 21Z 16 April 2024.

KDVN base reflectivity valid during the time of significant tornado activity in southeast Iowa.

KDVN 0.5 degree SRV valid during the time of sigtor activity in southeast Iowa. The velocity signature seems to be consistent with an EF2 strength tornado.

A NAMNEST sounding valid 21Z 16 April 2024 at Burlington, IA. The sounding may be located within the cool side of a subtle warm-frontal zone which might explain the slightly stable boundary layer profile. Otherwise, strong shear and sufficient CAPE were present for rotating storms and possible tornadoes.

A NAMNEST hodograph derived from the sounding above. Predicted tornado pathlength and longevity are displayed within the inset. Those values were slightly larger than what was observed with the EF2 tornado that occurred west of Burlington, IA.