By:
Bob Henson
8:17 PM GMT on June 26, 2015
In most parts of the country, summer thunderstorms are most common in the afternoon or evening. Things are a bit different from the eastern Great Plains into the western Great Lakes. Across this core swath of the Midwest, lightning is most likely to zigzag across the summer sky during the wee hours of the morning. Along with disrupting countless nights of sleep, these middle-of-the-night thunderstorms are renowned for torrential rain, large hail, and destructive wind. One of several vivid examples from the past week was the supercell that moved across western South Dakota on Friday night, June 19,
killing livestock with softball-sized hail before morphing into an larger complex that sent
wind gusts of 60 to 100 mph across the southern half of the state well past midnight, finally reaching southwest Minnesota around 3:00 a.m. See radar loop
here (also embedded at the bottom of this post).
Figure 1. A powerful supercell, shown here near Belle Fourche, SD, raced across South Dakota on Friday night, June 19, while evolving into a mesoscale convective system (MCS). Image credit:
Cody Lere, courtesy of NWS Rapid City.
This year a major field study--primiarly supported by the National Science Foundation, with additional support from NASA, NOAA, and DOE--is zeroing in on large, long-lived storm clusters known as mesoscale convective systems (MCSs) as they prowl the central Plains from late evening deep into the night. The experiment, dubbed Plains Elevated Convection at Night (PECAN), started on June 1 and will continue through July 15 if the weather remains favorable. PECAN is a sprawling Kansas-based project, with an operations center at Fort Hays State University, three aircraft (a University of Wyoming King Air, NOAA P-3, and NASA DC-8), and a fleet of mobile teams, including nine portable Doppler radars and four vertical profiling units (PECAN Integrated Sounding Arrays, or PISAs). Six other PISAs are based at fixed sites. The project’s official domain extends from northern Oklahoma to southern Nebraska, although the P-3 flew into South Dakota to investigate the storm of June 19 and a successful mission took place in Iowa on Wednesday night, June 24.
PECAN’s overarching goal is to “advance the understanding and forecast skill of the processes that initiate and maintain nocturnal convection in the Great Plains.” It’s no mystery that nighttime MCSs are a huge influence on central U.S. climate. Roughly half of the summertime rainfall over the Great Plains is deposited by MCSs. Often, the systems develop night after night along a preferred west-to-east corridor whose location can shift north or south in a given year. Most nighttime MCSs over the plains are elevated, meaning they draw energy from just above the lowest few thousand feet of the atmosphere (the boundary layer). As the sun goes down, the boundary layer cools more quickly than the air above it, which can lead to an often-sharp inversion that keeps the two layers separated. Above the boundary layer, storm-fueling warmth and moisture from the Gulf of Mexico can sweep unimpeded into MCSs, borne on a southerly low-level jet stream that routinely ramps up around nightfall. Winds in this low-level jet stream often top 60 mph, which can keep an MCS going for hours on end. Some MCSs can persist for more than 12 hours and travel more than 1,000 miles.
Figure 2. Radar data from WSI, covering the June-to-August periods from 1996 to 2002, was used to produce this climatology that shows the time of day when precipitation (inferred from radar reflectivities of more than 15 dBz) were most likely. Much or most of this rain would be in the form of thunderstorms. Areas in black and dark blue/purple denote peak activity between around midnight and 6:00 a.m. local solar time. Image credit:
David Ahijevych, NCAR.
Some weather patterns are reliable MCS producers, making a given night’s forecast fairly straightforward. Other setups are more ambiguous, leading to the risk of unpleasant, dangerous surprises in the middle of the night. “In some cases it's clear that the MCS will move in a certain direction, but the forecasters may struggle with how long it will last,” said Mike Coniglio (NOAA National Severe Storms Laboratory), one of the PECAN co-investigators. “Other times the MCS motion can be quite a forecast challenge, since it depends on both the fine-scale details of the cold air produced by the cluster of thunderstorms as well as the details of the vertical wind profile surrounding the system, which itself often changes drastically in the few hours after sunset.” Still other times, an MCS will stay potent well through the night, producing severe weather and even tornadoes, despite background conditions that suggest a stable layer near the ground would inhibit the severe weather.
Figure 3. On June 4, NCAR’s S-Pol research radar monitored this rapidly developing thunderstorm in western Kansas, which merged with other cells as part of an MCS after sunset. S-Pol is stationed southwest of Hays for the PECAN project. Image credit: Carlye Calvin/UCAR.
To address the forecast question from several angles, PECAN is oriented around four distinct mission types. Each night of operations will include one or more sub-experiments organized around these four areas:
--The birth and early life of elevated storms: How do various types of disturbances kick off thunderstorms in the late evening, as the boundary layer is just beginning to detach from the atmosphere above it?
--The internal structure and microphysics of MCSs: How do the rain-cooled downdrafts from MCSs affect storm evolution? How well does the new dual-polarization software in NWS NEXRAD radars classify the size and types of raindrops, ice crystals, and cloud droplets found in MCSs?
--Bores and other wave-like features: MCSs can trigger atmospheric features that propagate well beyond the storm complex itself. One such feature that’s not well known to the public is the undular bore, which ripples outward against the prevailing wind much like a tidal bore does. These wavelike features can help nourish an MCS by forcing downstream air above the boundary layer to rise. What drives these features, and how can they be better folded into prediction efforts?
--Storm- and MCS-scale modeling: Data gathered from the field in PECAN will be teamed with computer models at various resolutions, to analyze what types of future observation platforms might be best suited for improving forecasts of MCS evolution. According to Coniglio, current storm-scale models tends to produce downdrafts that are too cold and spread out too extensively, which makes it harder to predict MCS behavior.
Figure 4. Undergraduates Leslie Cain (Fort Hays State University) and Shiou-Rong Chu (National Taiwan University) calibrated and launched radiosondes for PECAN from a fixed monitoring site near Brewster, Kansas. Image credit: Carlye Calvin/UCAR.
Life on the graveyard shift
This spring’s thunderstorm action shifted rather abruptly from torrential rain-producing storms in Oklahoma during May to MCSs swinging across the upper Midwest in June. Although Kansas hasn’t been the prime focal point of MCS development, enough storms have been moving through the northern part of the PECAN domain to keep the researchers busy, and the aircraft component extends the project range somewhat. PECAN is budgeted for up to 27 observing periods, and Thursday night was the 16th night of operations, with about three weeks left to go.
PECAN’s unusual hours are well suited for undergraduate and graduate students who know all about pulling all-nighters. Several dozen students have been launching radiosondes and deploying other equipment as part of PECAN’s mobile teams. Unlike chase-oriented projects such as VORTEX2, where mobile units move with storms throughout the afternoon and evening, PECAN’s mobile units generally stay put after their late-evening deployment. Once data gathering is done for the night, which can be as late as 3:00 or 4:00 am, it’s time to catch a few winks before gearing up in time for the next day’s 3:00 pm weather briefing. “Some of us have resorted to putting aluminum foil over the bedroom windows to block out the morning light and try to get on a roughly 3-am-to-11-am sleep schedule,” said Coniglio.
Figure 5. The PECAN PIE group, one of the teams of students and scientists documenting storms across the central Great Plains for weeks on end as part of PECAN. Left to right: Anthony Torres (University of Michigan), Erin Dougherty (2015 graduate, University of Virginia), and postdoctoral fellows David Bodine and Kristen Rasmussen (NCAR Advanced Study Program).
“I’ve just been pretending I'm in another time zone,” said Anthony Torres (University of Michigan), an undergraduate taking part in PECAN through UCAR’s SOARS program. “It's been a challenge remembering what day it is, what's considered breakfast, lunch, dinner, etc.” Torres and several other participants have dubbed themselves PECAN PIE, with the PIE standing for Precipitation Instrumentation Experiment. On Thursday night, they were in the midst of a multi-day deployment ranging across several states, using instruments called disdrometers that measure the sizes and fall speeds of individual raindrops. The PECAN PIE team has set up a Facebook page and their own WU blog, which they’ll soon be updating.
“We always see something interesting in the data,” says Tammy Weckwerth, one of six PECAN principal investigators. “It may be the structure and evolution of the MCSs, or the organization of the convection initiation events, or seemingly endless bores emanating out from storms, or interesting surface boundaries, or mid-level moisture tongues, or widespread waves propagating throughout the region. No matter when we collect data or what the conditions are, there is always something intriguing going on in the atmosphere.”
The PECAN Field Catalog, maintained by NCAR’s Earth Observing Laboratory, includes many links to outreach-oriented material. See also the NSF and NCAR news releases on PECAN and Jon Erdman’s comprehensive writeup on MCSs and PECAN at weather.com. Thanks to Jon for providing additional background for this blog post.
Bob Henson