Paul Markowski
· Distinguished Professor of MeteorologyVerifiedPennsylvania State University · Department of Meteorology and Atmospheric Science
Active 1998–2026
About
Paul Markowski is a Distinguished Professor of Meteorology and the Department Head at Penn State's Department of Meteorology and Atmospheric Science. His research group studies the dynamics and prediction of convective storms and their attendant hazards, with a particular focus on tornadoes. His work employs state-of-the-art observations, computer simulations, and theoretical approaches to better understand atmospheric convection, mesoscale meteorology, and severe weather phenomena. His research interests include atmospheric convection, mesoscale meteorology, and applying basic research findings to operational meteorology problems. He has contributed to the understanding of atmospheric dynamics, mesoscale meteorology, synoptic meteorology, and atmospheric convection, aiming to improve the prediction and understanding of severe weather events.
Research topics
- Geology
- Meteorology
- Physics
- Climatology
- Atmospheric sciences
- Mechanics
- Environmental science
- Classical mechanics
- Geodesy
- Geometry
Selected publications
Numerical Simulations of Supercell Storms Employing the Thin Boundary Layer Equations
Journal of the Atmospheric Sciences · 2026-01-12
articleSenior authorAbstract Three relatively high-resolution (75-m horizontal grid spacing), large-eddy simulation (LES) ensembles of tornadic supercell storms are used to investigate the effect of a more realistic treatment of the lower boundary condition on supercell storm simulations. These ensembles vary only in the parameterization of near-surface turbulence: one ensemble uses the semislip scheme, whereas the other two employ versions of the “thin boundary layer equations” (TBLE) approach. Each simulated storm is a long-lived supercell, with an intense mesocyclone and at least one tornado-like vortex (TLV). The primary differences observed across the ensembles are (i) the amplitude of near-surface turbulent eddies in the environmental boundary layer and (ii) the intensity of the TLVs, with both TBLE schemes resulting in stronger turbulent eddies in the environment and stronger TLVs within the simulated storms. Although substantial variations in precipitation distribution and cold pool strength are also found among the simulations, the differences in these larger-scale storm attributes from one ensemble to another are not statistically significant. Significance Statement The assumptions implicit in the semislip boundary condition, which is the lower boundary condition most commonly used in convective storm simulations that include surface drag, are problematic. An alternative approach to modeling the effects of the lower boundary, borrowed from the engineering large-eddy simulation community, is adopted for numerical simulations of supercell storms. The approach at least avoids some of the thorny assumptions inherent in the semislip condition. Though it is not yet possible to know which storm simulations are ultimately the “best,” an awareness of their sensitivities to the handling of the unresolved/underresolved flow in the near-surface layer is an important first step toward improving their realism.
Monthly Weather Review · 2026-02-25
articleAbstract This case study analyzes a nontornadic supercell observed on 9–10 June 2009 in southwest Kansas during the Verification of the Origins of Rotation in Tornadoes Experiment 2 (VORTEX2). Time-series multi-Doppler radar analyses and diabatic Lagrangian analysis retrievals document the kinematic and thermal–microphysical evolution of the storm’s strengthening low-level mesocyclone during the period 2342–2351 UTC, an apparent “tornadogenesis failure” event around 2351 UTC, and subsequent storm decay through 0024 UTC. An analyzed current of low- and midlevel streamwise vorticity enters the supercell updraft, appearing similar to the streamwise vorticity current (SVC) identified in previous supercell simulations. However, the present SVC primarily feeds into the midlevel updraft and mesocyclone, with relatively limited inflow to the low-level occlusion updraft and mesocyclone. Lagrangian vector vorticity dynamical calculations demonstrate that baroclinity, differential hydrometeor loading, and horizontal stretching all play significant roles in the generation and amplification of streamwise vorticity associated with this SVC. The origins of concentrated vertical vorticity in this mature low-level mesocyclone are consistent with off-trajectory separation of streamwise vorticity in downdraft previously identified in supercell simulations. However, the off-trajectory vorticity vector displacement in the present case is forced by differential hydrometeor loading instead of thermal gradients, since the cold pool is penetrating the low-level mesocyclone core. Another unique feature of this study is the detailed validation of surface radar-derived airflow and retrieved thermal fields with observations from a dense array of surface in situ measurement platforms in the storm. Significance Statement This study investigates the origins of strongly rotating low-altitude winds via airflow, temperature, and precipitation analyses of a nontornadic supercell (a long-lived thunderstorm with rotating updrafts). Although computer simulations provide highly detailed process understanding of complicated supercell evolutions, these simulated processes have been difficult to quantify in real supercells owing to a lack of required observations. We identify “currents” of horizontal vorticity—rotating wind in a vertical plane—that develop along the edges of the supercell’s rain-cooled outflow and precipitation core, similar to previous simulated rotation development processes. These vorticity currents are subsequently tipped in the observed storm’s updraft to impart its vertical rotation, although ingesting cold outflow likely prevents the observed rotating updraft from forming a tornado.
Will a Worm Turn? Observations of Coherent Surface-Layer Vertical Vorticity in Supercell Inflow
Bulletin of the American Meteorological Society · 2026-02-05
articleSenior authorAbstract A recent supercell modeling study showed that coherent turbulent structures associated with significant near-surface vertical vorticity in the ambient environment may serve as seeds for tornadoes. These structures, colloquially coined “vorticity worms” owing to their elongated and meandering appearance, develop within the surface layer in strong flow and are distinct from the larger-scale horizontal convective rolls often seen in the mixed layer. Elongated coherent turbulent structures are a common feature of the atmosphere and have been heavily studied by the boundary layer community, but before the recent simulation, they have not been previously implicated as facilitators of tornadogenesis. Observations of these structures around supercells have been limited, and historically, the observing platforms most often used in storm-focused data collection have not prioritized the precipitation-free inflow surface layer especially at the spatial and temporal resolution required to capture these structures and their interactions with storms. In recent field deployments, the National Severe Storms Laboratory used a truck-mounted pulsed Doppler lidar to perform near-ground scans to search for vorticity worms in real supercell inflow to compare with the simulation findings. Multiple deployments revealed vorticity worms and near-surface horizontal vorticity in supercell inflow that were consistent with the recently published supercell simulations that motivated this work. In two presented cases, vorticity worms were associated with the development of pretornadic, near-ground vortices, although neither circulation reached tornadic strength. These observations provide credibility to the recent simulations and suggest that continued observations of these coherent structures and research into their role in tornadogenesis are worth pursuing. Significance Statement New simulations suggest that a source of rotation for tornadoes may originate in the ambient environment in the form of streak-like coherent structures near the ground. We used a mobile Doppler lidar to document the existence of these streak-like structures around storms and present cases where these structures developed into regions of enhanced rotation when they moved near the storm’s updraft. These observations provide credibility to the recent simulations and suggest that more research and observations of these structures around storms are needed to confirm if these structures play a role in tornado formation.
2025-08-08
articleOpen accessSenior authorCorrespondingThe initiation of supercell thunderstorms presents a forecasting challenge due to the complexities of the interaction between growing cumulus clouds and wind shear. Previous research on convection initiation has largely focused on identifying the processes and environmental characteristics that control when and where long-lived thunderstorms will develop. However, once a persistent thunderstorm updraft does form in a sheared environment, several “milestones” of supercell evolution – including updraft splitting and cold pool formation – often must occur before the onset of severe weather hazards. Therefore, a better understanding of these developmental milestones may provide useful information for the forecasting of both convection initiation and supercell hazards.The goal of this research is to document early supercell evolution in a variety of environments. To address this goal, a matrix of CM1 simulations (dx = 100 m) with variable environmental thermodynamic and kinematic profiles was developed and analyzed. Preliminary results from these simulations indicate that there is large variability in early supercell evolution as a function of the environment. First, the timing of updraft splitting is sensitive to environmental stability. Updrafts split faster in environments with greater low-level stability, suggesting an important role of wake entrainment processes in eroding the center of the incipient updraft. Second, cold pool development is sensitive to both environmental stability and hodograph curvature. Most notably, in relatively stable environments characterized by strongly curved hodographs, the incipient updraft overruns the developing cold pool, resulting in a poorly-organized supercell by the end of the simulation. Finally, sensitivity tests with different microphysics schemes indicate that the simulated early supercell evolution is strongly dependent on the microphysics scheme used. Given that characteristics of early supercell evolution can vary markedly as a function of the storm environment and microphysics, additional field observations in this period of storm evolution may be particularly beneficial.
Quarterly Journal of the Royal Meteorological Society · 2025-04-04
articleOpen accessAbstract High‐resolution mesoscale model simulations of two archetypal quasi‐linear convective systems producing outbreaks of three or more tornadoes in the United Kingdom are performed to determine vortexgenesis mechanisms. Type 1 events are associated with north–south‐oriented cold fronts with regularly spaced misocyclones along them. In one type 1 event, a near‐surface vortex sheet broke down into near‐equally spaced misovortices having a wavelength of about 7.5 times the width of the shear zone. Rayleigh's and Fjørtoft's instability criteria were met preceding the development of the vortices, suggesting the presence of horizontal shearing instability (HSI). Lagrangian calculations of vorticity tendency showed that parcels entered the misovortices at lower heights, acquired their vorticity via tilting, before being further enhanced by stretching as the parcel ascended. These results implied HSI was the initial mechanism for the amplification of perturbations along the vortex sheet in the type 1 event. In contrast, type 2 events are associated with west–east‐oriented cold fronts with disorganized, elongated cyclonic–anticyclonic vorticity couplets evolving into a small number of cyclonic and anticyclonic misovortices with irregular misovortices. In one type 2 event, Fjørtoft's instability criterion was not met. Lagrangian vorticity‐tendency calculations showed that parcels acquired vorticity similar to type 1 events, where parcels entered the misovortices at lower heights, acquired their vorticity via tilting, before being further enhanced by stretching as the parcel ascended. However, the magnitude of tilting was typically larger in the type 2 event. Comparing these two events showed two possible mechanisms for misovortexgenesis in UK tornado outbreaks: misovortices in type 1 events form and grow via HSI along the front, whereas misovortices in type 2 events are not due to HSI.
2025-08-08
preprintOpen access1st authorCorrespondingThis presentation will summarize and discuss the key findings from a recent numerical modeling study (Markowski 2024, Journal of the Atmospheric Sciences), which examined tornadogenesis in a supercell simulation incorporating a turbulent environmental boundary layer. The simulation configuration represents a departure from the set-up used in almost all prior simulations, in that past simulations almost always have had a laminar environmental boundary layer. Substantial near-surface vertical vorticity (ζ > 0.03 s⁻¹ at z = 7.5 m) is present in the form of elongated streaks oriented with the southerly ground-relative winds. These ζ streaks coincide with undulations in predominantly horizontal, westward-directed environmental vortex lines, shaped by vertical motions linked to coherent turbulent structures—features long recognized in the boundary layer and turbulence literature. The ζ streaks act as preferred sites for tornadogenesis, and may even facilitate it, as environmental ζ can be quickly intensified by the strong convergence beneath supercell updrafts. Interestingly, the simulation lacks evidence of the traditional "baroclinic mechanism" of tornadogenesis, despite the supercell's structure and evolution closely resembling that seen in cases in which the baroclinic mechanism is operating. I will try to make sense of what all of these findings might mean. The presentation also will highlight unexpected differences in cold pool behavior between storms initialized with turbulent versus laminar boundary layers.
Axisymmetric Analysis of Tornado-Like Vortices in Simulated Supercells
Journal of the Atmospheric Sciences · 2025-10-08
articleSenior authorAbstract Present-day simulations of supercell thunderstorms have high resolution and incorporate the physical processes known to be conducive to simulating tornado-like vortices (TLVs). In such supercell simulations, TLVs are identified by the strength, duration, and location of the vertical vorticity in the simulated supercell. To bring the analysis of these supercell-produced TLVs a step closer to observations and theory, the TLV in an advanced supercell simulation is identified as the tornado center, and the Cartesian model velocities are transformed to cylindrical coordinates and azimuthally averaged. The azimuthally averaged TLV exhibits many of the observed and theoretically modeled features of tornadoes, including an end-wall vortex with strong maximum vertical and azimuthal velocities ( w max and υ max ) close to the ground with transition to the weaker core velocities ( w c and υ c ) aloft through vortex breakdown. A theory for the rotating-flow boundary layer with radial and azimuthal inflow velocities modeled on the axisymmetric analysis is shown to produce good qualitative agreement with the analyzed axisymmetric TLVs; however, the theoretical w max and υ max are too large as the theory does not account for vortex breakdown. Estimation of the corner flow swirl ratio suggests a limit of υ max / υ c slightly greater than unity; since υ c is a feature of the mesocyclone, there is a limit on the extent to which the amplified velocities of the end-wall velocities can be realized. To augment the diversity of cases, the present analysis is applied to a simplified set of supercell simulations; the present theory explains several features of the axisymmetric vortices.
Monthly Weather Review · 2024-03-11 · 1 citations
articleSenior authorAbstract Previous work found that cold pools in ordinary convection are more sensitive to the microphysics scheme when the lifting condensation level (LCL) is higher owing to a greater evaporation potential, which magnifies microphysical uncertainties. In the current study, we explore whether the same reasoning can be applied to supercellular cold pools. To do this, four perturbed-microphysics ensembles are run, with each using an environment with a different LCL. Similar to ordinary convection, the sensitivity of supercellular cold pools to the microphysics increases with higher LCLs, though the physical reasoning for this increase in sensitivity differs from a previous study. Using buoyancy budgets along parcel trajectories that terminate in the cold pool, we find that negative buoyancy generated by microphysical cooling is partially countered by a decrease in environmental potential temperatures as the parcel descends. This partial erosion of negative buoyancy as parcels descend is most pronounced in the low-LCL storms, which have steeper vertical profiles of environmental potential temperature in the lower atmosphere. When this erosion is accounted for, the strength of the strongest cold pools in the low-LCL ensemble is reduced, resulting in a narrower distribution of cold pool strengths. This narrower distribution is indicative of reduced sensitivity to the microphysics. These results suggest that supercell behavior and supercell hazards (e.g., tornadoes) may be more predictable in low-LCL environments. Significance Statement Thunderstorms typically produce “pools” of cold air beneath them owing in part to the evaporation of rain and melting of ice produced by the storm. Past work has found that in computer simulations of thunderstorms, the cold pools that form beneath thunderstorms are sensitive to how rain and ice are modeled in the simulation. In this study, we show that in the strongest thunderstorms that are capable of producing tornadoes, this sensitivity is reduced when the humidity in the lowest few kilometers above the surface is increased. Exploring why the sensitivity is reduced when the humidity increases provides a deeper understanding of the relationship between humidity and cold pool strength, which is important for severe storm forecasting.
‘Twisters’ movie: Two tornado scientists take us inside the real world of storm chasing
2024-07-11
articleSenior authorOn the vortex dynamical contribution to the emission of infrasound from tornadoes
The Journal of the Acoustical Society of America · 2024-08-01
articleOpen accessSenior authorPrevious field experiments support the claim that a tornado can radiate discernible infrasound between 0.5 and 10 Hz. The physical mechanisms of tornado sound generation are still not fully understood, although several potential mechanisms have been proposed. In this paper, the theory of vortex sound is applied to the sound radiation from two numerical tornado simulations based on large eddy simulations. There are two different vortex-related mechanisms in distinct frequency regimes. It is found that rotation of a non-axisymmetric vorticity field produces low-frequency infrasound less than 2.0 Hz. High-frequency tornado infrasound can be attributed to more complex vortex dynamics such as vortex merging.
Recent grants
NSF · $222k · 2004–2008
NSF · $1.2M · 2018–2023
NSF · $1.1M · 2022–2026
NSF · $748k · 2007–2014
Frequent coauthors
- 52 shared
Yvette Richardson
Pennsylvania State University
- 23 shared
Erik N. Rasmussen
NOAA National Severe Storms Laboratory
- 21 shared
Joshua Wurman
University of Illinois Urbana-Champaign
- 16 shared
David C. Dowell
NOAA Earth System Research Laboratory
- 14 shared
Jerry M. Straka
- 10 shared
James Marquis
Pacific Northwest National Laboratory
- 8 shared
Paul Robinson
University of Illinois Urbana-Champaign
- 8 shared
Karen Kosiba
University of Alabama in Huntsville
- Resume-aware match score
- Save to shortlist
- AI-drafted outreach
See your match with Paul Markowski
PhdFit ranks faculty by your research interests, methods, and publications — grounded in their actual work, not templates.
- Free to start
- No credit card
- 30-second signup