Ensemble sensitivity analysis (ESA) is a computationally inexpensive technique to diagnose how the initial atmospheric state affects a later forecast metric. Recent study has found success applying this method to convective phenomena from the mesoalpha scale to the storm scale. Some of these works have analyzed short-term, ensemble predictions of thunderstorms and their near-storm environments within the Warn-on-Forecast System (WoFS). However, several challenges remain largely unexplored that may complicate ESA usage within the system, including a multiphysics configuration and relatively small ensemble size. As such, this study applies ESA to forecasts of individual mesocyclones to 1) better understand its utility within the WoFS framework and 2) uncover near-storm controls on predicted mesocyclone strength. Storm-relative sensitivities to many diagnostic variables are calculated at short (30-90 min) lead times for an updraft helicity response function. Their spatial and statistical distributions are analyzed as they relate to near-storm dynamics and member-to-member planetary boundary layer (PBL) scheme stratification. ESA is found to provide useful insight into sources of short-term uncertainty of predicted mesocyclone strength. These sources also differed notably case to case, ranging from environmental differences to storm interactions, and were often induced by PBL multiphysics. Further, ESA illuminates how WoFS models storm-induced kinematic feedbacks, finding patterns similar to previous supercell dynamics research. These results indicate that ESA is an effective metric to identify sources of uncertainty in future WoFS case studies. Additionally, the utility of ESA-based forecast improvement techniques like ensemble subsetting should be explored within WoFS. SIGNIFICANCE STATEMENT: The Warn-on-Forecast System is an ensemble forecasting system designed to provide probabilistic guidance for individual thunderstorms during the time between National Weather Service watches and warnings. To help better understand how this system models storms and their interactions with surrounding conditions, a statistical technique called ensemble sensitivity analysis is employed and its usefulness within the system is assessed. Four cases studies are presented in this work to study its utility and the sensitivities of individual rotating thunderstorms within different background conditions. This research strives to develop a foundation upon which to justify the use of ensemble sensitivity within the Warn-on-Forecast System for potential forecast improvement applications that may improve warnings for severe storms.

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 lowand 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 down-draft 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 lowlevel 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.

Targeted Observation by Radars and Uncrewed Aircraft Aystems (UAS) of Supercells (TORUS) aimed to improve the conceptual model of supercell thunderstorms through advancing the understanding of the role of storm-generated airmass boundaries and coherent structures in the development of near-surface rotation. Research questions guiding the field phase of TORUS focused on left-flank vertical vorticity sheets, streamwise vorticity currents, left-flank convergence boundaries, and rear-flank internal surges. Research questions also aimed to address the relationship between inflow modification and supercell characteristics. Across three field seasons (2019, 2022, and 2023), data on 46 supercell thunderstorms were collected through coordinated deployments of radars, lidars, mobile mesonets, UAS, manned aircraft, radiosondes, and swarmsondes. More than 200 scientists and engineers (many of whom were students) participated in the TORUS field deployments. The scientific motivation for TORUS, experiment design, and examples of data/ analysis are presented in this article. SIGNIFICANCE STATEMENT: Targeted Observation by Radars and Uncrewed Aircraft Systems (UAS) of Supercells (TORUS) was a collaborative research project funded by the National Science Foundation and the National Oceanic and Atmospheric Administration to advance understanding of supercells. TORUS involved more than 200 scientists and engineers (many of whom were students) who led data collection on 46 supercell thunderstorms across three field seasons. This effort constituted the most deployments of UAS within supercells and, on 17 May 2019, likely yielded the longest continuous airborne multi-Doppler radar sampling of a Great Plains supercell ever conducted.

On 15 June 2019, a dissipating left-moving supercell was sampled during the first year of the Targeted Observation by Radars and Uncrewed Aerial Systems (UASs) of Supercells field campaign near Vega, Texas. A unique data-set was captured, providing a rare opportunity to explore physical processes related to storm demise. The storm weakened rapidly as the main convective tower detached from its cloud base, resembling the "downscale transition" discussed in prior studies. Ground-based mobile mesonet and mobile radar observations revealed an intensifying cold pool surge moments prior to complete dissipation. Mobile radiosonde observations revealed an increase in low-level stability and low-level storm-relative helicity prior to local sunset. A series of numerical simulations were performed using the Bryan Cloud Model to understand supercell evolution by controlling changes to the background state using base-state substitution. Results showed that changes to low-level stability had a greater influence on storm evolution than changes to the wind profile. Trajectory analyses reveal that updraft parcel origin heights ascended only when changes in low-level stability were introduced, with elevated parcels becoming drier and less buoyant, weakening the main updraft through the effects of entrainment. Thus, it is hypothesized that dissipation resulted from reduced updraft buoyancy and dry air entrainment from elevated parcels through increased low-level stability. It is also hypothesized that the main convective tower separated from the parent cloud base due to drier updraft parcels reaching the main updraft from higher altitudes. SIGNIFICANCE STATEMENT: An instance of storm dissipation was analyzed near Vega, Texas, using ground-based instrumentation and mobile radar observations from the Targeted Observation by Radars and Uncrewed Aerial Systems of Supercells field campaign nearing the time of local sunset. Observations showed that the storm's cold air intensified and moved into the storm's inflow region moments prior to the evaporation of the main updraft and its cloud base. Numerical simulations were performed to replicate the storm's behavior and showed that the cooler, stable air in the storm's environment directly contributed to the storm's dissipation rather than the evolving environmental wind profile. During the dissipation, drier parcels from above the surface likely reduced the vertical motion within the updraft and caused the main cloud tower to evaporate before its cloud base.

Past studies have shown baroclinic zones between warm ambient inflow and cooled downdraft air to be key in the generation of horizontal vorticity and in potential tornadogenesis if tilted and stretched in the vertical by updrafts. While previous work has focused on Great Plains supercells, a much narrower body of work has considered similar vorticity sources in high-shear, low-CAPE supercells and quasi-linear convective systems (QLCSs). In this study, 166 Texas Tech field campaign, including 76 mesovortex (23 tornadic and 53 nontornadic) and 35 high-shear, low-CAPE supercell intercepts, are statistically and spatiotemporally analyzed. Larger virtual potential temperature gradients were observed near rotating segments of the QLCS. No statistically significant differences in cold pool characteristics are evident globally between nontornadic and tornadic mesovortices, though significant differences do exist when considering specific periods of the mesovortex life cycle and mesovortex-relative position. Three fine-scale array (;1-km spacing) mesovortex intercepts during the nontornadic, time-of-tornado, and posttornadic phases support the statistical analysis, possibly implicating baroclinic vorticity generation as a viable source of vorticity to achieve tornadogenesis within QLCS mesovortices. Further, this study found supercells in high-shear, low-CAPE regimes uniformly have weaker cold pools than supercells in larger CAPE environments. After QLCS-supercell mergers, the resultant QLCS cold pool tended to weaken immediately downstream of the merger but with no disruption in the tornado potential of the QLCS segment.

Quasi-linear convective systems (QLCSs) are responsible for approximately a quarter of all tornado events in the United States, but no field campaigns have focused specifically on collecting data to understand QLCS tornadogenesis. The Propagation, Evolution, and Rotation in Linear Storms (PERiLS) project was the first observational study of tornadoes associated with QLCSs ever undertaken. Participants were drawn from more than 10 universities, laboratories, and institutes, with over 100 students participating in field activities. The PERiLS field phases spanned 2 years, late winters and early springs of 2022 and 2023, to increase the probability of intercepting significant tornadic QLCS events in a range of large-scale and local environments. The field phases of PERiLS collected data in nine tornadic and nontornadic QLCSs with unprecedented detail and diversity of measurements. The design and execution of the PERiLS field phase and preliminary data and ongoing analyses are shown.

On 28 May 2019, a tornadic supercell, observed as part of Targeted Observation by UAS and Radars of Supercells (TORUS) produced an EF-2 tornado near Tipton, Kansas. The supercell was observed to interact with multiple preexisting airmass boundaries. These boundaries and attendant air masses were examined using unoccupied aircraft system (UAS), mobile mesonets, radiosondes, and dual-Doppler analyses derived from TORUS mobile radars. The cool-side air mass of one of these boundaries was found to have higher equivalent potential temperature and backed winds relative to the warm-side air mass; features associated with mesoscale air masses with high theta-e (MAHTEs). It is hypothesized that these characteristics may have facilitated tornadogenesis. The two additional boundaries were produced by a nearby supercell and appeared to weaken the tornadic supercell. This work represents the first time that UAS have been used to examine the impact of preexisting airmass boundaries on a supercell, and it provides insights into the influence environmental heterogeneities can have on the evolution of a supercell.

Over the last decade, supercell simulations and observations with ever-increasing resolution have provided new insights into the vortex-scale processes of tornado formation. This article incorporates these and other recent findings into the existing three-step model by adding an additional fourth stage. The goal is to provide an updated and clear picture of the physical processes occurring during tornadogenesis. Specifically, we emphasize the importance of the low-level wind shear and mesocyclone for tornado potential, the organization and interaction of relatively small-scale pretornadic vertical vorticity maxima, and the transition to a tornado-characteristic flow. Based on these insights, guiding research questions are formulated for the decade ahead. SIGNIFICANCE STATEMENT: This article provides a nontechnical overview of how tornadoes form. Sequentially, the most important processes include the initial creation of rotating updrafts, the development of disorganized patches of rotation at the surface, the organization of these patches into a more defined, symmetric vortex, and the final transition into a fully developed tornado in which air turns abruptly upward very near the surface. Based on this proposed conceptual model, guiding research questions are formulated for the decade ahead.

This study aims to objectively identify storm-scale characteristics associated with tornado-like vortex (TLV) formation in an ensemble of high-resolution supercell simulations. An ensemble of 51 supercells is created using Cloud Model version 1 (CM1). The first member is initialized using a base state populated by the Rapid Update Cycle (RUC) proximity sounding near El Reno, Oklahoma, on 24 May 2011. The other 50 ensemble members are created by randomly perturbing the base state after a supercell has formed. There is considerable spread between ensemble members, with some supercells producing strong, long-lived TLVs, while others do not produce a TLV at all. The ensemble is ana-lyzed using the ensemble sensitivity analysis (ESA) technique, uncovering storm-scale characteristics that are dynamically relevant to TLV formation. In the rear flank, divergence at the surface southeast of the TLV helps converge and contract existing vertical vorticity, but there is no meaningful sensitivity to rear -flank outflow temperature. In the forward flank, warm temperatures within the cold pool are important to TLV production and magnitude. The longitudinal positioning of strong streamwise vorticity is also a clear indicator of TLV formation and strength, especially within 5 min of when the TLV is measured.

The issues of Monthly Weather Review would be mostly empty if it were not for the large number of high-quality submissions each year. To retain our position as one of the leading meteorological journals in the world, we need to attract and retain these valuable contributions. However, of the approximately 400 annual submissions, only 56.7% were eventually published in 2021. The percentage of papers being rejected has slowly increased from 33.7% in 2007 (Schultz 2010a) to 37.6% in 2021, with withdrawals and transfers making up the remaining few percent. Although some of those rejections are because the papers are off topic forMonthly Weather Review, most rejections are because the science does not meet our minimum standards or may need to be explained better. Defending the rejection of papers, the great fluid-mechanics scientist Batchelor (1981, p. 16), the founder and chief editor of the Journal of Fluid Mechanics, once wrote,

Many numerical studies have focused on the importance of baroclinically generated vorticity at the edge of cold pools in supercellular tornadogenesis, and observational work has consistently found that strongly tornadic supercells have less dense, more buoyant cold pools than weakly or nontornadic supercells. However, there is a lack of observational studies that consider potential relationships between cold pool characteristics (e.g., density) and tornado production within linear systems, such as mesoscale convective system (MCS) or quasi-linear convective system (QLCS) events. This study presents two tornadic QLCS events that were observed during the Verification of the Origins of Rotation in Tornadoes Experiment-Southeast (VORTEX-SE) field project in 2016 and 2017. Supercell and hybrid modes were also observed and compared to the observations from the linear systems. No obvious differences in the thermodynamic deficits of the tornadic and nontornadic samples were found, likely due to the weakness of the produced tornadoes (≤EF1) and the small tornadic sample size (five cold pools). Comparison across storm mode did find some differences, with QLCS cold pools producing larger virtual potential temperature and psuedoequivalent potential temperature deficits than those observed in supercells. More importantly, our findings suggest that, in a QLCS, the magnitude of density gradients along the leading edge of the cold pool may be related to tornadogenesis by virtue of the implied baroclinic vorticity generation.

Ensemble forecasts are generated with and without the assimilation of near-surface observations from a portable, mesoscale network of StickNet platforms during the Verification and Origins of Rotation in Tornadoes EXperiment – Southeast (VORTEX-SE). Four VORTEX-SE intensive observing periods are selected to evaluate the impact of StickNet observations on forecasts and predictability of deep convection within the southeast United States. StickNet observations are assimilated with an experimental version of the HighResolution RapidRefresh Ensemble (HRRRE) in one experiment, and withheld in a control forecast experiment. Overall, StickNet observations are found to effectively reduce mesoscale analysis and forecast errors of temperature and dewpoint. Differences in ensemble analyses between the two parallel experiments are maximized near the StickNet array and then either propagate away with the mean low-level flow through the forecast period or remain quasi-stationary, reducing local analysis biases. Forecast errors of temperature and dewpoint exhibit periods of improvement and degradation relative to the control forecast, and error increases are largely driven on the storm scale. Convection predictability, measured through subjective evaluation and objective verification of forecast updraft helicity, is driven more by when forecasts are initialized (i.e., more data assimilation cycles with conventional observations) rather than the inclusion of StickNet observations in data assimilation. It is hypothesized that the full impact of assimilating these data is not realized in part due to poor sampling of forecast sensitive regions by the StickNet platforms, as identified through ensemble sensitivity analysis.

The forward-flank convergence boundary (FFCB) in supercells has been well documented in many observational and modeling studies. It is theorized that the FFCB is a focal point fore baroclinic generation of vorticity. This vorticity is generally horizontal and streamwise in nature, which can then be tilted and converted to mid-level (3-6 km AGL) vertical vorticity. Previous modeling studies of supercells often show horizontal streamwise vorticity present behind the FFCB, with higher resolution simulations resolving larger magnitudes of horizontal vorticity. Recently, studies have shown a particularly strong realization of this vorticity called the streamwise vorticity current (SVC). In this study, a tornadic supercell is simulated with the Bryan Cloud Model at 125-m horizontal grid spacing, and a coherent SVC is shown to be present. Simulated range-height indicator (RHI) data show the strongest horizontal vorticity is located on the periphery of a steady-state Kelvin-Helmholtz billow in the FFCB head. Additionally, similar structure is found in two separate observed cases with the Texas Tech University Ka-band (TTUKa) mobile radar RHIs. Analyzing vorticity budgets for parcels in the vicinity of the FFCB head in the simulation, stretching of vorticity is the primary contributor to the strong streamwise vorticity, while baroclinic generation of vorticity plays a smaller role.

Denotes content that is immediately available upon publication as open access.© 2020 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Ensemble sensitivity analysis (ESA) is applied to select types of observations, in various locations and in advance of forecast convection, to systematically evaluate the effectiveness of ESA-based observation targeting for 10 convection forecasts. To facilitate the analysis, observing system simulation experiments and perfect models are utilized to generate synthetic targeted observations of temperature and pressure for future assimilation with an ensemble prediction system. Various observation assimilation experiments are carried out to assess the impacts of nonlinearity, covariance localization, and numerical noise on ESA-based observation-impact predictions. It is discovered that localization applied during data assimilation restricts targeted-observation increments onto the forecast responses of composite reflectivity and 3-hourly accumulated precipitation, making impact predictions poor. In addition, numerical noise introduced by nonlinear perturbation evolution tends to reduce the correlations between observed and predicted impacts; small, random-perturbation experiments often yielded similar impacts on forecasts as targeted observations. Nonlinearity also manifests in the observation impacts when comparing targeted observations with nontargeted, randomly chosen observations; random observations have seemingly the same impact on forecasts as targeted observations. The results, under idealized conditions and simplified ensemble configurations, demonstrate that ESA-based targeting for nonlinear convection forecasts may be most applicable at short time scales. Important implications for ESA-based targeting methods employed with real-time ensemble systems are also discussed.

This study investigates whether the thermodynamics of supercell rear-flank outflow can be inferred from the propagation speed and vertical structure of the rear-flank gust front. To quantify the relationship between outflow thermodynamic deficit and gust front structure, CM1 is applied as a two-dimensional cold pool model to assess the vertical slope of cold pools of varying strength in different configurations of ambient shear. The model was run with both free-slip and semislip lower boundary conditions and the results were compared to observations of severe thunderstorm outflow captured by the Texas Tech University Ka-band mobile radars. Simulated cold pools in the free-slip model achieve the propagation speeds predicted by cold pool theory, while cold pool speeds in the semislip model propagate slower. Density current theory is applied to the observed cold pools and predicts the cold pool speed to within about 2 m s−1. Both the free-slip and semislip model results reveal that, in the same sheared flow, the edge of a strong cold pool is less inclined than that of a weaker cold pool. Also, a cold pool in weak ambient shear has a steeper slope than the same cold pool in stronger ambient shear. Nonlinear regressions performed on data from both models capture the proper dependence of slope on buoyancy and shear, but the free-slip model does not predict observed slopes within acceptable error, and the semislip model overpredicts the cold pool slope for all observed cases, but with uncertainty due to shear estimation.

This paper analyzes the impact of the educational conditions of Brazil’s Bolsa Família Program on the school enrollment, age-grade discrepancy, and labor of children benefiting from the program. The main hypotheses of this paper is that a child who lives in a household that receives the benefit has higher chances of being in school, lower chances to have age-grade discrepancy, and lower chances of working. Data used are from the 2010 Brazilian Demographic Census. Logistical models were estimated for each dependent variable (school enrollment, age-grade discrepancy, and child labor) and for three household income thresholds. Independent variables account for characteristics related to the household, mother, child, and whether the household was receiving Bolsa Família. The income thresholds are a maximum household per capita income of 70 Brazilian Reais, 140 Brazilian Reais (the official maximum value for eligibility into the Bolsa Família in 2010), and 280 Brazilian Reais. Models were also estimated separated by the rural and urban areas in the official income threshold. Results follow initial hypotheses of higher chances of school enrollment and lower chances of age-grade discrepancy among children who receive Bolsa Família. However, models also suggest higher chances of child labor among beneficiaries of the program.

A recent study found that surface hodographs over the Great Plains of the United States turn in a counterclockwise direction with time. This observed turning is opposite of the clockwise turning observed (and expected, based on theory) at higher altitudes. Using a mesoscale forecast model, the same study shows that it has the same hodograph behavior as found in the observations. The study further shows that the reason for this anomalous counterclockwise turning is the decoupling of the surface layer from the boundary layer after sunset and its recoupling after sunrise. The present paper presents a simple model for this behavior by extending a recent analytical model for the diurnal oscillation to include the surface-layer effect. In addition, selected solution features are analyzed in terms of several of the nondimensional input parameters.