Mobile, polarimetric radar data were collected on a series of tornadoes that occurred near Dodge City, Kansas. A poststorm survey revealed a series of tornadic debris swaths in several dirt fields and high-resolution pictures of the tornado documented the visual characteristics of the tornado and the lofted debris cloud. The main rotational couplet associated with the tornado was identified in the single-Doppler velocities; however, no secondary rotational couplets were resolved in the low-level data performed during two consecutive volume scans. Numerical simulations have suggested that cycloidal damage swaths can result when debris is deposited as the low-level inflow turns upward in the corner region of the updraft annulus of the tornado core. This mechanism can dominate even when suction vortices are present in the simulations and can produce these swaths in the absence of these smaller-scale vortices. It is hypothesized that the observed cycloidal damage swaths were a result of the low-level inflow in the corner region of the tornado and not by the existence of suction vortices. Polarimetric data were combined with photographs of the tornado in order to document the lofted debris cloud and its relationship with the funnel. This analysis provided an opportunity to investigate whether recent findings describing the cross-correlation coefficient ρhv and differential reflectivity ZDR signatures of the lofted debris cloud could be replicated. Regions of low ρhv at the periphery of the funnel cloud suggesting high debris loading and a column of negative ZDR centered on the tornado believed to be produced by common debris alignment were noted. It is well known that some tornadoes produce smaller-scale vortices that rotate around the central axis of the main circulation. In addition, numerous aerial photographs have documented cycloidal debris marks within tornado damage tracks that traverse open fields. The prevailing theory shown in numerous textbooks is that these marks are produced by these vortices. The current study suggests that this widely accepted model for producing these marks may be incorrect. It is suggested that these cycloidal marks are produced by the main tornado circulation and not by the smaller-scale vortices in this case.

Ice crystal number is a critical ingredient in the potential climate impact of persistent contrails and contrail-induced cirrus. We perform an extensive set of large-eddy simulations (LES) of ice nucleation and growth within aircraft exhaust jets with an emphasis on assessing the importance of detailed plume mixing on the effective ice-number emission index (EIiceno) produced for different conditions. Parameter variations considered include ambient temperature, pressure, and humidity; initial aerosol origin (exhaust or ambient), number, and properties; and aircraft engine size. The LES are performed in a temporal representation with binned microphysics including the basics of activation of underlying aerosol, droplet growth, and freezing. We find that a box-model approach reproduces EIiceno from LES well for sufficiently low aerosol numbers or when crystal production is predominantly on ambient aerosol. For larger exhaust aerosol number the box model generally overestimates EIiceno and can underestimate the fraction from ultrafine aerosol. The effects of different parameters on EIiceno can largely be understood with simpler analytic models that are formulated in low and high aerosol-number limits. The simulations highlight the potential importance of "cold'' contrails, ambient ultrafine aerosols, crystal loss due to competition between different-sized crystals, and limitations on reducing EIiceno. We find EIiceno insensitive to engine size for lower aerosol numbers, but decreasing with increasing engine size for higher aerosol numbers. Temporal versus spatial representations for jet LES are compared in an appendix.

AbstractPast numerical simulation studies found that debris loading from sand-sized particles may substantially affect tornado dynamics, causing reductions in near-surface wind speeds up to 50%. To further examine debris loading effects, simulations are performed using a large-eddy simulation model with a two-way drag force coupling between air and sand. Simulations encompass a large range of surface debris fluxes that cause negligible to substantial impact on tornado dynamics for a high-swirl tornado vortex simulation.Simulations are considered for a specific case with a single vortex flow type (swirl ratio, intensity, and translation velocity) and a fixed set of debris and aerodynamic parameters. Thus, it is stressed that these findings apply to the specific flow and debris parameters herein and would likely vary for different flows or debris parameters. For this specific case, initial surface debris fluxes are varied over a factor of 16 384, and debris cloud mass varies by only 42% of this range becaus...

Large-eddy simulations with size-resolved microphysics are used to model persistent aircraft contrails and contrail-induced cirrus from a few wing spans behind the aircraft until their demise after many hours. Schemes for dynamic local ice binning and updating coupled radiation dynamically as needed in individual columns were developed for numerical efficiency, along with a scheme for maintaining realistic ambient turbulence over long times. These capabilities are used to study some of the critical dynamics involved in contrail evolution and to explore the simulation features required for adequate treatment of different components. A "quasi 3D" approach is identified as a useful approximation of the full dynamics, reducing the computation to allow a larger parameter space to be studied. Ice crystal number loss involving competition between different crystal sizes is found to be significant for both young contrails and aging contrail cirrus. As a consequence, the sensitivity to the initial number of ice crystals in the contrail above a threshold is found to decrease significantly overtime, and uncertainties in the ice deposition coefficient and Kelvin effect for ice crystals assume an increased importance. Atmospheric turbulence is found to strongly influence contrail properties and lifetime in some regimes. Water from fuel consumption is found to significantly reduce aircraft-wake-induced ice crystal loss in colder contrails. Ice crystal shape effects, coupled radiation, and precipitation dynamics are also considered. An extensive set of simulations exploring a large parameter space with this model are analyzed in a companion paper.

More than 200 large-eddy simulations of long-lived contrails from several-seconds age until their demise have been performed and their lifetime-integrated behavior has been analyzed. The simulations employ size-resolved microphysics and include variations of effective ice crystal number emission index, temperature, relative humidity with respect to ice, stratification, shear, supersaturated-layer depth, uplift/subsidence, and coupled radiation. Basic scaling behaviors are analyzed for contrail lifetime, width, ice mass, and surface area. Lifetimes exceeding 40h, widths exceeding 100 km, and ice masses exceeding 50 kg m(-1) of flight path were sometimes encountered. Distinct behavior regimes produced by radiative forcing are identified and found to be predicted by a simplified model. The lifetime-integrated ice crystal surface area per length of flight path S-Sigma is used as an approximate metric of contrail significance, and a simple, physically based model is derived. Over much of the parameter space, S-Sigma is found to vary approximately simply as the product of the maximum contrail depth and the effective number of ice crystals per flight path; other parameters have their impact on S-Sigma dominantly through their effects on these two quantities. Model and simulation results highlight the importance of crystal number loss mechanisms, the interaction between shear and ice sedimentation, the depth of the supersaturated layer below flight level, and the potential integrated significance of "cold" subvisible contrails. The results can aid in estimating the effects of more complex contrail scenarios or mitigation strategies and in understanding some aspects of natural-cirrus dynamics.

Aeolian processes are the most active processes modifying the surface of Mars under present day climatic conditions. Besides wind streak changes and dune and ripple migrations, active dust devils occur frequently leaving numerous tracks on the Martian surface. These dust devil tracks (DDTs) are characterized by albedo changes with respect to their surroundings and are suggested to be caused by erosion of dust exposing coarser grained material. Here we show that DDTs with a cycloidal pattern analyzed in situ in southern Peru are formed by erosion of very coarse sands at the outer margins and its subsequent annular deposition in the central parts of dust devils. Field observations are supported by large-eddy simulations using typical dust devil parameters resembling the cycloidal morphology of the DDTs. Cycloidal DDTs observed on Mars resembling the Peruvian DDTs suggest an equivalent formation mechanism. Our results imply that the formation of DDTs on Mars are not solely due to dust erosion but also depositional processes and dust devils are strong enough to redistribute coarser grained material such as sands; hence they might contribute to the modification of the present day Martian landscape.

The properties of the near-surface inflow are known to be a critical factor in determining tornado structure and intensity (see e.g., Lewellen and Lewellen (2007a)), so it is natural to expect that topography might significantly impact tornado behavior near the surface. Tornado damage surveys over nontrivial terrain seem to support this general conclusion. For example Forbes (1998) in his study of damage tracks in the May 31, 1985 tornado outbreak in western Pennsylvania (to our knowledge the most extensive field survey of potential topographic effects) highlighted four phenomena he attributed to topographic influence:

Motivated by simulations of slow-growing contrail cirrus, the solution of the diffusional growth equations for a population of spherical ice crystals or water droplets is reexamined. For forcing specified by the evolution of the total water content above saturation within a parcel (whether driven by vertical motions, radiative heating, turbulent mixing, etc.) three behavior regimes are identified: "very fast growth" that cannot equilibrate, "fast growth" with a narrowing size spectrum, and "slow growth" with a broadening spectrum. The boundaries between regimes, time scales involved, and evolution of the condensate mass, number, and supersaturation are determined. The slow-growth regime represents an example of "spectral ripening," with crystal or droplet numbers falling in time because of surface tension effects. Surprisingly the diffusional growth equations for the size spectrum evolution can be solved exactly in this case: in appropriate coordinates the spectral shape becomes steady, crystal or droplet numbers fall as a forcing-dependent power law, and the mean particle mass grows linearly with time. Dependence on different physical variables, fluctuating forcing, and modifications due to kinetic theory corrections are all considered. In the limit of zero external forcing on the parcel the size-spectrum solution is mathematically equivalent to a classic result in the theory of Ostwald ripening of solid solutions. It is argued that the slow-growth regime may be important in the evolution of contrail cirrus and perhaps in setting upper limits on droplet number densities in stratiform boundary layer clouds. The theoretical results are compared with parcel model simulations for illustration and to study numerical issues in binned microphysics models.

Twelve large-eddy simulations, with a wide range of microphysical representations, are compared to each other and to independent measurements. The measurements and the initial and forcing data for the simulations are taken from the undisturbed period of the Rain in Cumulus over the Ocean (RICO) field study. A regional downscaling of meteorological analyses is performed so as to provide forcing data consistent with the measurements. The ensemble average of the simulations plausibly reproduces many features of the observed clouds, including the vertical structure of cloud fraction, profiles of cloud and rain water, and to a lesser degree the population density of rain drops. The simulations do show considerable departures from one another in the representation of the cloud microphysical structure and the ensuant surface precipitation rates, increasingly so for the more simplified microphysical models. There is a robust tendency for simulations that develop rain to produce a shallower, somewhat more stable cloud layer. Relations between cloud cover and precipitation are ambiguous.

Figure 1: A sample of simulated surface tracks for a collection of intense, medium-swirl tornadoes, primarily varying the mass density of individual debris particles (increasing from left to right) and translation velocity (increasing from top to bottom). Dimensionless parameters are Sc = 3.0; Aa = 7.7; left to right: Av = 23.8, 15.9, 10.6, 7.2; and, top to bottom: At = 0.06, 0.17, 0.28 (see [1] for definitions). Grayscale intensities have been normalized to provide consistent appearances across the entire range of translation velocities.

Much of our understanding of tornado structure and dynamics is based on analysis of vortices that are near axisymmetry. Theoretically it reduces a threedimensional problem to a two-dimensional one (or even one-dimension since cylindrical symmetry often prevails). In field or laboratory observations it allows wind-fields to be inferred from a much more limited set of velocity measurements than would be required far from axisymmetry. Moreover the axisymmetric limit is still sufficiently rich to illustrate and study much of the complex dynamics of vortices that is seen in general. Nonetheless, better understanding and analysis methods for vortices far from axisymmetry would be extremely useful. Vortices in the field can deviate significantly from the axisymmetric limit, complicating their categorization even when the dynamics is basically that of the axisymmetric case. More fundamentally, while all of the important dynamics found in the axisymmetric limit is also important in the general case the converse is not true: important new dynamics can arise because of the asymmetry that has no counterpart in the symmetric case. As part of a longstanding effort to better understand the near-surface behavior of tornadoes we have conducted numerous large-eddy simulations (LES) with both realistic and idealized boundary conditions, often generating examples exhibiting asymmetric behavior in different guises. In this work we consider some aspects of the classification and analysis of tornado-like vortices away from the axisymmetric limit, gleaned from study of some of these LES sets. In the simulations the asymmetric behavior can arise from asymmetries imposed through the lateral boundary conditions, asymmetric forcing within the domain (e.g., surface friction acting on a translating vortex), or through vortex formation driven by internal convergence or shear instability. The cases can be inherently time varying or quasi-steady. Here we emphasize an important subclass: the embedding of a concentrated vortex within a larger-scale

This paper reports an intercomparison study on undisturbed trade wind cumulus convection under steady-state conditions as observed during the Barbados Oceanographic and Meteorological Experiment (BOMEX) with 10 large eddy simulation (LES) models. A main objective of this study is to obtain a quantitative assessment of the quality of the turbulent dynamics for this type of boundary layer clouds as produced by the different LES codes. A 6-h simulation shows excellent model-to-model agreement of the observed vertical thermodynamical structure, reasonable agreement of variances and turbulent fluxes, and good agreement of quantities conditionally sampled within the model clouds, such as cloud cover, liquid water, and cloud updraft strength. In the second part of this paper the LES dataset is used to evaluate simple models that are used in parameterizations of current general circulation models (GCMs). Finally, the relation of this work to subsequent LES studies of more complicated regimes is discussed, and guidance is given for the design of future observational studies of shallow cumulus boundary layers.

Cloud water sedimentation and drizzle in a stratocumulus-topped boundary layer are the focus of an intercomparison of large-eddy simulations. The context is an idealized case study of nocturnal stratocumulus under a dry inversion, with embedded pockets of heavily drizzling open cellular convection. Results from 11 groups are used. Two models resolve the size distributions of cloud particles, and the others parameterize cloud water sedimentation and drizzle. For the ensemble of simulations with drizzle and cloud water sedimentation, the mean liquid water path (LWP) is remarkably steady and consistent with the measurements, the mean entrainment rate is at the low end of the measured range, and the ensemble-average maximum vertical wind variance is roughly half that measured. On average, precipitation at the surface and at cloud base is smaller, and the rate of precipitation evaporation greater, than measured. Including drizzle in the simulations reduces convective intensity, increases boundary layer stratification, and decreases LWP for nearly all models. Including cloud water sedimentation substantially decreases entrainment, decreases convective intensity, and increases LWP for most models. In nearly all cases, LWP responds more strongly to cloud water sedimentation than to drizzle. The omission of cloud water sedimentation in simulations is strongly discouraged, regardless of whether or not precipitation is present below cloud base.

Debris clouds provide an important visual signature of tornadoes and can potentially significantly affect the wind structure, damage potential, and Doppler radar measurements of tornado wind speeds. To study such issues, the dynamics of finescale debris have been added to an existing high-resolution large-eddy simulation model of tornado dynamics. A so-called “two-fluid” or “Eulerian–Eulerian” approach is employed, together with a surface layer model for lofting and depositing debris. In this paper the debris implementation is described, three critical dimensionless parameters governing tornado debris effects are identified, and sample results from a large set of simulations of tornadoes with idealized debris are presented. The results demonstrate that the accumulation of small-scale debris within the surface layer and corner flow can significantly alter the wind speeds and flow structure of the tornado vortex within a few hundred meters of the surface. They suggest that the total mass of the debris cloud can reach tens of thousands of tons. Near the surface, the debris mass loading can be well above 1, the peak mean velocities can be reduced by as much as half, and the total momentum (air plus debris) can either significantly increase or decrease. Local air and debris velocities can differ significantly and in a nontrivial fashion, thereby complicating the interpretation of Doppler radar measurements of tornado structure. Debris fluctuations, centrifuging, negative buoyancy, and angular momentum transport are all significant mechanisms for the debris effects. A negative physical feedback reduces the sensitivity of the results to changes in the parameterization of the surface debris fluxes. The realistic simulation of tornado debris clouds and surface damage tracks should prove useful in identifying the dynamics governing their observed counterparts.