Search Results
You are looking at 41 - 50 of 67 items for
- Author or Editor: George H. Bryan x
- Refine by Access: All Content x
Abstract
Simulations of supercell thunderstorms in a sheared convective boundary layer (CBL), characterized by quasi-two-dimensional rolls, are compared with simulations having horizontally homogeneous environments. The effects of boundary layer convection on the general characteristics and the low-level mesocyclones of the simulated supercells are investigated for rolls oriented either perpendicular or parallel to storm motion, as well as with and without the effects of cloud shading.
Bulk measures of storm strength are not greatly affected by the presence of rolls in the near-storm environment. Though boundary layer convection diminishes with time under the anvil shadow of the supercells when cloud shading is allowed, simulations without cloud shading suggest that rolls affect the morphology and evolution of supercell low-level mesocyclones. Initially, CBL vertical vorticity perturbations are enhanced along the supercell outflow boundary, resulting in nonnegligible near-ground vertical vorticity regardless of roll orientation. At later times, supercells that move perpendicular to the axes of rolls in their environment have low-level mesocyclones with weaker, less persistent circulation compared to those in a similar horizontally homogeneous environment. For storms moving parallel to rolls, the opposite result is found: that is, low-level mesocyclone circulation is often enhanced relative to that in the corresponding horizontally homogeneous environment.
Abstract
Simulations of supercell thunderstorms in a sheared convective boundary layer (CBL), characterized by quasi-two-dimensional rolls, are compared with simulations having horizontally homogeneous environments. The effects of boundary layer convection on the general characteristics and the low-level mesocyclones of the simulated supercells are investigated for rolls oriented either perpendicular or parallel to storm motion, as well as with and without the effects of cloud shading.
Bulk measures of storm strength are not greatly affected by the presence of rolls in the near-storm environment. Though boundary layer convection diminishes with time under the anvil shadow of the supercells when cloud shading is allowed, simulations without cloud shading suggest that rolls affect the morphology and evolution of supercell low-level mesocyclones. Initially, CBL vertical vorticity perturbations are enhanced along the supercell outflow boundary, resulting in nonnegligible near-ground vertical vorticity regardless of roll orientation. At later times, supercells that move perpendicular to the axes of rolls in their environment have low-level mesocyclones with weaker, less persistent circulation compared to those in a similar horizontally homogeneous environment. For storms moving parallel to rolls, the opposite result is found: that is, low-level mesocyclone circulation is often enhanced relative to that in the corresponding horizontally homogeneous environment.
Abstract
The structure and intensity of tornado-like vortices are examined using large-eddy simulations (LES) in an idealized framework. The analysis focuses on whether the simulated boundary layer contains resolved turbulent eddies, and whether most of the vertical component of turbulent momentum flux is resolved rather than parameterized. Initial conditions are first generated numerically using a “precursor simulation” with an axisymmetric model. A three-dimensional “baseline” LES is then integrated using these initial conditions plus random perturbations. With this baseline approach, the inner core of the simulated vortex clearly contains resolved turbulent eddies (as expected); however, the boundary layer inflow has very weak resolved turbulent eddies, and the subgrid model accounts for most of the vertical turbulent momentum flux (contrary to the design of these simulations). To overcome this problem, a second precursor simulation is conducted in which resolved turbulent fluctuations develop within a smaller, doubly periodic LES domain. Perturbation flow fields from this precursor LES are then “injected” into the large-domain LES at a specified radius. With this approach, the boundary layer inflow clearly contains resolved turbulent fluctuations, often organized as quasi-2D rolls, which persist into the inner core of the simulation; thus, the simulated tornado-like vortex and its inflowing boundary layer can be characterized as LES. When turbulence is injected, the inner-core vortex structure is always substantially different, the boundary layer inflow is typically deeper, and in most cases the maximum wind speeds are reduced compared to the baseline simulation.
Abstract
The structure and intensity of tornado-like vortices are examined using large-eddy simulations (LES) in an idealized framework. The analysis focuses on whether the simulated boundary layer contains resolved turbulent eddies, and whether most of the vertical component of turbulent momentum flux is resolved rather than parameterized. Initial conditions are first generated numerically using a “precursor simulation” with an axisymmetric model. A three-dimensional “baseline” LES is then integrated using these initial conditions plus random perturbations. With this baseline approach, the inner core of the simulated vortex clearly contains resolved turbulent eddies (as expected); however, the boundary layer inflow has very weak resolved turbulent eddies, and the subgrid model accounts for most of the vertical turbulent momentum flux (contrary to the design of these simulations). To overcome this problem, a second precursor simulation is conducted in which resolved turbulent fluctuations develop within a smaller, doubly periodic LES domain. Perturbation flow fields from this precursor LES are then “injected” into the large-domain LES at a specified radius. With this approach, the boundary layer inflow clearly contains resolved turbulent fluctuations, often organized as quasi-2D rolls, which persist into the inner core of the simulation; thus, the simulated tornado-like vortex and its inflowing boundary layer can be characterized as LES. When turbulence is injected, the inner-core vortex structure is always substantially different, the boundary layer inflow is typically deeper, and in most cases the maximum wind speeds are reduced compared to the baseline simulation.
Abstract
A new microphysics scheme has been developed based on the prediction of bulk particle properties for a single ice-phase category, in contrast to the traditional approach of separating ice into various predefined species (e.g., cloud ice, snow, and graupel). In this paper, the new predicted particle properties (P3) scheme, described in Part I of this series, is tested in three-dimensional simulations using the Weather Research and Forecasting (WRF) Model for two contrasting well-observed cases: a midlatitude squall line and winter orographic precipitation. Results are also compared with simulations using other schemes in WRF. Simulations with P3 can produce a wide variety of particle characteristics despite having only one free ice-phase category. For the squall line, it produces dense, fast-falling, hail-like ice near convective updraft cores and lower-density, slower-falling ice elsewhere. Sensitivity tests show that this is critical for simulating high precipitation rates observed along the leading edge of the storm. In contrast, schemes that represent rimed ice as graupel, with lower fall speeds than hail, produce lower peak precipitation rates and wider, less distinct, and less realistic regions of high convective reflectivity. For the orographic precipitation case, P3 produces areas of relatively fast-falling ice with characteristics of rimed snow and low- to medium-density graupel on the windward slope. This leads to less precipitation on leeward slopes and more on windward slopes compared to the other schemes that produce large amounts of snow relative to graupel (with generally the opposite for schemes with significant graupel relative to snow). Overall, the new scheme produces reasonable results for a reduced computational cost.
Abstract
A new microphysics scheme has been developed based on the prediction of bulk particle properties for a single ice-phase category, in contrast to the traditional approach of separating ice into various predefined species (e.g., cloud ice, snow, and graupel). In this paper, the new predicted particle properties (P3) scheme, described in Part I of this series, is tested in three-dimensional simulations using the Weather Research and Forecasting (WRF) Model for two contrasting well-observed cases: a midlatitude squall line and winter orographic precipitation. Results are also compared with simulations using other schemes in WRF. Simulations with P3 can produce a wide variety of particle characteristics despite having only one free ice-phase category. For the squall line, it produces dense, fast-falling, hail-like ice near convective updraft cores and lower-density, slower-falling ice elsewhere. Sensitivity tests show that this is critical for simulating high precipitation rates observed along the leading edge of the storm. In contrast, schemes that represent rimed ice as graupel, with lower fall speeds than hail, produce lower peak precipitation rates and wider, less distinct, and less realistic regions of high convective reflectivity. For the orographic precipitation case, P3 produces areas of relatively fast-falling ice with characteristics of rimed snow and low- to medium-density graupel on the windward slope. This leads to less precipitation on leeward slopes and more on windward slopes compared to the other schemes that produce large amounts of snow relative to graupel (with generally the opposite for schemes with significant graupel relative to snow). Overall, the new scheme produces reasonable results for a reduced computational cost.
Abstract
The triggering of convective orographic rainbands by small-scale topographic features is investigated through observations of a banded precipitation event over the Oregon Coastal Range and simulations using a cloud-resolving numerical model. A quasi-idealized simulation of the observed event reproduces the bands in the radar observations, indicating the model’s ability to capture the physics of the band-formation process. Additional idealized simulations reinforce that the bands are triggered by lee waves past small-scale topographic obstacles just upstream of the nominal leading edge of the orographic cloud. Whether a topographic obstacle in this region is able to trigger a strong rainband depends on the phase of its lee wave at cloud entry. Convective growth only occurs downstream of obstacles that give rise to lee-wave-induced displacements that create positive vertical velocity anomalies wc and nearly zero buoyancy anomalies bc as air parcels undergo saturation. This relationship is quantified through a simple analytic condition involving wc , bc , and the static stability N 2 m of the cloud mass. Once convection is triggered, horizontal buoyancy gradients in the cross-flow direction generate circulations that align the bands parallel to the flow direction.
Abstract
The triggering of convective orographic rainbands by small-scale topographic features is investigated through observations of a banded precipitation event over the Oregon Coastal Range and simulations using a cloud-resolving numerical model. A quasi-idealized simulation of the observed event reproduces the bands in the radar observations, indicating the model’s ability to capture the physics of the band-formation process. Additional idealized simulations reinforce that the bands are triggered by lee waves past small-scale topographic obstacles just upstream of the nominal leading edge of the orographic cloud. Whether a topographic obstacle in this region is able to trigger a strong rainband depends on the phase of its lee wave at cloud entry. Convective growth only occurs downstream of obstacles that give rise to lee-wave-induced displacements that create positive vertical velocity anomalies wc and nearly zero buoyancy anomalies bc as air parcels undergo saturation. This relationship is quantified through a simple analytic condition involving wc , bc , and the static stability N 2 m of the cloud mass. Once convection is triggered, horizontal buoyancy gradients in the cross-flow direction generate circulations that align the bands parallel to the flow direction.
Abstract
This study is the first in a series that investigates the effects of turbulence in the boundary layer of a tornado vortex. In this part, axisymmetric simulations with constant viscosity are used to explore the relationships between vortex structure, intensity, and unsteadiness as functions of diffusion (measured by a Reynolds number Re
r
) and rotation (measured by a swirl ratio S
r
). A deep upper-level damping zone is used to prevent upper-level disturbances from affecting the low-level vortex. The damping zone is most effective when it overlaps with the specified convective forcing, causing a reduction to the effective convective velocity scale W
e
. With this damping in place, the tornado-vortex boundary layer shows no sign of unsteadiness for a wide range of parameters, suggesting that turbulence in the tornado boundary layer is inherently a three-dimensional phenomenon. For high Re
r
, the most intense vortices have maximum mean tangential winds well in excess of W
e
, and maximum mean vertical velocity exceeds 3 times W
e
. In parameter space, the most intense vortices fall along a line that follows 
Abstract
This study is the first in a series that investigates the effects of turbulence in the boundary layer of a tornado vortex. In this part, axisymmetric simulations with constant viscosity are used to explore the relationships between vortex structure, intensity, and unsteadiness as functions of diffusion (measured by a Reynolds number Re
r
) and rotation (measured by a swirl ratio S
r
). A deep upper-level damping zone is used to prevent upper-level disturbances from affecting the low-level vortex. The damping zone is most effective when it overlaps with the specified convective forcing, causing a reduction to the effective convective velocity scale W
e
. With this damping in place, the tornado-vortex boundary layer shows no sign of unsteadiness for a wide range of parameters, suggesting that turbulence in the tornado boundary layer is inherently a three-dimensional phenomenon. For high Re
r
, the most intense vortices have maximum mean tangential winds well in excess of W
e
, and maximum mean vertical velocity exceeds 3 times W
e
. In parameter space, the most intense vortices fall along a line that follows
Abstract
The stratocumulus cloud–capped boundary layer under a sharp inversion is a challenging regime for large-eddy simulation (LES). Here, data from the first research flight of the Second Dynamics and Chemistry of the Marine Stratocumulus field study are used to evaluate the effect of different LES turbulence closures. Six different turbulence models, including traditional TKE and Smagorinsky models and more advanced models that employ explicit filtering and reconstruction, are tested. The traditional models produce unrealistically thin clouds and a decoupled boundary layer as compared with other more advanced models. Traditional models rely on specified subfilter-scale (SFS) Prandtl and Schmidt numbers to obtain SFS eddy diffusivity from eddy viscosity, whereas dynamic models can compute SFS eddy diffusivity independently through dynamic procedures. The effective SFS Prandtl number in dynamic models is found to be ~0.5 below the cloud and ~10 inside the cloud layer, implying minimized mixing in the cloud. In contrast, the SFS Prandtl number in traditional models is about 1/3 throughout the entire boundary layer, suggesting spuriously strong mixing in the cloud. The SFS Schmidt number in the dynamic models also changes independently from the SFS Prandtl number, whereas in traditional models they are identical, meaning that the efficiency of the turbulent mixing of water content is forced to be the same as that of heat. Since it is very difficult to know in advance the SFS Prandtl and Schmidt numbers in a given flow, dynamic models may provide a more realistic representation of scalar mixing in LES.
Abstract
The stratocumulus cloud–capped boundary layer under a sharp inversion is a challenging regime for large-eddy simulation (LES). Here, data from the first research flight of the Second Dynamics and Chemistry of the Marine Stratocumulus field study are used to evaluate the effect of different LES turbulence closures. Six different turbulence models, including traditional TKE and Smagorinsky models and more advanced models that employ explicit filtering and reconstruction, are tested. The traditional models produce unrealistically thin clouds and a decoupled boundary layer as compared with other more advanced models. Traditional models rely on specified subfilter-scale (SFS) Prandtl and Schmidt numbers to obtain SFS eddy diffusivity from eddy viscosity, whereas dynamic models can compute SFS eddy diffusivity independently through dynamic procedures. The effective SFS Prandtl number in dynamic models is found to be ~0.5 below the cloud and ~10 inside the cloud layer, implying minimized mixing in the cloud. In contrast, the SFS Prandtl number in traditional models is about 1/3 throughout the entire boundary layer, suggesting spuriously strong mixing in the cloud. The SFS Schmidt number in the dynamic models also changes independently from the SFS Prandtl number, whereas in traditional models they are identical, meaning that the efficiency of the turbulent mixing of water content is forced to be the same as that of heat. Since it is very difficult to know in advance the SFS Prandtl and Schmidt numbers in a given flow, dynamic models may provide a more realistic representation of scalar mixing in LES.
Abstract
Large-eddy simulation (LES) has been an essential tool in the development of theory and parameterizations for clouds, but when applied to stratocumulus clouds under sharp temperature inversions, many LES models produce an unrealistically thin cloud layer and a decoupled boundary layer structure. Here, explicit filtering and reconstruction are used for simulation of stratocumulus clouds observed during the first research flight (RF01) of the Second Dynamics and Chemistry of the Marine Stratocumulus field study (DYCOMS II). The dynamic reconstruction model (DRM) is used within an explicit filtering and reconstruction framework, partitioning subfilter-scale motions into resolvable subfilter scales (RSFSs) and unresolvable subgrid scales (SGSs). The former are reconstructed, and the latter are modeled. Differing from traditional turbulence models, the reconstructed RSFS stress/fluxes can produce backscatter of turbulence kinetic energy (TKE) and, importantly, turbulence potential energy (TPE). The modeled backscatter reduces entrainment at the cloud top and, meanwhile, strengthens resolved turbulence through preserving TKE and TPE, resulting in a realistic boundary layer with an adequate amount of cloud water and vertically coupled turbulent eddies. Additional simulations are performed in the terra incognita, when the grid spacing of a simulation becomes comparable to the size of the most energetic eddies. With 20-m vertical and 1-km horizontal grid spacings, simulations using DRM provide a reasonable representation of bulk properties of the stratocumulus-capped boundary layer.
Abstract
Large-eddy simulation (LES) has been an essential tool in the development of theory and parameterizations for clouds, but when applied to stratocumulus clouds under sharp temperature inversions, many LES models produce an unrealistically thin cloud layer and a decoupled boundary layer structure. Here, explicit filtering and reconstruction are used for simulation of stratocumulus clouds observed during the first research flight (RF01) of the Second Dynamics and Chemistry of the Marine Stratocumulus field study (DYCOMS II). The dynamic reconstruction model (DRM) is used within an explicit filtering and reconstruction framework, partitioning subfilter-scale motions into resolvable subfilter scales (RSFSs) and unresolvable subgrid scales (SGSs). The former are reconstructed, and the latter are modeled. Differing from traditional turbulence models, the reconstructed RSFS stress/fluxes can produce backscatter of turbulence kinetic energy (TKE) and, importantly, turbulence potential energy (TPE). The modeled backscatter reduces entrainment at the cloud top and, meanwhile, strengthens resolved turbulence through preserving TKE and TPE, resulting in a realistic boundary layer with an adequate amount of cloud water and vertically coupled turbulent eddies. Additional simulations are performed in the terra incognita, when the grid spacing of a simulation becomes comparable to the size of the most energetic eddies. With 20-m vertical and 1-km horizontal grid spacings, simulations using DRM provide a reasonable representation of bulk properties of the stratocumulus-capped boundary layer.
Abstract
A large-eddy simulation (LES) framework with an “eddy injection” technique has been developed that ensures a majority of turbulent kinetic energy in numerically simulated tornado-like vortices is represented by resolved eddies. This framework is used to explore the relationships between environmental forcing mechanisms, surface boundary conditions, and tornado vortex structure, intensity, and wind gusts. Similar to previous LES studies, results show that the maximum time- and azimuthal-mean tangential winds {V}max can be well in excess of the “thermodynamic speed limit,” which is 66 m s−1 for most of the simulations. Specifically, {V}max exceeds this speed by values ranging from 21% for a large, high-swirl vortex to 59% for a small, low-swirl vortex. Budgets of mean and eddy angular and radial momentum are used to show that resolved eddies in the tornado core act to reduce the wind speed at the location of {V}max, although they do transport angular momentum downward into the lowest levels of the boundary layer, increasing low-level swirl.
Three measures of tornado intensity are introduced: maximum time–azimuthal-mean surface (10 m) horizontal wind speed ({S10}max), maximum 3-s gusts of S10 (S10-3s), and maximum vertical 3-s gusts at 10 m (W10-3s). While {S10}max is considerably less than {V}max, transient features in the boundary layer can generate S10-3s in excess of 150 m s−1, and W10-3s in excess of 100 m s−1. For high-swirl vortices, the extreme gusts are confined closer to the center, well inside the radius of maximum azimuthal-mean surface winds. For the low-swirl vortex, both the strongest mean winds and the extreme gusts are restricted to a very narrow core.
Abstract
A large-eddy simulation (LES) framework with an “eddy injection” technique has been developed that ensures a majority of turbulent kinetic energy in numerically simulated tornado-like vortices is represented by resolved eddies. This framework is used to explore the relationships between environmental forcing mechanisms, surface boundary conditions, and tornado vortex structure, intensity, and wind gusts. Similar to previous LES studies, results show that the maximum time- and azimuthal-mean tangential winds {V}max can be well in excess of the “thermodynamic speed limit,” which is 66 m s−1 for most of the simulations. Specifically, {V}max exceeds this speed by values ranging from 21% for a large, high-swirl vortex to 59% for a small, low-swirl vortex. Budgets of mean and eddy angular and radial momentum are used to show that resolved eddies in the tornado core act to reduce the wind speed at the location of {V}max, although they do transport angular momentum downward into the lowest levels of the boundary layer, increasing low-level swirl.
Three measures of tornado intensity are introduced: maximum time–azimuthal-mean surface (10 m) horizontal wind speed ({S10}max), maximum 3-s gusts of S10 (S10-3s), and maximum vertical 3-s gusts at 10 m (W10-3s). While {S10}max is considerably less than {V}max, transient features in the boundary layer can generate S10-3s in excess of 150 m s−1, and W10-3s in excess of 100 m s−1. For high-swirl vortices, the extreme gusts are confined closer to the center, well inside the radius of maximum azimuthal-mean surface winds. For the low-swirl vortex, both the strongest mean winds and the extreme gusts are restricted to a very narrow core.
Abstract
This study investigates droplet size distribution (DSD) characteristics from condensational growth and transport in Eulerian dynamical models with bin microphysics. A hierarchy of modeling frameworks is utilized, including parcel, one-dimensional (1D), and three-dimensional large-eddy simulation (LES). The bin DSDs from the 1D model, which includes only vertical advection and condensational growth, are nearly as broad as those from the LES and in line with observed DSD widths for stratocumulus clouds. These DSDs are much broader than those from Lagrangian microphysical calculations within a parcel framework that serve as a numerical benchmark for the 1D tests. In contrast, the bin-modeled DSDs are similar to the Lagrangian microphysical benchmark for a rising parcel in which Eulerian transport is not considered. These results indicate that numerical diffusion associated with vertical advection is a key contributor to broadening DSDs in the 1D model and LES. This DSD broadening from vertical numerical diffusion is unphysical, in contrast to the physical mixing processes that previous studies have indicated broaden DSDs in real clouds. It is proposed that artificial DSD broadening from vertical numerical diffusion compensates for underrepresented horizontal variability and mixing of different droplet populations in typical LES configurations with bin microphysics, or the neglect of other mechanisms that broaden DSDs such as growth of giant cloud condensation nuclei. These results call into question the ability of Eulerian dynamical models with bin microphysics to investigate the physical mechanisms for DSD broadening, even though they may reasonably simulate overall DSD characteristics.
Abstract
This study investigates droplet size distribution (DSD) characteristics from condensational growth and transport in Eulerian dynamical models with bin microphysics. A hierarchy of modeling frameworks is utilized, including parcel, one-dimensional (1D), and three-dimensional large-eddy simulation (LES). The bin DSDs from the 1D model, which includes only vertical advection and condensational growth, are nearly as broad as those from the LES and in line with observed DSD widths for stratocumulus clouds. These DSDs are much broader than those from Lagrangian microphysical calculations within a parcel framework that serve as a numerical benchmark for the 1D tests. In contrast, the bin-modeled DSDs are similar to the Lagrangian microphysical benchmark for a rising parcel in which Eulerian transport is not considered. These results indicate that numerical diffusion associated with vertical advection is a key contributor to broadening DSDs in the 1D model and LES. This DSD broadening from vertical numerical diffusion is unphysical, in contrast to the physical mixing processes that previous studies have indicated broaden DSDs in real clouds. It is proposed that artificial DSD broadening from vertical numerical diffusion compensates for underrepresented horizontal variability and mixing of different droplet populations in typical LES configurations with bin microphysics, or the neglect of other mechanisms that broaden DSDs such as growth of giant cloud condensation nuclei. These results call into question the ability of Eulerian dynamical models with bin microphysics to investigate the physical mechanisms for DSD broadening, even though they may reasonably simulate overall DSD characteristics.
Abstract
Supersaturation fluctuations in the atmosphere are critical for cloud processes. A nonlinear dependence on two scalars—water vapor and temperature—leads to different behavior than single scalars in turbulent convection. For modeling such multiscalar processes at subgrid scales (SGS) in large-eddy simulations (LES) or convection-permitting models, a new SGS scheme is implemented in CM1 that solves equations for SGS water vapor and temperature fluctuations and their covariance. The SGS model is evaluated using benchmark direct-numerical simulations (DNS) of turbulent Rayleigh–Bénard convection with water vapor as in the Michigan Tech Pi Cloud Chamber. This idealized setup allows thorough evaluation of the SGS model without complications from other atmospheric processes. DNS results compare favorably with measurements from the chamber. Results from LES using the new SGS model compare well with DNS, including profiles of water vapor and temperature variances, their covariance, and supersaturation variance. SGS supersaturation fluctuations scale appropriately with changes to the LES grid spacing, with the magnitude of SGS fluctuations decreasing relative to those at the resolved scale as the grid spacing is decreased. Sensitivities of covariance and supersaturation statistics to changes in water vapor flux relative to thermal flux are also investigated by modifying the sidewall conditions. Relative changes in water vapor flux substantially decrease the covariance and increase supersaturation fluctuations even away from boundaries.
Abstract
Supersaturation fluctuations in the atmosphere are critical for cloud processes. A nonlinear dependence on two scalars—water vapor and temperature—leads to different behavior than single scalars in turbulent convection. For modeling such multiscalar processes at subgrid scales (SGS) in large-eddy simulations (LES) or convection-permitting models, a new SGS scheme is implemented in CM1 that solves equations for SGS water vapor and temperature fluctuations and their covariance. The SGS model is evaluated using benchmark direct-numerical simulations (DNS) of turbulent Rayleigh–Bénard convection with water vapor as in the Michigan Tech Pi Cloud Chamber. This idealized setup allows thorough evaluation of the SGS model without complications from other atmospheric processes. DNS results compare favorably with measurements from the chamber. Results from LES using the new SGS model compare well with DNS, including profiles of water vapor and temperature variances, their covariance, and supersaturation variance. SGS supersaturation fluctuations scale appropriately with changes to the LES grid spacing, with the magnitude of SGS fluctuations decreasing relative to those at the resolved scale as the grid spacing is decreased. Sensitivities of covariance and supersaturation statistics to changes in water vapor flux relative to thermal flux are also investigated by modifying the sidewall conditions. Relative changes in water vapor flux substantially decrease the covariance and increase supersaturation fluctuations even away from boundaries.
