Influence of atmospheric vertical wind and turbulence on noctilucent clouds

Noctilucent clouds (NLCs), also called Polar Mesospheric Clouds (PMC), are high-altitude ice clouds in the summer mesopause region, mainly at polar latitudes. Their formation is sensitive to local atmospheric conditions, such as water vapor and temperature, as well as dynamics. This study investigates the impact of vertical transport, specifically turbulence and vertical winds, on NLC microphysics and background water vapor distribution using the three-dimensional transport model MIMAS (Mesospheric Ice Microphysics And tranSport model). We conducted numerical simulations of NLCs considering heterogeneous nucleation only with varying turbulence and vertical wind conditions, including cases with a 2× increase and a 0.5× reduction in the values of both parameters. Our results show that increased turbulence broadens the vertical distribution of ice particles, forming smaller ice particles that result in less bright clouds and a reduced freeze-drying effect, with a ∼7% decrease in peak mean particle radius and a ∼30% drop in peak mean brightness. In contrast, increased vertical winds enhance the upward transport of water vapor and reduce the net downward sedimentation speed of ice particles, thereby increasing water vapor concentration at mesopause altitudes and allowing particles to remain longer in supersaturated regions. This promotes the growth of larger ice particles and results in a ∼26% increase in peak mean particle radius and a ∼260% increase in peak mean brightness. These results highlight the different roles of turbulence and vertical winds in shaping NLC properties and background water vapor distribution, with implications for improving the modeling of NLCs and improving our understanding of how NLCs respond to long-term changes in upper atmospheric conditions.

1 Introduction

Noctilucent clouds (NLCs), also known as polar mesospheric clouds (PMCs), are the highest clouds in the Earth’s atmosphere, forming in the mesosphere at altitudes around 80–85 km. Occurring during the summer months at latitudes greater than 50°N, NLCs are known for their luminous, wave-like glow in the twilight sky (Thomas, 1984; Gadsden and Schröder, 1989). NLCs form under extremely cold conditions in the upper mesosphere (below ∼150 K), where water vapor becomes supersaturated and condenses onto nanoscale dust particles, usually of meteoritic origin. These conditions occur at high latitudes during summer, as upward transport generally increases during summer, leading to adiabatic cooling in the mesopause. Gravity wave activity in this region amplifies this process, making the mesopause the coldest region on Earth during summer (Lübken, 1999). Although the water vapor concentration at these altitudes is very low (<10ppmv), the exceptionally low temperatures cause supersaturation and allow the formation of ice particles. As a result, their formation is influenced by a combination of environmental factors, such as temperature, water vapor concentration, and background dynamics (Baumgarten et al., 2010). These factors are crucial in determining the height, brightness, and particle size of NLCs (Rapp and Thomas, 2006). This makes them a unique natural tracer for studying the impacts of global change on the upper mesosphere (Thomas and Olivero, 1989; Russell et al., 2014; Hervig et al., 2016; Lübken et al., 2018). In other words, observations of mesospheric ice clouds are an important tool for characterizing the polar summer mesopause region (Rapp and Thomas, 2006).

The upper mesosphere is characterized by significant turbulence and vertical wind fluctuations, primarily driven by gravity waves and tides. Gravity wave activity in the mesosphere exhibits seasonal and altitude-dependent variations, with some evidence of long-term trends that may be influenced by changes in mesospheric winds and lower atmospheric conditions (Hoffmann et al., 2010; Jacobi et al., 2025). Generated by mountain wave drag, convective storms, and wind shear, these waves propagate upward from the troposphere and stratosphere, depositing energy and momentum as they dissipate in the mesosphere (Fritts and Alexander, 2003). This energy deposition not only drives the large-scale circulation but also leads to localized turbulence and changes in vertical wind patterns, which are crucial to understanding mesospheric dynamics. The variations in gravity wave activity, therefore, influence the transport and mixing of water vapor, affecting its availability for the nucleation and growth of ice particles. Vertical winds, in particular, play a key role in ice particle sedimentation and growth. As a result, they can alter the altitude and optical properties of NLCs, and their effects are often directly visible in NLC structures (Witt, 1962; Fritts et al., 1993; Fiedler et al., 2003; Baumgarten and Fritts, 2014). While most previous studies on NLCs have focused on the role of temperature variations and increased water vapor concentrations in their formation and microphysical characteristics, the impact of atmospheric dynamics, particularly turbulence and vertical winds, has received comparatively less attention. Despite their significance, the effects of these dynamic processes on global-scale NLC properties remain poorly understood (Baumgarten et al., 2008; Fritts et al., 2022).

Understanding the interplay between turbulence, vertical winds, and the background atmosphere is essential for accurately characterizing NLC formation and its impact on water vapor profiles. Due to their sensitivity to gravity wave activity, NLCs are effective tracers for examining wave-induced processes in the upper atmosphere and investigating how turbulence and vertical transport influence NLCs, thus providing valuable insight into gravity wave dynamics and their contribution to mesospheric variability (Dalin et al., 2016; Kjellstrand et al., 2022). This study uses model simulations to explore how these processes affect the mean properties of NLCs and the background water vapor distribution. This will contribute to more accurate modeling of NLC dynamics and improve predictions of future trends in their behavior.

2 Model setup and numerical experiments

2.1 Model description

The overall model framework integrates two model components: the Leibniz Institute Middle Atmosphere (LIMA) model and the Mesospheric Ice Microphysics And TranSport (MIMAS) model. The main concept behind this approach is to use LIMA to simulate the background atmospheric conditions in the Northern Hemisphere (Berger, 2008; Lübken et al., 2009; Lübken and Berger, 2011; Lübken et al., 2013), which supplies the necessary background dynamics, including temperature, wind fields, air pressure, and density for the mesospheric region (e.g., Lübken et al., 2021). These conditions are then input into the MIMAS model to calculate the properties of NLCs (see Figure 1 in Vellalassery et al., 2023).

Figure 1

Figure 1. Schematic diagram showing the key vertical transport processes that influence ice particles (represented as black dots) in the upper mesosphere. These include upward or downward motion driven by vertical winds, mixing caused by turbulence, and the downward sedimentation of ice particles due to gravity.

MIMAS is a three-dimensional Lagrangian transport and microphysics model developed for NLC studies (e.g., Berger and von Zahn, 2002; Kiliani et al., 2013; Lübken et al., 2018). MIMAS simulates the life cycle of ice particles in the polar mesopause by tracking their interactions with the background atmosphere. NLC simulations focus on the Northern Hemisphere between 37° and 90° latitude. The model has a horizontal resolution of 1° in latitude and 3° in longitude (180° West to 180° East), and a vertical resolution of 100 m covering altitudes from 77.8 km to 94.1 km (Berger and Lübken, 2015; Lübken et al., 2018; Lübken et al., 2021). In MIMAS, NLC formation occurs when the mesospheric environment becomes supersaturated with water vapor. Under these conditions, meteoric dust particles serve as nucleation sites, triggering condensation and subsequent ice particle growth. In MIMAS, 40 million dust particles are available for microphysical processes, serving as potential nuclei for ice particles. The size distribution of these dust particles is based on Hunten et al. (1980), who investigated the properties of “meteoritic smoke particles” in detail and modeled their formation, including the processes of coagulation, sedimentation, and turbulent diffusion. The movement of these particles is influenced by advection, sedimentation, and turbulent diffusion. While horizontal transport is handled purely through advection, vertical motion is also affected by both sedimentation and eddy diffusion. Ice particle formation and growth in the model are based on standard microphysical processes, including the Kelvin effect, which sets a size-dependent supersaturation threshold for particle growth, and the degree of water vapor saturation (Berger and Lübken, 2015; Gadsden and Schröder, 1989), assuming spherical ice particles. In the mesosphere, ice particles may form through heterogeneous nucleation on meteoric dust or through homogeneous nucleation (Murray and Jensen, 2010). However, homogeneous nucleation requires extremely high supersaturation values (S > 1,000), which are not achieved under typical NLC conditions. Although ion-induced nucleation has been proposed (Sugiyama, 1994), Gumbel et al. (2003) concluded that it is unlikely to make a significant contribution. Because heterogeneous nucleation is regarded as the dominant and physically relevant mechanism for NLC formation (Gadsden, 1982; Turco et al., 1982; Tanaka et al., 2022), this is the only nucleation pathway implemented in MIMAS (Berger and von Zahn, 2002). Recent studies examining mesospheric variability (e.g., Zhang et al., 2022; Duft et al., 2015) also implicitly rely on heterogeneous nucleation as the dominant pathway.

MIMAS follows the trajectories and growth of 40 million particles (dust plus ice), with updates occurring every 3 min for transport processes and every 90 s for microphysics and water vapor advection. At both the upper and lower vertical boundaries, microphysical particles (dust/ice) that reach the model boundary are reseeded back into randomly selected locations within the initial seeding region (83 km to the mesopause, 55–90°N) (von Zahn and Berger, 2003; Kiliani, 2014). This maintains a quasi-steady population of meteoric smoke particles throughout the NLC season. Collisions and aggregation are not included in MIMAS, as particle number densities in the mesosphere are low (∼10⁴–10⁵ m^-3), making coagulation negligible (Turco et al., 1982). In MIMAS, latent heat released during ice particle condensation or sublimation is not coupled to the model thermodynamics and is therefore not considered in this study. Previous studies indicate that the latent heat release associated with mesospheric ice particles is likely negligible (Espy and Jutt, 2002). Regarding water vapor, MIMAS uses initial water vapor profiles that are constant at each pressure level at the beginning of the season (May 10). The water vapor profile was obtained from the model simulations of Körner and Sonnemann (2001), which correspond to an average of over 10 days (Berger and von Zahn, 2002). Water vapor transport, a key driver of NLC formation, is modeled via advection by background winds and turbulent diffusion. In MIMAS, water vapor originates primarily from vertical transport from the troposphere and in situ chemical production via methane (CH₄) oxidation, a dominant source in the mesosphere (Brasseur and Solomon, 2005). Below the lower boundary of MIMAS, two factors determine the mixing ratio of H₂O in the stratosphere: first, the transport of H₂O from the troposphere, and second, the oxidation of methane (CH₄), where each CH₄ molecule produces two H₂O molecules. Through photochemical processes, methane almost entirely converts into H₂O in the mesosphere (Lübken et al., 2018). The mesospheric water vapor is advected by wind fields, dispersed through vertical turbulent diffusion, and reduced via photodissociation caused by solar radiation. In MIMAS, the advection of particles is governed by solving the equation of motion for each particle n, where the position evolves based on the local wind velocity. The basic form of the motion is:

${\mathit{dX}}_{\mathit{n}}/\mathit{dt}=\mathit{u}\left({\mathit{X}}_{\mathit{n}}\right)$

Where ${\mathit{X}}_n=\left(x_n,y_n,z_n\right)$ is the three-dimensional position of the particle, and u(X_n) is the wind vector at that position. To explicitly account for all directional components, the full set of motion equations is:

${\mathit{dx}}_{\mathit{n}}/\mathit{dt}=\mathit{u}\left({\mathit{X}}_{\mathit{n}}\right)$

${\mathit{dy}}_{\mathit{n}}/\mathit{dt}=\mathit{v}\left({\mathit{X}}_{\mathit{n}}\right)$

${\mathit{dz}}_{\mathit{n}}/\mathit{dt}=\mathit{w}\left({\mathit{X}}_{\mathit{n}}\right)+{\mathit{w}}_{\mathit{s},\mathit{n}}+{\mathit{w}}_{\mathit{t},\mathit{n}}$

Here, u, v, and w represent the zonal, meridional, and vertical wind components, respectively. The terms w_s,n and w_t,n denote the sedimentation velocity and turbulent diffusion contribution for particle n (see Figure 1). MIMAS employs a size-dependent sedimentation velocity as implemented by Berger and von Zahn (2002). Sedimentation speed depends on the particle radius, the ambient temperature, and the air density.

In addition to the advection with wind, water vapor in MIMAS is continually mixed by turbulent eddy diffusion. In one dimension, this is described by Fick’s second law:

$\mathrm{\partial }\mathrm{c}/\mathrm{\partial }\mathrm{t}=K_{zz}{\mathrm{\partial }}^{2}c/\mathrm{\partial }{\mathrm{z}}^2$

Where c is the tracer concentration, and K_zz is the eddy diffusion coefficient. The general form in multiple dimensions is:

$\mathrm{\partial }\mathrm{c}/\mathrm{\partial }\mathrm{t}=K_{zz}{\nabla }^{2}c$

Turbulent diffusion is assumed to be isotropic. However, due to the very different spatial scales used in the model, approximately 100 km in the horizontal direction and 100 m in the vertical direction, the horizontal component of diffusion is negligible compared to the vertical component. For this reason, only vertical diffusion is implemented explicitly in MIMAS, and the turbulence is assumed to be horizontally homogeneous. The values used for the vertical eddy diffusion coefficient, referred to as K_zz, vary with altitude and are adapted from the sounding rocket climatology at 69°N presented in Lübken (1997). MIMAS uses K_zz values reduced to 25% of those in Lübken (1997) to account for the intermittency of turbulence (Berger and von Zahn, 2002). Below 85 km, turbulence is kept constant at the value specified for 85 km, because in this region the observed turbulence profile drops off too sharply (see Supplementary Figure S1), where turbulence is important for counteracting the freeze-drying effect (Berger and von Zahn, 2002). We refer to this turbulence profile in MIMAS as the reference turbulence throughout this paper. This profile was determined based on numerical experiments and comparison with observations (Berger and von Zahn, 2002). The model does not include latitudinal or temporal variations in K_zz, as implementing such dependencies would require a detailed understanding of the gravity wave breaking processes that drive mesospheric turbulence, and observations are available only for the latitude of about 69°N (Lübken, 1997). The particle turbulent motion scheme used in MIMAS is represented using a random walk scheme and explained in detail in Berger and von Zahn (2002). NLCs as modeled by LIMA/MIMAS are in general agreement with observations concerning: height, latitudinal extent, and long-term variations (Lübken et al., 2008; Lübken and Berger, 2011; Lübken et al., 2021).

Summarizing the model evolution, MIMAS was first introduced by Berger and von Zahn (2002) as a 3D Lagrangian model tracking ice particle microphysics in the polar mesopause. It was coupled to the LIMA background model in 2006 (Berger and Lübken, 2006) to incorporate realistic dynamics and diurnal variability. In Berger and Lübken (2015), the particle ensemble was increased to 40 million, and the domain was extended to cover 37°–90°N with 100 m vertical resolution. In 2018, methane oxidation was added as a major mesospheric H₂O source, allowing for the simulation of long-term NLC trends under climate forcing. The current version of the model used in this study relies on the model setup used in Lübken et al. (2018), and more detailed information about MIMAS and its earlier versions and validation with observations can be found in previous studies (Berger and von Zahn, 2002; Berger, 2008; Berger and Lübken, 2011; Berger and Lübken, 2015; Lübken et al., 2018; Lübken et al., 2021; Vellalassery et al., 2023; Vellalassery et al., 2024).

2.2 Numerical experiments

In this study, we investigate how variations in mesospheric dynamics influence NLC formation using numerical experiments performed with the MIMAS model. Specifically, we focus on the sensitivity of NLC formation to vertical transport processes, namely, turbulent diffusion and vertical wind, since these are key factors controlling the vertical distribution of water vapor and ice particles in the mesosphere.

To examine their individual effects, we conduct model runs in which either turbulent diffusion or vertical wind is modified, while all other model inputs are held constant. We performed several numerical experiments and selected two scenarios to illustrate the results: one with turbulence reduced to 0.5× and enhanced to 2× the reference value, and another with vertical wind reduced to 0.5× and enhanced to 2× the reference value. In total, four sensitivity runs were performed (0.5× and 2× for turbulence, and 0.5× and 2× for vertical wind), and these are the only model runs analysed in this study. MIMAS does not include an explicit eddy-viscosity formulation for momentum transport. In MIMAS, only the eddy diffusivity (K_zz) is varied, while momentum dissipation via gravity-wave friction in LIMA is kept fixed. Consequently, changes in K_zz alter the effective ratio between eddy diffusivity and eddy viscosity relative to the reference case. The reference values refer to the default values currently used in the MIMAS model under standard conditions (see Supplementary Figure S1 for the values of the eddy diffusion coefficient and July-averaged vertical wind). These experiments allow us to assess how variations in vertical transport affect the formation and growth of ice particles, as well as the redistribution of mesospheric water vapor. The impact of these variations is analyzed by comparing the model outputs and is discussed in the following sections. All analyses in this paper focus on the month of July, as it represents the peak of the NLC season during the summer (Lübken et al., 2018); however, we also briefly show the seasonal evolution. We have performed the numerical experiments for one representative year (2008) for comparability with earlier studies (Vellalassery et al., 2023; Vellalassery et al., 2024). The averaged profiles presented in Figures 2, 5, 6, 8 of this paper correspond to data averaged across the latitude band 69°N ± 3° and the month of July.

Figure 2

Figure 2. Zonal and monthly averaged water vapor profiles for July at 69°N from MIMAS for different turbulence (see insert). The solid lines show the background water vapor concentration with NLCs; the dashed lines show the water vapor concentration without NLCs.

3 Results and discussion

3.1 Impact of turbulence on water vapor

Figure 2 shows the zonal and monthly averaged water vapor (H₂O) profiles from MIMAS under three different turbulence conditions. A significant difference in the water vapor profiles is observed between simulations with and without NLCs. When comparing profiles with NLCs (solid lines) and without NLCs (dashed lines), the formation of NLCs leads to a reduction in water vapor concentration (dehydration) at altitudes above approximately 83 km, and an increase (hydration) below this altitude.

For water vapor to nucleate around dust particles, a region supersaturated with water vapor is required. The degree of saturation of air with water vapor is calculated as S = P_H2O/P_ice(T), where P_H2O is the partial pressure of water vapor, given by C_H2O · P_atm (where C_H2O is the water vapor mixing ratio), and P_ice(T) is the saturation vapor pressure over a flat ice surface. When S > 1, the environment is supersaturated, meaning ice particles can grow. When S < 1, the air is subsaturated, leading to the sublimation of ice particles. The ice particles start to form by nucleation around dust particles in the mesopause altitudes (around 88 km), where the value of S is maximum due to the extremely low temperatures. Then these particles start settling down due to gravity. During their sedimentation, they grow further by consuming water vapor from the background atmosphere, thereby depleting the water vapor from the background atmosphere. When they reach altitudes where the background temperature rises above the frost point (S < 1), ice particles sublimate and release the water vapor back to the background atmosphere. This leads to a significantly drier mesopause region compared to a scenario without NLC formation (for example, at 83 km, water vapor concentration is approximately 3 × 10⁹ molecules cm^-3, compared to about 5 × 10⁷ molecules cm^-3 at 88 km (Kiliani et al., 2013)). This process is referred to as the freeze-drying effect, discussed in several studies (Lübken et al., 2009; Vellalassery et al., 2024).

Regarding the sensitivity of water vapor profiles to different turbulence conditions, Figure 2 illustrates that without NLC formation (dashed lines), variations in turbulence have no significant impact on the vertical distribution of water vapor. With NLC formation (solid lines), however, the effect of turbulence is significantly pronounced on the background water vapor profiles. The hydration at sublimating altitudes decreases with an increase in turbulence (yellow line). i.e., The amplitude of water vapor enhancement in the sublimation zone (below ∼82 km) decreases with increasing turbulence. As illustrated in Figure 3, turbulence causes vertical mixing that moves particles up and down across short distances. When this mixing occurs well above the S = 1 threshold (e.g., above ∼83 km), particles remain within the water vapor supersaturated zone. In this region, increased turbulence enhances the vertical movements of ice particles but does not substantially alter their growth rates, since particles remain within the supersaturated zone (S > 1). In contrast, near the S = 1 boundary (around 82 km), an increase in turbulence pushes particles in and out of the saturation zone. As shown in the diagram, particles move back and forth between S > 1 (growth) and S < 1 (sublimation). Since sublimation occurs faster than growth (Rapp and Thomas, 2006; Kiliani et al., 2013), it plays an important role in determining particle properties in this region. Therefore, even small changes in turbulence near the S = 1 threshold can lead to significant differences in the microphysical growth of ice particles. This behavior is further modulated by the Kelvin effect, which describes the dependence of saturation vapor pressure on particle size. For small ice particles, surface tension generates an internal pressure, known as Laplace pressure, which increases the equilibrium vapor pressure at the ice particle surface. Consequently, smaller particles require a higher saturation ratio to remain in equilibrium with their environment. This means that as particles become smaller, the threshold S value needed for them to grow again increases (see Figure 2 in Berger and von Zahn, 2002). The S = 1 line shown in Figure 3 refers to saturation over a flat ice surface; the size-dependent supersaturation threshold due to the Kelvin effect is applied only in the microphysical growth calculation and is therefore not reflected in the figure.

Figure 3

Figure 3. Schematic representation of the effect of turbulent mixing on ice particles near the saturation ratio threshold (S = 1) in the mesopause region. The blue curve shows the vertical profile of saturation ratio (S) with altitude. The dashed vertical line indicates the saturation threshold (S = 1). The dotted horizontal line represents the lower boundary of water vapor supersaturation separating the ice formation and growth region (S > 1) from the sublimation region (S < 1). The circular arrows illustrate turbulent vertical motions of ice particles.

For example, turbulence can push particles downwards much faster (in the order of m/s) than their sedimentation velocity of about 10 cm/s for 50 nm particles at an altitude of 83 km (Kiliani et al., 2013). This rapid downward movement transports them into the subsaturated region (S < 1), where they quickly sublimate and decrease in size (Kiliani et al., 2013). If turbulence pushes these smaller particles back upward to their previous altitude, they may not grow again, even if the local S is still slightly above 1. This is because their reduced size now requires a higher S for growth due to the Kelvin effect, and the actual S is no longer sufficient. As a result, the particles continue to sublime rather than regrow. Therefore, increased turbulence at the lower boundary of saturation and below leads to a net ice loss and limits the maximum growth of ice particles. This reduces the maximum size that particles can grow, resulting in less water vapor being released during ice particle sublimation compared to cases with reduced turbulence.

3.2 Impact of turbulence on NLC properties

To investigate the response of NLC properties to turbulence variations, we analyzed time series of key NLC parameters. These include the number of ice particles, particle radius, and the backscatter coefficient (β) at 532 nm (hereafter also referred to as “NLC brightness”) across three different turbulence conditions. Figures 4a–c show the time series of ice particle number density profiles. The number density of ice particles represents the total number of ice particles per cubic centimeter. The figures indicate that the number of ice particles peaks just below the mesopause, around 87 km altitude.

Figure 4

Figure 4. Time series of NLC properties (number of ice particles, ice particle radius, and brightness) as a function of geometric height, zonally averaged at 69°N for the 2008 NLC season. The results are shown for three runs with different turbulence (see inset). (a–c) Number of ice particles, (d–f) ice particle radius, and (g–i) NLC brightness.

As mentioned earlier, the size distribution of dust particles (condensation nuclei, CN) in our model follows Hunten et al. (1980), spanning sizes from 1.2 to 3.7 nm, divided into five bins. The distribution is heavily weighted toward smaller particles: 86.5% in 1.2–1.7 nm, 11.7% in 1.7–2.2 nm, 1.7% in 2.2–2.7 nm, 0.2% in 2.7–3.2 nm, and 0.02% in 3.2–3.7 nm (see Berger and von Zahn, 2002, for more details). The number of ice particles peaks at mesopause altitudes due to a combination of temperature, water vapor saturation, and CN availability. The mesopause exhibits the lowest atmospheric temperatures, leading to the highest saturation ratio (S) of water vapor, promoting ice nucleation. For ice to form, a CN must be larger than a certain minimum radius (Berger and von Zahn, 2002). This means the CN must be big enough for water vapor to condense. The higher the saturation ratio (S), the smaller this minimum radius becomes (Berger and von Zahn, 2002). Therefore, even small CN can form ice particles at mesopause altitudes, where S is highest. As in the Hunten distribution, 86.5% of all particles have radii <1.7 nm, which contain large numbers of CN larger than the minimum radius. The presence of abundant small CNs and rapid nucleation leads to the formation of many ice particles. Since they all compete for the limited amount of water vapor, which restricts the growth of individual particles, leading to smaller ice particles but larger in number near the mesopause. As altitude decreases, the saturation ratio (S) also decreases due to rising temperatures, which significantly slows down the formation of new ice particles. This explains why the number of ice particles peaks around mesopause and gradually decreases at lower altitudes. At typical NLC altitudes around 83 km, the model calculated particle number densities range from 50 to 200 cm^-3, which reasonably agrees with the lidar observations (e.g., von Cossart et al., 1999; Baumgarten et al., 2008; Hervig et al., 2009).

The vertical distribution of ice particle number density changes with turbulence variation. Under low turbulence conditions, the ice particles are more concentrated within a narrow altitude range, and a pronounced peak in the number of ice particles occurs around the mesopause at 88 km (Figure 4a). With increasing turbulence (Figures 4b,c), the high-density region broadens, and the ice particles spread over a wider altitude range. For better representation, Figure 5a shows the averaged profiles of the number density of ice particles. It clearly illustrates that the vertical distribution of ice particles becomes wider, and ice particles are more dispersed with increasing turbulence.

Figure 5

Figure 5. Zonal and monthly averaged profiles for July at 69°N, showing the effect of different turbulence scenarios (as indicated in the inset). The panels show: ((a), left) number of ice particles, ((b), middle) ice particle radius, and ((c), right) NLC brightness (backscatter coefficient at 532 nm).

Turbulence moves particles upwards and downwards, which, on average, would not affect the number of particles at a given altitude if the concentration of particles is constant with altitude. However, the concentration peaks around 87 km. As a result, an increase in turbulence causes ice particles to move from regions of higher concentration to lower. This causes the ice particles to spread over a larger vertical range, flattening the peak in number density. The total column density was calculated for different turbulence, and no significant difference was observed. The results indicate that the vertical distribution of the number of ice particles is highly sensitive to turbulence variations at mesopause altitudes. I.e., During the early phase of ice particle growth, vertical motion is dominated by turbulence, leading to a broad distribution of particle altitudes.

Figures 4d–f show time series of the vertical distribution of ice particle radius for different turbulence conditions, and Figure 5b shows zonally and monthly averaged profiles for ice particle radius comparing different turbulence conditions. In all cases, the maximum radius occurs at about 82–83 km altitude. Comparing different turbulence scenarios, an increase in turbulence leads to a decrease in the maximum ice particle radius. Also, the altitude of large-sized particles slightly shifts upward with increasing turbulence. As already explained, most of the ice particles are smaller and form near the coldest regions close to the mesopause (86–88 km) and settle downwards. During sedimentation, the particles are influenced by atmospheric transport mechanisms, including mean vertical and horizontal winds, as well as small-scale motions such as waves and turbulence. Not all particles that form near the mesopause settle and grow to their maximum size. Smaller ice particles sublimate more quickly due to the Kelvin effect. They release their water content as they descend into warmer regions where the saturation ratio decreases. This released vapor becomes available to slightly larger, more stable particles, enabling their continued growth. Therefore, with decreasing altitude, although the saturation ratio S decreases due to rising temperature, the increasing availability of water vapor from small ice particles’ sublimation enhances the growth of ice particles. Ice particles reach their maximum size shortly before sublimation, typically below ∼82 km (sublimation zone), where temperatures rise above the frost point. In particular, the region between 83 and 84 km, just above the sublimation zone, has been identified as the zone of maximum growth rate. More than 80% of the total ice particle growth occurs within this narrow region over a short time period (Berger and von Zahn, 2002). Hence, the net sedimentation speed of ice particles and their residence time within this maximum growth zone are key parameters controlling NLC properties. For example, 50 nm ice particles at ∼83 km settle at ∼10 cm/s, much slower than turbulent velocities (m/s). As discussed, enhanced turbulence near the S = 1 boundary increases particle motion across it, accelerating sublimation, reducing particle size, and shortening their residence time in the region of most efficient growth. This explains why the maximum radius of ice particles decreases with increasing turbulence and why there is a slight upward shift in the altitude of the maximum radius.

To further confirm these findings, we performed numerical experiments with MIMAS in which turbulence was selectively increased over narrow altitude ranges of 1 km and 500 m, while keeping other layers unchanged (not shown in this paper). For instance, turbulence was increased in layers from 80 to 81 km, 81–82 km, and 89–90 km (1 km resolution), and from 80.0 to 80.5 km, 80.5–81.0 km, etc., up to 89.5–90.0 km (500 m resolution). The results clearly showed that the maximum ice particle radius decreases most when turbulence increased between 81.5 and 82.5 km, near the S = 1 boundary. In contrast, when turbulence was increased in the 83–84 km region, within the maximum growth region but above the S = 1 boundary, ice particle radius increased slightly. This can be attributed to the fact that turbulence-induced mixing within the maximum growth zone may slightly increase the residence time of particles (as long as they are not pushed out of the supersaturated region), slightly enhancing their growth. These results demonstrate that the impact of turbulence on ice particle radius strongly depends on the altitude at which turbulence increases. Enhanced turbulence around and below the S = 1 boundary significantly reduces ice particle size, while increased turbulence above the S = 1 boundary and within the maximum growth zone can slightly enhance growth. The overall effect of increased turbulence is a reduction in the maximum radius of ice particles within NLCs.

Figures 4g–i illustrate the time series of NLC brightness under different turbulence conditions. Although the maximum number of ice particles forms at higher altitudes, the maximum bright NLCs’ altitude is around 82–83 km. This is because the radius of the ice particles primarily determines the NLC brightness; therefore, the maximum brightness corresponds to the altitude where the radius is largest, just above the sublimation zone (Kiliani et al., 2013). This is consistent with previous observational studies, which show that brighter NLCs are more strongly associated with larger particle sizes than with higher number densities (Baumgarten and Fiedler, 2008; Rapp and Thomas, 2006). NLC brightness is strongly affected by turbulence variations. Figure 5c shows the averaged profile of NLC brightness, highlighting a significant decrease in brightness with increasing turbulence. Additionally, the altitude of maximum brightness shifts higher, and the brightness distribution broadens due to turbulence-induced dispersion. As mentioned, the sharp decline in NLC brightness (Figure 5c) with increasing turbulence is caused by a reduction in particle radius. NLC backscatter strongly depends on particle radius (r). For particles smaller than ∼30 nm, Rayleigh scattering dominates (β ∝ r⁶), while for particles around 30–50 nm, Mie scattering applies (β ∝ r^4-5), making the backscatter extremely sensitive to changes in particle size (e.g., Hervig et al., 2009). Therefore, even small changes in radius lead to significant changes in optical scattering. Doubling the reference turbulence causes an approximate 7% reduction in the peak mean ice particle radius, leading to a ∼30% decrease in the maximum mean brightness.

3.3 Impact of vertical wind on water vapor

In general, mean vertical winds in the upper mesosphere are weak, typically a few centimeters per second, and exhibit strong variability with latitude, season, and wave activity (Becker, 2012; Chau et al., 2021). At the summer mesosphere (around 80–90 km altitude), the vertical wind is generally upward, driven by gravity wave breaking and the resulting meridional circulation (Lindzen, 1981; Garcia and Solomon, 1985). This upward motion transports water vapor from lower altitudes and plays a crucial role in NLC formation (Berger and von Zahn, 2002). In LIMA, the mean vertical wind velocity in July is positive and in the range of 0–2 cm/s (see Supplementary Figure S1). We investigate how the variations in vertical wind affect water vapor distributions at NLC altitudes in the presence and absence of NLCs. Figure 6 illustrates the mean water vapor profiles under conditions with and without NLCs for different vertical wind conditions. The vertical wind profile was modified by scaling the reference wind by factors of 0.5 (reduced VW) and 2 (enhanced VW). Without NLCs (dashed lines), increasing the vertical wind enhances the upward transport of water vapor, leading to higher concentrations at ice-forming altitudes. This is expected, as stronger upward winds carry more water vapor from lower to higher altitudes. When NLCs are present, the background water vapor distribution becomes more sensitive to changes in vertical wind. An increase in vertical wind not only transports more water vapor upward, but also consequently promotes NLC formation, which intensifies the freeze-drying effect. As a result, water vapor concentrations increase notably below ∼83 km, where ice particles typically sublimate.

Figure 6

Figure 6. Zonal and monthly averaged water vapor profiles for July at 69°N from MIMAS for different vertical winds (VW, see insert). The solid lines show the background water vapor concentration with NLCs, and the dashed lines show the water vapor concentration without NLCs.

3.4 Impact of vertical wind on NLC properties

Vertical winds, influenced by gravity waves, play a crucial role in determining ice particle sedimentation speed and growth. In this section, we investigate the response of NLC properties to varying vertical wind. The profiles of ice particle number density (Figures 7a–c, 8a) show an apparent sensitivity to changes in vertical wind. In contrast to turbulence increase, which broadens the vertical distribution of ice particle numbers and alters their peak value and distribution, an increase in vertical wind primarily causes a slight upward shift in the entire profile without significantly affecting its shape or values (Figure 8a). This vertical shift suggests stronger upward winds push ice particles to higher altitudes. As a result, the altitudes of maximum number density also elevate. These findings support the interpretation that NLC formation is primarily controlled by local supersaturation conditions and the availability of condensation nuclei, whereas vertical winds mainly influence the vertical positioning and residence time (and thereby the size) of the ice particles (Rapp and Thomas, 2006; Berger and von Zahn, 2007).

Figure 7

Figure 7. Time series of NLC properties (number of ice particles, ice particle radius, and brightness) as a function of geometric height, zonally averaged at 69°N for the 2008 NLC season. The results are shown for three runs with different vertical winds (see inset). (a–c) Number of ice particles, (d–f) ice particle radius, and (g–i) NLC brightness.

Figure 8

Figure 8. Zonal and monthly averaged profiles for July at 69°N, showing the effect of different vertical wind scenarios (as indicated in the inset). The panels show: ((a), left) number of ice particles, ((b), middle) mean ice particle radius, and ((c), right) NLC brightness (backscatter coefficient at 532 nm).

The mean radius profiles (Figures 7d–f, 8b) illustrate a clear sensitivity to vertical wind variations. Unlike the effects of turbulence, an increase in vertical wind results in a larger radius of ice particles at all altitudes. The most notable changes occur near the altitude where the average radius peaks. At reduced vertical wind (0.5× reference), the mean particle radius remains below 25 nm. In contrast, under enhanced vertical wind (2× reference), values exceed 35 nm. There are two primary causes for this behavior. First, stronger vertical winds transport more water vapor upward, increasing saturation and lowering the boundary of the saturation zone. Second, the interplay between vertical wind and downward sedimentation alters the effective vertical transport of ice particles. Upward-directed vertical winds reduce the net downward sedimentation velocity, allowing ice particles to remain longer in supersaturated layers. For example, the average vertical wind speeds in the model at 83 km are approximately 2 cm/s, while the typical sedimentation velocity of larger (50 nm) ice particles at the same altitude is approximately 10 cm/s (Kiliani, 2014). Therefore, doubling the upward winds can considerably reduce the net downward speed, thus prolonging the growth period, especially in the maximum growth region (83–84 km). This balance between ice particle vertical transport processes determines the overall ice water content and the microphysical properties of NLCs (Rapp and Thomas, 2006).

The NLC brightness profiles (Figures 7g–i, 8c) show a strong sensitivity to vertical wind variations, with brightness increasing significantly as the vertical winds increase. Additionally, the vertical extent of optically bright NLC layers increases. This enhancement in β is primarily driven by the increase in particle radius. As explained for the turbulence case, the NLC optical brightness is approximately proportional to the 5th to 6th power of the particle radius; even modest reductions in particle size result in a substantial decrease in β. Doubling the vertical wind leads to an approximate 26% increase in the peak mean ice particle radius, resulting in a ∼260% increase in the maximum mean brightness. These results emphasise the key role of mesospheric dynamics in controlling the optical visibility of NLCs. Specifically, strong upwelling, often driven by tides or gravity waves, can lead to brighter, more optically dense, and persistent cloud layers (DeLand et al., 2003; Kiliani et al., 2013).

4 Summary and conclusions

This study investigated how variations in mesospheric turbulence and vertical winds influence the formation and properties of NLCs and the distribution of background water vapor. Using the MIMAS model, driven by mesospheric conditions from LIMA, we performed numerical experiments to isolate the individual effects of turbulence and vertical wind on NLC microphysical properties. The background dynamics and NLC properties calculated by our LIMA/MIMAS setup have been demonstrated to reproduce key climatological features of NLCs, including occurrence, altitude, brightness distribution, and interannual variability, and have been validated in previous studies, showing consistency with other models as well as with observations (e.g., Berger, 2008; Berger and Lübken, 2015; Lübken et al., 2018; Lübken et al., 2021). Specifically, we conducted simulations of a complete summer season in which the turbulent diffusion coefficient and vertical wind speed were systematically reduced and enhanced relative to their reference values in the model. The key findings of this study are summarized as follows.

Without NLCs, turbulence has a negligible effect on the background water vapor profiles. However, when NLCs are present, increased turbulence interferes with the sedimentation of ice particles (particularly near the lower boundary of the supersaturated region (S = 1)), limiting their growth and the associated freeze-drying effect, which in turn alters the distribution of background water vapor. Turbulence strongly influences the vertical distribution of ice particles. Under low turbulence, more ice particles are confined to a narrow altitude range, producing a pronounced peak near the mesopause. With increasing turbulence, vertical mixing disperses particles over a broader altitude range, widening and flattening the peak, while the total column-integrated ice particle number remains largely unchanged. This occurs because turbulence diffusion causes particles to move from regions of higher concentration to lower, redistributing them vertically (Turco et al., 1982). The NLC brightness (backscatter coefficient (β) at 532 nm) is very sensitive to turbulence variations. Ice particles in NLCs reach their maximum size just before sublimation, with the most rapid growth occurring at 83–84 km. Increasing turbulence, particularly around the S = 1 boundary, enhances vertical mixing. This causes frequent crossings of particles between supersaturated and subsaturated regions, leading to a net loss of water vapor due to rapid sublimation, which reduces the maximum growth of ice particles. Because NLC brightness depends strongly on particle radius, this leads to a corresponding decrease in brightness. These findings are consistent with previous modeling studies and lidar observations that report broader particle distributions and reduced brightness during periods of increased turbulence (Baumgarten et al., 2010; Kiliani et al., 2013). The Increase in vertical winds enhances NLCs in two ways. First, they transport more water vapor to NLC-forming altitudes, increasing its availability and creating more favourable conditions for NLC development. Second, upward vertical winds counteract the gravitational settling of ice particles, slowing their downward sedimentation and allowing them to remain longer in supersaturated regions. This extends the growth phase of ice particles, resulting in larger particle sizes and increased NLC brightness.

It should be noted, however, that this approach involves certain simplifications. In the real atmosphere, vertical wind and turbulence are not entirely independent; they are dynamically coupled and can vary non-uniformly with altitude. Scaling these parameters uniformly does not fully capture the spatial and temporal variability seen in observations. Furthermore, in reality, changes in vertical wind can influence the thermal structure of the mesosphere. In this study, our simulations do not account for temperature changes associated with varying vertical wind or turbulence. In the real atmosphere, upward winds, such as those induced by gravity waves, not only transport water vapor to higher altitudes but also lead to adiabatic cooling as rising air parcels expand in lower-pressure regions. This cooling effect enhances supersaturation and further promotes ice particle growth (Jensen et al., 1989). Since this mechanism is not accounted for in our simulations, the observed changes in NLC properties with increasing vertical wind are solely attributed to increased water vapor availability and the reduced net sedimentation velocity of ice particles. If adiabatic cooling were included, the resulting temperature decrease would likely enhance NLC formation even further. Therefore, our current findings probably underestimate the impact of increased vertical winds. The current version of MIMAS includes only heterogeneous nucleation on meteoric smoke particles; homogeneous nucleation is not considered because it is expected to play only a minor role under typical polar mesopause conditions (Murray and Jensen, 2010; Tanaka et al., 2022). Despite these limitations, this study aims to isolate and understand the first-order effects of vertical transport on NLC formation, assuming that all other factors remain unchanged.

Our results demonstrate that small-scale turbulence and vertical wind have a significant impact on the formation and evolution of noctilucent clouds. Laboratory studies (Duft et al., 2015; Sato et al., 2025) have shown that ice nucleation and early growth are highly sensitive to supersaturation and nanoscale processes, which aligns well with our finding that turbulence and vertical wind affect the duration of particles in the growth region. Feofilov and Petelina (2010) found that NLC brightness and occurrence depend on temperature and water vapor. In our simulations, turbulence and vertical wind influence these factors by altering the distribution of water vapor and ice particles. Wilms et al. (2016) showed with model simulations that gravity-wave variability limits the time available for particle growth, and our results indicate that turbulence and vertical wind can produce a similar effect.

A comparison was carried out between MIMAS water vapor profiles during the NLC season and SOFIE satellite data (not shown). This analysis indicates that the model tends to overestimate water vapor depletion at ice-forming altitudes. Possible reasons for this discrepancy include limitations in representing background dynamics, particularly gravity wave activity, since slight variations in turbulence or vertical wind can strongly impact water vapor and NLCs. The observed wave-like patterns in NLCs likely reflect these underlying wind patterns and turbulence. A detailed investigation of these mechanisms highlights the potential of using NLCs as effective tracers for studying gravity waves and atmospheric dynamics in the upper mesosphere. From the response to the increased turbulence and vertical wind cases, we conclude that (including unexpected responses of NLC), improved background dynamics is only part of the solution. Currently, MIMAS includes only water vapor transport and photochemistry. Future work aims to improve the model by incorporating more realistic background dynamics (e.g., from Upper Atmosphere–ICOsahedral Nonhydrostatic model (UA-ICON)) and extending the water vapor chemistry.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. The permanent DOI for the dataset is: https://dx.doi.org/10.22000/gss0g6yxt7ydx3xz.

Author contributions

AV: Writing – original draft, Formal Analysis, Supervision, Writing – review and editing, Investigation, Visualization, Conceptualization, Methodology. GB: Writing – review and editing, Supervision, Funding acquisition, Resources. MG: Writing – review and editing, Supervision. F-JL: Supervision, Writing – review and editing. GU: Writing – review and editing, Writing – original draft, Formal Analysis, Conceptualization, Visualization, Investigation.

Funding

The author(s) declared that financial support was received for this work and/or its publication. This research is supported by the project W1 (Gravity Wave Parameterization for the Atmosphere) of the Collaborative Research Centre TRR 181 “Energy Transfers in Atmosphere and Ocean” funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) - Projektnummer 274762653.

Acknowledgements

We thank the Deutsche Forschungsgemeinschaft (DFG) for funding support and the IT department at IAP for their assistance with computational resources.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fspas.2026.1721062/full#supplementary-material

References

Baumgarten, G., and Fiedler, J. (2008). Vertical structure of particle properties and water content in noctilucent clouds. Geophys. Res. Lett. 35, L10811. doi:10.1029/2007GL033084

CrossRef Full Text | Google Scholar

Baumgarten, G., and Fritts, D. C. (2014). Quantifying Kelvin-Helmholtz instability dynamics observed in noctilucent clouds: 1. Methods and observations. J. Geophys. Res. Atmos. 119 (15), 9324–9337. doi:10.1002/2014JD021832

CrossRef Full Text | Google Scholar

Baumgarten, G., Fiedler, J., Lübken, F., and von Cossart, G. (2008). Particle properties and water content of noctilucent clouds and their interannual variation. J. Geophys. Res. Atmos. 113 (6). doi:10.1029/2007JD008884

CrossRef Full Text | Google Scholar

Baumgarten, G., Fiedler, J., and Rapp, M. (2010). On microphysical processes of noctilucent clouds (NLC): observations and modeling of mean and width of the particle size-distribution. Atmos. Chem. Phys. 10 (14), 6661–6668. doi:10.5194/acp-10-6661-2010

CrossRef Full Text | Google Scholar

Becker, E. (2012). Dynamical control of the middle atmosphere. Space Sci. Rev. 168 (1–4), 283–314. doi:10.1007/s11214-011-9841-5

CrossRef Full Text | Google Scholar

Berger, U. (2008). Modeling of middle atmosphere dynamics with LIMA. J. Atmos. Solar-Terrestrial Phys. 70 (8–9), 1170–1200. doi:10.1016/j.jastp.2008.02.004

CrossRef Full Text | Google Scholar

Berger, U., and Lübken, F.-J. (2006). Weather in mesospheric ice layers. Geophys. Res. Lett. 33 (4). doi:10.1029/2005GL024841

CrossRef Full Text | Google Scholar

Berger, U., and Lübken, F. J. (2011). Mesospheric temperature trends at mid-latitudes in summer. Geophys. Res. Lett. 38 (22). doi:10.1029/2011GL049528

CrossRef Full Text | Google Scholar

Berger, U., and Lübken, F. J. (2015). Trends in mesospheric ice layers in the Northern Hemisphere during 1961-2013. J. Geophys. Res. 120 (21). doi:10.1002/2015JD023355

CrossRef Full Text | Google Scholar

Berger, U., and von Zahn, U. (2002). Icy particles in the summer mesopause region: three-dimensional modeling of their environment and two-dimensional modeling of their transport. J. Geophys. Res. Space Phys. 107 (A11). doi:10.1029/2001JA000316

CrossRef Full Text | Google Scholar

Berger, U., and von Zahn, U. (2007). Three-dimensional modeling of the trajectories of visible noctilucent cloud particles: an indication of particle nucleation well below the mesopause. J. Geophys. Res. Atmos. 112 (16). doi:10.1029/2006JD008106

CrossRef Full Text | Google Scholar

Brasseur, G. P., and Solomon, S. (2005). “Chemical concepts in the atmosphere,” in Aeronomy of the middle atmosphere: chemistry and physics of the stratosphere and mesosphere (Dordrecht: Springer Netherlands), 11–50. doi:10.1007/1-4020-3824-0_2

CrossRef Full Text | Google Scholar

Chau, J. L., Marino, R., Feraco, F., Urco, J. M., Baumgarten, G., Lübken, F., et al. (2021). Radar observation of extreme vertical drafts in the polar summer mesosphere. Geophys. Res. Lett. 48 (16), e2021GL094918. doi:10.1029/2021GL094918

CrossRef Full Text | Google Scholar

Dalin, P., Gavrilov, N., Pertsev, N., Perminov, V., Pogoreltsev, A., Shevchuk, N., et al. (2016). A case study of long gravity wave crests in noctilucent clouds and their origin in the upper tropospheric jet stream. J. Geophys. Res. Atmos. 121 (23), 14102–14118. doi:10.1002/2016JD025422

CrossRef Full Text | Google Scholar

DeLand, M. T., Shettle, E. P., Thomas, G. E., and Olivero, J. J. (2003). Solar backscattered ultraviolet (SBUV) observations of polar mesospheric clouds (PMCs) over two solar cycles. J. Geophys. Res. Atmos. 108 (8). doi:10.1029/2002jd002398

CrossRef Full Text | Google Scholar

Duft, D., Nachbar, M., Eritt, M., and Leisner, T. (2015). A linear trap for studying the interaction of nanoparticles with supersaturated vapors. Aerosol Sci. Technol. 49 (9), 683–691. doi:10.1080/02786826.2015.1063583

CrossRef Full Text | Google Scholar

Espy, P. J., and Jutt, H. (2002). Equilibrium temperature of water-ice aerosols in the high-latitude summer mesosphere. J. Atmos. Solar-Terrestrial Phys. 64 (17), 1823–1832. doi:10.1016/S1364-6826(02)00191-8

CrossRef Full Text | Google Scholar

Feofilov, A. G., and Petelina, S. v. (2010). Relation between mesospheric ice clouds, temperature, and water vapor determined from Odin/OSIRIS and TIMED/SABER data. J. Geophys. Res. Atmos. 115 (18). doi:10.1029/2009JD013619

CrossRef Full Text | Google Scholar

Fiedler, J., Baumgarten, G., and von Cossart, G. (2003). Noctilucent clouds above ALOMAR between 1997 and 2001: occurrence and properties. J. Geophys. Res. Atmos. 108 (8). doi:10.1029/2002jd002419

CrossRef Full Text | Google Scholar

Fritts, D. C., and Alexander, M. J. (2003). Gravity wave dynamics and effects in the middle atmosphere. Rev. Geophys. 41 (1). doi:10.1029/2001RG000106

CrossRef Full Text | Google Scholar

Fritts, D. C., Isler, J. R., Thomas, G. E., and Andreassen, Ø. (1993). Wave breaking signatures in noctilucent clouds. Geophys. Res. Lett. 20 (19), 2039–2042. doi:10.1029/93GL01982

CrossRef Full Text | Google Scholar

Fritts, D. C., Wang, L., Lund, T. S., Thorpe, S., Kjellstrand, C. B., Kaifler, B., et al. (2022). Multi-scale Kelvin-Helmholtz instability dynamics observed by PMC Turbo on 12 July 2018: 2. DNS modeling of KHI dynamics and PMC responses. J. Geophys. Res. Atmos. 127 (18), e2021JD035834. doi:10.1029/2021JD035834

CrossRef Full Text | Google Scholar

Gadsden, M. (1982). Noctilucent clouds. Space Sci. Rev. 33 (3–4). doi:10.1007/BF00196999

CrossRef Full Text | Google Scholar

Gadsden, M., and Schröder, W. (1989). “Noctilucent clouds,” in Physics and chemistry in space (Berlin, Heidelberg: Springer-Verlag), 18. doi:10.1007/978-3-642-48626-5

CrossRef Full Text | Google Scholar

Garcia, R. R., and Solomon, S. (1985). The effect of breaking gravity waves on the dynamics and chemical composition of the mesosphere and lower thermosphere. J. Geophys. Res. 90 (D2), 3850–3868. doi:10.1029/JD090iD02p03850

CrossRef Full Text | Google Scholar

Gumbel, J., Siskind, D. E., Witt, G., Torkar, K. M., and Friedrich, M. (2003). Influences of ice particles on the ion chemistry of the polar summer mesosphere. J. Geophys. Res. Atmos. 108 (8). doi:10.1029/2002jd002413

CrossRef Full Text | Google Scholar

Hervig, M. E., Stevens, M. H., Gordley, L. L., Deaver, L. E., Russell, J. M., III, and Bailey, S. M. (2009). Relationships between polar mesospheric clouds, temperature, and water vapor from solar occultation for ice experiment (SOFIE) observations. J. Geophys. Res. Atmos. 114 (20). doi:10.1029/2009JD012302

CrossRef Full Text | Google Scholar

Hervig, M. E., Berger, U., and Siskind, D. E. (2016). Decadal variability in PMCs and implications for changing temperature and water vapor in the upper mesosphere. J. Geophys. Res. Atmos. 121 (5), 2383–2392. doi:10.1002/2015JD024439

CrossRef Full Text | Google Scholar

Hoffmann, P., Becker, E., Singer, W., and Placke, M. (2010). Seasonal variation of mesospheric waves at northern middle and high latitudes. J. Atmos. Solar-Terrestrial Phys. 72 (14–15), 1068–1079. doi:10.1016/j.jastp.2010.07.002

CrossRef Full Text | Google Scholar

Hunten, D. M., Turco, R. P., and Toon, O. B. (1980). Smoke and dust particles of meteoric origin in the mesosphere and stratosphere. J. Atmos. Sci. 37 (6), 1342–1357. doi:10.1175/1520-0469(1980)037<1342:SADPOM>2.0.CO;2

CrossRef Full Text | Google Scholar

Jacobi, C., Karami, K., Kuchar, A., Ern, M., Renkwitz, T., Latteck, R., et al. (2025). Long-term changes and trends of mesosphere/lower thermosphere gravity waves over Collm, Germany. Adv. Radio Sci. 23, 21–31. doi:10.5194/ars-23-21-2025

CrossRef Full Text | Google Scholar

Jensen, E. J., Thomas, G. E., and Toon, O. B. (1989). On the diurnal variation of noctilucent clouds. J. Geophys. Res. 94 (D8), 10769–10780. doi:10.1029/JD094iD08p10769

CrossRef Full Text | Google Scholar

Kiliani, J. (2014). 3-D modeling of noctilucent cloud evolution and relationship to the ambient atmosphere. Kühlungsborn: University of Rostock. Doctoral dissertation.

Google Scholar

Kiliani, J., Baumgarten, G., Lübken, F. J., Berger, U., and Hoffmann, P. (2013). Temporal and spatial characteristics of the formation of strong noctilucent clouds. J. Atmos. Solar-Terrestrial Phys. 104, 151–166. doi:10.1016/j.jastp.2013.01.005

CrossRef Full Text | Google Scholar

Kjellstrand, C. B., Fritts, D. C., Miller, A. D., Williams, B. P., Kaifler, N., Geach, C., et al. (2022). Multi-scale Kelvin-Helmholtz instability dynamics observed by PMC Turbo on 12 July 2018: 1. Secondary instabilities and billow interactions. J. Geophys. Res. Atmos. 127 (18), e2021JD036232. doi:10.1029/2021JD036232

CrossRef Full Text | Google Scholar

Körner, U., and Sonnemann, G. R. (2001). Global three-dimensional modeling of the water vapor concentration of the mesosphere-mesopause region and implications with respect to the noctilucent cloud region. J. Geophys. Res. Atmos. 106 (D9). doi:10.1029/2000JD900744

CrossRef Full Text | Google Scholar

Lindzen, R. S. (1981). Turbulence and stress owing to gravity wave and tidal breakdown. J. Geophys. Res. Oceans 86 (C10), 9707–9714. doi:10.1029/jc086ic10p09707

CrossRef Full Text | Google Scholar

Lübken, F. J. (1997). Seasonal variation of turbulent energy dissipation rates at high latitudes as determined by in situ measurements of neutral density fluctuations. J. Geophys. Res. Atmos. 102 (D12). doi:10.1029/97JD00853

CrossRef Full Text | Google Scholar

Lübken, F. J. (1999). Thermal structure of the arctic summer mesosphere. J. Geophys. Res. Atmos. 104 (D8), 9135–9149. doi:10.1029/1999JD900076

CrossRef Full Text | Google Scholar

Lübken, F. J., and Berger, U. (2011). Latitudinal and interhemispheric variation of stratospheric effects on mesospheric ice layer trends. J. Geophys. Res. Atmos. 116 (D4). doi:10.1029/2010JD015258

CrossRef Full Text | Google Scholar

Lübken, F. J., Baumgarten, G., Fiedler, J., Gerding, M., Höffner, J., and Berger, U. (2008). Seasonal and latitudinal variation of noctilucent cloud altitudes. Geophys. Res. Lett. 35 (6). doi:10.1029/2007GL032281

CrossRef Full Text | Google Scholar

Lübken, F. J., Berger, U., and Baumgarten, G. (2009). Stratospheric and solar cycle effects on long-term variability of mesospheric ice clouds. J. Geophys. Res. Atmos. 114 (21). doi:10.1029/2009JD012377

CrossRef Full Text | Google Scholar

Lübken, F. J., Berger, U., and Baumgarten, G. (2013). Temperature trends in the midlatitude summer mesosphere. J. Geophys. Res. Atmos. 118 (24). doi:10.1002/2013JD020576

CrossRef Full Text | Google Scholar

Lübken, F. J., Berger, U., and Baumgarten, G. (2018). On the anthropogenic impact on long-term evolution of noctilucent clouds. Geophys. Res. Lett. 45 (13). doi:10.1029/2018GL077719

CrossRef Full Text | Google Scholar

Lübken, F. J., Baumgarten, G., and Berger, U. (2021). Long term trends of mesopheric ice layers: a model study. J. Atmos. Solar-Terrestrial Phys. 214. doi:10.1016/j.jastp.2020.105378

CrossRef Full Text | Google Scholar

Murray, B. J., and Jensen, E. J. (2010). Homogeneous nucleation of amorphous solid water particles in the upper mesosphere. J. Atmos. Solar-Terrestrial Phys. 72 (1), 51–61. doi:10.1016/j.jastp.2009.10.007

CrossRef Full Text | Google Scholar

Rapp, M., and Thomas, G. E. (2006). Modeling the microphysics of mesospheric ice particles: assessment of current capabilities and basic sensitivities. J. Atmos. Solar-Terrestrial Phys. 68, 715–744. doi:10.1016/j.jastp.2005.10.015

CrossRef Full Text | Google Scholar

Russell, J. M., Rong, P., Hervig, M. E., Siskind, D. E., Stevens, M. H., Bailey, S. M., et al. (2014). Analysis of northern midlatitude noctilucent cloud occurrences using satellite data and modeling. J. Geophys. Res. Atmos. 119 (6), 3238–3250. doi:10.1002/2013JD021017

CrossRef Full Text | Google Scholar

Sato, R., Noguchi, K., Koshida, H., Ishibashi, A., and Hama, T. (2025). Three-step growth of vapor-deposited ice under mesospheric temperature and water vapor conditions. J. Phys. Chem. Lett. 16, 7675–7684. doi:10.1021/acs.jpclett.5c01536

PubMed Abstract | CrossRef Full Text | Google Scholar

Sugiyama, T. (1994). Ion-recombination nucleation and growth of ice particles in noctilucent clouds. J. Geophys. Res. Space Phys. 99 (3), 3915–3929. doi:10.1029/93JA02822

CrossRef Full Text | Google Scholar

Tanaka, K. K., Mann, I., and Kimura, Y. (2022). Formation of ice particles through nucleation in the mesosphere. Atmos. Chem. Phys. 22 (8), 5639–5650. doi:10.5194/acp-22-5639-2022

CrossRef Full Text | Google Scholar

Thomas, G. E. (1984). Solar mesosphere explorer measurements of polar mesospheric clouds (noctilucent clouds). J. Atmos. Terr. Phys. 46 (9), 819–824. doi:10.1016/0021-9169(84)90062-X

CrossRef Full Text | Google Scholar

Thomas, G. E., and Olivero, J. J. (1989). Climatology of polar mesospheric clouds. 2. Further analysis of solar mesosphere explorer data. J. Geophys. Res. 94 (D12), 14673–14681. doi:10.1029/jd094id12p14673

CrossRef Full Text | Google Scholar

Turco, R. P., Toon, O., Whitten, R., Keesee, R., and Hollenbach, D. (1982). Noctilucent clouds: simulation studies of their genesis, properties and global influences. Planet. Space Sci. 30 (11), 1147–1181. doi:10.1016/0032-0633(82)90126-X

CrossRef Full Text | Google Scholar

Vellalassery, A., Baumgarten, G., Grygalashvyly, M., and Lübken, F. J. (2023). Greenhouse gas effects on the solar cycle response of water vapour and noctilucent clouds. Ann. Geophys. 41 (2), 289–300. doi:10.5194/angeo-41-289-2023

CrossRef Full Text | Google Scholar

Vellalassery, A., Baumgarten, G., Grygalashvyly, M., and Lübken, F. J. (2024). Long-term evolution in noctilucent clouds’ response to the solar cycle: a model-based study. Atmosphere 15 (1), 88. doi:10.3390/atmos15010088

CrossRef Full Text | Google Scholar

von Cossart, G., Fiedler, J., and von Zahn, U. (1999). Size distributions of NLC particles as determined from 3-color observations of NLC by ground-based lidar. Geophys. Res. Lett. 26 (11), 1513–1516. doi:10.1029/1999GL900226

CrossRef Full Text | Google Scholar

von Zahn, U., and Berger, U. (2003). Persistent ice cloud in the midsummer upper mesosphere at high latitudes: three-dimensional modeling and cloud interactions with ambient water vapor. J. Geophys. Res. Atmos. 108 (8). doi:10.1029/2002jd002409

CrossRef Full Text | Google Scholar

Wilms, H., Rapp, M., and Kirsch, A. (2016). Nucleation of mesospheric cloud particles: sensitivities and limits. J. Geophys. Res. Space Phys. 121, 2621–2644. doi:10.1002/2015JA021764

CrossRef Full Text | Google Scholar

Witt, G. (1962). Height, structure and displacements of noctilucent clouds. Tellus A Dyn. Meteorology Oceanogr. 14 (1), 1–18. doi:10.3402/tellusa.v14i1.9524

CrossRef Full Text | Google Scholar

Zhang, L., Tinsley, B., and Zhou, L. (2022). Responses of CIPS/AIM noctilucent clouds to the interplanetary magnetic field. Atmos. Chem. Phys. 22 (20), 13355–13370. doi:10.5194/acp-22-13355-2022

CrossRef Full Text | Google Scholar