Next Article in Journal
A Prompt Decarbonization Pathway for Shipping: Green Hydrogen, Ammonia, and Methanol Production and Utilization in Marine Engines
Next Article in Special Issue
Peculiarities of the Vertical Structure of Atmospheric Aerosol Fields in the Basin of Lake Baikal According to Lidar Observations
Previous Article in Journal
Application of Radar Radial Velocity Data Assimilation in the Forecasts of Typhoon Linfa Based on Different Horizontal Length Scale Factors
Previous Article in Special Issue
Coherent Backscattering by Large Ice Crystals of Irregular Shapes in Cirrus Clouds
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Atmosphere
Volume 14
Issue 3
10.3390/atmos14030583
atmosphere-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

A Study on the Wide Range of Relative Humidity in Cirrus Clouds Using Large-Ensemble Parcel Model Simulations

by
Miao Zhao
and
Xiangjun Shi
*
School of Atmospheric Sciences, Nanjing University of Information Science and Technology, Nanjing 210044, China
*
Author to whom correspondence should be addressed.
Atmosphere 2023, 14(3), 583; https://doi.org/10.3390/atmos14030583
Submission received: 7 February 2023 / Revised: 3 March 2023 / Accepted: 14 March 2023 / Published: 17 March 2023
(This article belongs to the Special Issue Properties of Cirrus Cloud by Lidars: Observation and Theory)
Browse Figures
Review Reports Versions Notes

Abstract

:
This study investigates the possible mechanisms related to the wide range of relative humidity in cirrus clouds (RHi). Under the closed adiabatic assumption, the impacts of vertical motion and ice crystal deposition/sublimation on RHi are investigated through in situ observations and parcel model simulations. Vertical motion is an active external force that changes the RHi, and ice crystal deposition/sublimation plays a role in mitigating the change in the RHi. They are the two most important mechanisms involved in controlling the RHi fluctuation during cirrus evolution and could well explain the wide range of RHi in wave-related cirrus clouds. Furthermore, a comparison of statistical cloud characteristics from both observations and simulations shows that a very low value (e.g., 0.001) for the water vapor ice deposition coefficient is highly unlikely.

1. Introduction

Cirrus clouds (including thin cirrus) cover about 30% of the Earth’s surface and significantly impact the transport of water vapor across the tropopause and the radiative transfer in the upper troposphere [1,2,3,4,5,6,7,8,9]. Unlike in warm clouds, the number density (Ni) of ice crystals (ICs) in cirrus clouds is relatively low (usually much less than 1000 L−1), and the in-cloud relative humidity over ice (RHi) has a wide range (50% to 150% in most cases) [5,10,11,12,13,14]. The in-cloud RHi dominates the deposition/sublimation process of ICs and plays an important role in cloud evolution [15,16,17]. Although some observational studies have provided some possible reasons for the wide range of in-cloud RHi, there are very few studies that quantitatively investigate the main mechanisms related to RHi [18,19,20,21,22,23,24].
The main influencing mechanisms of in-cloud RHi can be divided into two categories: mechanisms related to exchanges with the surrounding ambiance and mechanisms unassociated with the surrounding ambiance. The entrainment and mixing process on cloud edges can affect the in-cloud water through a mass exchange [25,26,27]. It is noteworthy that the entrainment and mixing process remains poorly understood and that its measurement and quantification are extremely difficult [28,29,30,31]. Thus, this study only focuses on the mechanisms unassociated with the surrounding ambiance. With the same water vapor pressure, the change in air temperature (T) caused by vertical motion could affect the RHi because the saturated water vapor pressure changes sharply with T [32,33,34,35]. In cirrus clouds, the IC deposition/sublimation process occurs as long as the water vapor is saturated/unsaturated [32]. The IC deposition/sublimation directly changes the in-cloud RHi via the changing water vapor. Meanwhile, the in-cloud RHi also determines the efficiency of the ICs to scavenge/release water vapor [16,36]. Taken overall, under conditions where neither a thermal exchange nor a mass exchange with the surrounding ambiance is possible (i.e., under the closed adiabatic assumption), the change in in-cloud RHi is mainly determined by vertical velocity (W) and IC deposition/sublimation. The main purpose of this study is to investigate how these two mechanisms impact RHi fluctuations and to determine whether the wide range of RHi in wave-related cirrus clouds (this kind of cirrus is relatively more consistent with the closed adiabatic assumption) can be explained by these two mechanisms. Note that neither in-site observations nor the cloud microphysics classical theory can support this study alone.
The cloud parcel model is an important tool for analyzing cloud evolution [13,37,38,39]. Furthermore, field campaigns with aircraft measurements have provided numerous in situ observations of in-cloud RHi, the ICs’ physical characteristics, and the basic atmospheric state [24,40]. Based on these observations and a few artificial settings, the impact of vertical motion and IC deposition/sublimation on in-cloud RHi can be analyzed with large-ensemble parcel model simulations. In this study, a closed adiabatic cloud parcel model was first introduced and then used to carry out large-ensemble simulations. By taking advantage of the model simulations, the contributions of microphysical (i.e., IC deposition/sublimation) and dynamical (i.e., vertical motion) processes could also be separately investigated. This paper is organized as follows: Section 2 introduces the in situ observations and cloud parcel model, and the experimental setup is described. The parcel model simulation results are analyzed in Section 3. Section 4 discusses the limitations of this study. Finally, conclusions are provided in Section 5.

2. Data and Methods

2.1. In Situ Observations

Multiple field campaigns have provided the microphysical properties of cirrus clouds [24,40]. The preferred choice is that of cirrus clouds which are consistent with the closed adiabatic assumption. Cirrus clouds generally have two basic generating mechanisms, namely the uplift of moist air by wave motions on a variety of scales and the deep convective outflow [13,17]. The cirrus from the deep convective outflow grows thinner at a rapid pace as convection ceases, with IC sedimentation being an important process in their life cycle [41]. It is clear that this kind of cirrus is not suited for this study. In this study, in situ observations from the NASA Mid-latitude Airborne Cirrus Properties Experiment (MACPEX) were used because the cirrus observed during the MACPEX were mostly wave-related. The MACPEX mission was an intensive, month-long (March–April 2011) campaign over the South-Central U.S., and it provided nine flights with sufficient science-quality data [42,43]. Excluding the flights that targeted the anvil cirrus (i.e., deep convective outflow), seven flights were selected for the analysis of cirrus microphysical properties and the initial conditions for parcel model simulations.
The Meteorological Measurement System (MMS; [44]) provided the basic atmospheric states needed by this study, which included the W, T, and pressure (P). The water vapor was measured with a Harvard Water Vapor instrument (HWV; [45]). RHi was calculated from the measured water vapor pressure using the saturation water vapor pressure formulas of Murphy and Koop [46]. A Two-Dimensional Stereo probe (2D-S; [47]) provided the Ni, ice water content (qi), and IC size distribution (the lower limit was 5 μm). The above-mentioned in situ observations were sampled at 1 Hz. Only the cirrus samples (Ni > 1 L−1 and T < 233 K) that were far from the cloud edges were analyzed in this study. The number of total samples was 19,679.

2.2. Cloud Parcel Model and Experimental Setups

A closed adiabatic cloud parcel model driven by a prescribed W time series was used in this study. The P lapse rate of the surrounding ambiance can be diagnosed based on the standard atmosphere. The vertical motion will affect the P of the air parcel, then T, and then RHi. The IC deposition/sublimation is described by the trend of the IC size (Ri, spherical assumption). Equations that describe the evolution of T, P, and Ri can be found in the textbook by Pruppacher and Klett [32]. There are 50 IC size bins; in this study, Ni and qi are the sums of each IC’s size bin. RHi was determined using the conservation equation of total water. The IC sublimation was regarded as the reverse process of deposition. It is noteworthy that the impacts of vertical motion and IC deposition/sublimation on RHi were well represented in the cloud parcel model because the relevant theoretical formulas are simple and robust. The parcel model was just used to investigate how these two mechanisms impact RHi fluctuations based on observed ICs, which would be used for the initial state of parcel model simulations. Besides the above mechanisms, only the ice nucleation process was added to the parcel model because the initial ice crystals might vanish in the sublimation phase during the simulation. The ice nucleation process included the heterogeneous freezing of ice nucleating particles (INPs) and the homogeneous nucleation of a supercooled aerosol solution. The residuals of the vanished ICs (excluding ICs from homogeneous freezing) would then 100% act as INPs during the subsequent simulation. The main purpose of this setup was to let the simulated Ni have a statistical characteristic similar to that of the initial Ni (i.e., observed Ni). The purpose of adding homogeneous freezing was to prevent the parcel model’s simulated RHi from unreasonably high values (much larger than the homogeneous threshold, 150~160%). Following the simulation setup of Krämer et al. [39], the concentration of sulfate aerosol particles (Na) (i.e., supercooled solution aerosol particles) used in the parcel model was assumed to be 300 cm−3. It is necessary to point out that large-ensemble simulation results are not sensitive to Na because the concentration of newly formed ICs from homogeneous freezing (Nihom) is not very sensitive to Na as long as the Na is much larger than the Nihom [38]. The model time step was set to 10 s. Correspondingly, the sample interval of the W time series used to drive the parcel model was 10 s.
In this study, the effects of vertical motion and IC deposition/sublimation were investigated using large-ensemble simulations, which could provide the statistical characteristics of cirrus clouds for a comparison with the observations. Each simulation ran for 6 h, which is around the lower limit of cirrus lifetime statistics (from several hours to more than 24 h [48]). The first-hour simulation was taken as the model spin-up, and the subsequent simulation results were used in our analysis. Because the main purpose of this study is to determine whether the wide range of RHi can be explained by the above two mechanisms, the observed data (especially IC and W) were used to constrain parcel model simulations. The initial state of the parcel model simulation was set by the observed T, P, RHi, and IC size distribution (including the values of Ni and qi). Because the IC size bin in the parcel model (tunable setting) is usually different from that in the observation, the original data from the observed IC size distribution bin were first smoothed with the gamma size distribution before being used for the model’s initial state. The observed W could not be directly used to drive the parcel model simulation because flight measurements were sampled from different places (~200 m between two adjacent samples). In other words, the observed W cannot provide wave frequency information. Therefore, only the amplitude of the W time series (i.e., vertical wind speed) used for driving the parcel model simulations came from observations. How to produce a reasonable W time series is carefully illustrated in the next paragraph.
During the parcel model simulation, the temperature perturbations (ΔTs, changes as compared to the initial value) were mostly determined based on the vertical distance of the cloud parcel from its initial position (ΔH, the time integration of W). Thus, the effects of vertical motion did not only depend on the wave amplitude but also on wave frequency. Here, the speed of vertical motion (i.e., wave amplitude) that was used to drive the parcel model was obtained from the observed W because the statistical distribution characteristic of the vertical wind speed at a local place had to be similar to that taken from flight measurements. The raw spectrum of vertical motion was generated by randomly selecting the W from the observation. In order to avoid producing a very weak ΔT time series, high-frequency turbulent waves (periods less than a few minutes) had to be inhibited to some extent. For this purpose, the value of the randomly selected W could not be far from the previous one (the difference was less than 0.5 m s−1). To finally allow the cloud parcel to return to its initial position, the average W (very small) was removed. The ΔH was limited to −500~500 m in order to prevent an unusually strong ΔT (beyond the range of −5~5 K; the dry adiabatic lapse rate was at ~1 K/100 m). Furthermore, to allow the ΔT time series to have obvious, alternating positive and negative variations during the 6-h simulation, a 6-h ΔH time series was required to have two peaks over 200 m and two valleys below −200 m. In this study, a total of 50 W time series (6-h series) were selected for the large-ensemble simulations. The statistical characteristics of these 50 W time series are shown in Figure 1. As expected, the occurrence frequency of W is very close to that in the observations. The ΔT is mostly dependent on the ΔH (i.e., time integration of W) under the closed adiabatic assumption [32]. In terms of statistical characteristics, the wave-driven ΔT is ~4 times that of the W. This relation between the ΔT and W is in accordance with the approximated formula (δW = 0.23 δT; δ indicates a standard deviation) based on observed fluctuations on the mesoscale [36], which has been widely used in model simulation studies (e.g., [8,37,49,50]). The waves with periods of 20~200 min are dominant. This is generally in agreement with the vertical winds measured with the Doppler lidar at one mid-latitude place, which is dominated by mesoscale gravity waves (20~150 min, [51]). Figure 1 also shows two of the 50 W time series. The high-frequency fluctuations (period of a few minutes) due to turbulence are obvious. This is consistent with the vertical air motion retrieved from ground radar observations [38,52]. In short, the statistical characteristics of W and the corresponding ΔT are generally in agreement with related background knowledge, and so the prescribed W time series used for large-ensemble simulations are reasonable.
In this study, one thousand observed samples were used to derive the initial conditions needed for the large-ensemble simulation (RHi, T, P, and ICs’ variables). The probability distribution characteristics of these samples are consistent with those of the total samples (shown in Section 3.2). With each initial condition, the parcel model was run 50 times, driven by the 50 W time series provided. Therefore, each experiment had 50,000 simulations (i.e., large-ensemble simulations). Here, one reference experiment and six sensitivity experiments were used to analyze the effects of vertical motion and IC deposition/sublimation (Table 1). The results of the REF experiment were compared with the observations and then investigated as to whether the statistical characteristics of the observations could be reproduced by parcel model simulations. Note that the water vapor ice deposition coefficient (αd), which determines the ability of ICs to scavenge/release water vapor, is full of uncertainty [53,54]. Gierens et al. [53] suggested that the αd might be much smaller than a unit (i.e., 1). Thus, the αd was set to 0.05 in the REF experiment. The Wamp, Wfre, and Wno experiments focused on investigating the effects of vertical motion. The Wamp experiment simulated cirrus evolution with a weaker vertical motion (where the amplitude was reduced by half). The Wfre experiment was used to test the impact of wave frequency. In order to double the wave frequency used in the Wfre experiment, the interval of the given W time series was set to 5 s rather than the initial 10 s. Correspondingly, each simulation was only run for 3 h. The Wno experiment simulated cirrus evolution without vertical motion. The effects of IC deposition/sublimation were investigated using the ICnosub, ICadL, and ICadH experiments. The comparison between the ICnosub and REF experiments was used to analyze the effect of the IC sublimation. The ICadL and ICadH experiments explored sensitivities to the αd. Because the efficiency of ICs in scavenging/releasing water vapor during the deposition/sublimation process depends on the value of the αd, the simulation results from the ICadL experiment (the αd is very small) can be interpreted as cirrus evolution with a very weak impact from IC deposition/sublimation.

3. Results

3.1. Case Studies

To better understand the cirrus statistical characteristics of large-ensemble simulations, the impacts of the vertical motion and microphysical processes on cirrus evolution are first illustrated with the use of some case simulations. The first W time series, shown in Figure 1 (Wave1), was used to drive seven simulations, which correspond to the seven experiments listed in Table 1. All these seven simulations had the same initial conditions: RHi = 110%, T = 225 K, P = 250 hPa, Ni = 100 L−1, the IC mass-weighted average radius (Riq)= 15 μm, and the shape parameter of the gamma IC size distribution (μi) was 5. Besides the seven simulations, the REF experiment was also conducted again, with initial conditions of Ni = 10 L−1 and Riq = 30 μm and was driven by another W time series (Wave2 in Figure 1). Thus, there were a total of eight case simulations.
Figure 2 shows the results of these case simulations. At the beginning of the REF (Wave1) simulation (0~0.2 h), all ICs grew because the RHi > 100%. As a result, the qi increased, and the RHi decreased. Thereafter, the ice parcel went down (0.3~1 h; see the ΔH in Figure 1). The RHi dropped to below 100%, mainly due to the positive ΔT driven by the ΔH (a negative correlation between the ΔT and ΔH). Meanwhile, the ICs began to sublimate. All ICs became smaller, and the small ICs vanished. Correspondingly, Ni dropped from 100 L−1 to 3 L−1. At about 1.8 h, the increasing RHi reached the heterogeneous freezing threshold (120~130%). The heterogeneous freezing occurred with the aid of the INPs, which came from the sublimated ICs. Correspondingly, Ni went back to 100 L−1. In the subsequent simulation process, Ni hardly changed. The size of each IC increased or decreased according to the RHi fluctuations, which were mainly driven by the ΔT. It is noteworthy that ΔT is mainly dominated by the ΔH (i.e., vertical motion) because the effect of latent heat from IC deposition/sublimation is much weaker than the ΔH (not shown). In terms of water vapor, IC deposition/sublimation (i.e., the change in qi) could directly change the RHi (the total water conservation) and then pull the RHi towards 100% (negative feedback). Taken overall, vertical motion can be considered an active external force that changes the RHi, while IC deposition/sublimation plays a role in mitigating the change in RHi.
At the beginning of the REF (Wave2) simulation (Figure 2), the RHi increased quickly and reached the homogeneous freezing threshold (150~160%). Homogeneous nucleation occurred and produced a lot of small ICs. Because the newly formed ICs (whose concentration was relatively high) grew by scavenging water vapor, the RHi sharply fell to 100%. In the subsequent simulation, these small ICs would be the first to vanish during the sublimation phase (i.e., RHi below 100%). This suggests that it is difficult for the small ICs produced by homogeneous freezing to survive for a long time if there are obvious vertical fluctuations.
As expected, both the Wamp and Wfre simulations showed that the maximum ΔH decreased as compared with the REF (Wave1) simulation (not shown). Therefore, the RHi fluctuations in the Wamp and Wfre simulations also became weak, and all ICs could survive during the sublimation phase (Figure 2). Furthermore, the maximum qi from the Wamp and Wfre simulations was much lower than that from the REF experiment. This suggests that stronger RHi fluctuations could help ICs to grow bigger (i.e., a larger Ri). The above phenomenon is clearer in the Wno simulation. Because of the IC deposition, the initial RHi (110%) was reduced to 100% in the first half hour of the Wno simulation. It is reasonable to infer that the RHi after 1 h of simulation is almost not related to the initial RHi. This is the reason why the first-hour simulation results are not used in the following statistical analysis. In short, vertical motion (i.e., external forcing) controls the change in RHi and determines cirrus evolution to some extent.
During the ICnosub simulation, all ICs could grow under RHi > 100% but could not sublimate under RHi < 100% (Figure 2). Correspondingly, the qi could only increase. As a result, the RHi gradually decreased during the air parcel vertical fluctuation because more and more water vapor was absorbed by the ICs when RHi > 100%. The ICadL and ICadH simulations explored the sensitivity to the αd. As mentioned above, IC deposition/sublimation could dampen the fluctuation of RHi (negative feedback). This effect increased/decreased with an increasing/decreasing the αd. During the first sublimation phase of the ICadH simulation, all ICs vanished and released water vapor to prevent the RHi from decreasing. As compared to the REF (Wave1) simulation, the fluctuation of the RHi weakened in the ICadH simulation due to the stronger efficiency of the IC deposition/sublimation. On the contrary, the fluctuation of the RHi became stronger in the ICadL simulation. Note that a strong fluctuation could allow the RHi to reach the homogeneous freezing threshold. After the occurrence of homogeneous nucleation, the huge number of ICs would bring the RHi down to 100% due to their strong efficiency in scavenging water vapor. The comparison among the ICadL, ICadH, and REF (Wave1) simulations clearly indicates that IC deposition/sublimation could dampen the fluctuation of RHi and make an impact on cirrus evolution.

3.2. Comparisons between the REF Experiment and Observations

This subsection illustrates the comparisons between the REF experiment (50,000 simulations) and observations (19,679 samples). It is noteworthy that the simulation results with Ni < 1 L−1 (only a very small part) were excluded from the statistical analysis, which is consistent with the observed samples. Furthermore, large-ensemble simulations provide very large samples (i.e., model output time points). Therefore, only a small part (200,000) of those samples (randomly selected) was used to create the plots below. It is necessary to point out that the probability distribution characteristic of these 200,000 samples is consistent with that of the total samples (not shown). The efficiency of ICs to scavenge/release water vapor is proportional to NiRi [55,56]. Here, NiRi is the sum of that from each IC’s size bin. In the following analysis, NiRi is used to indicate the effect of IC deposition/sublimation on RHi.
Figure 3 shows scatter plots of the W, RHi, and NiRi. As expected, in the scatter plot of RHi vs. NiRi, the 1000 samples selected for parcel model simulations have similar probability distribution characteristics with all observed samples. The scatter plot of RHi vs. W from the observations shows that the number of points (i.e., the samples defined by the RHi and W) in the first (RHi > 100%, and W > 0 m s−1) and third (RHi < 100%, and W < 0 m s−1) quadrants is larger than that in the second (RHi > 100%, and W < 0 m s−1) and fourth (RHi < 100%, and W > 0 m s−1). This might be explained by the fact that the RHi increased when the air parcel rose (i.e., W > 0 m s−1, and T is decreased), and the RHi decreased when the air parcel went down (i.e., W < 0 m s−1, and T is increased). This characteristic also appears in the corresponding scatter plot from the REF experiment. The observation data show that the number of samples with a higher NiRi (>1000 μm L−1) when RHi > 125% is obviously larger than that when RHi < 75%. The reason is that the Ri begins to increase when RHi > 100%, and Ni could possibly increase when RHi > 125% (if ice nucleation occurs). When RHi < 75%, ICs will have been sublimated for a while (Ri is decreased), and small ICs might vanish (Ni decreases). On the other hand, the number of samples with lower NiRi (<100 μm L−1) when RHi < 75% is obviously larger than that when RHi > 125%. Simulation results from the REF experiment also show these characteristics. In terms of the ICs’ impact on RHi, a higher NiRi indicates that the RHi could be quickly pulled back to 100%. Therefore, the RHi is much closer to 100% when NiRi > 10,000 μm L−1. This phenomenon is more obvious in the REF experiment. The scatter plots of W vs. NiRi from both the observations and the REF experiment show that a higher NiRi usually goes together with a stronger fluctuation range for the W. One possible reason is that the stronger W fluctuations could produce stronger RHi fluctuations, which might trigger ice nucleation events (i.e., Ni is increased). Another possible reason is that ICs could grow bigger (i.e., a larger Ri and qi) under stronger W fluctuations (see the case simulations). Generally speaking, these scatter plots from the REF experiment are similar to those from the observations.
Figure 4 shows the occurrence frequencies of RHi and NiRi. The RHi from the observations is mostly in the range of 50~150%. This is generally in agreement with previous studies [12,17,18]. The occurrence frequency of RHi from the REF experiment is very close to that from the observations. Furthermore, the occurrence frequency of NiRi from the REF experiment is also very close to that from the observations. Taken overall, the statistical characteristics of the W, RHi, and NiRi from the observations (Figure 1, Figure 3 and Figure 4) could almost be reproduced by the REF experiment (the large-ensemble simulations). This suggests that the wide range of RHi in the observations (wave-related cirrus clouds) can be well explained by the vertical motion and IC deposition/sublimation (two mechanisms under the closed adiabatic assumption).

3.3. Deep Analysis through Sensitivity Experiments

In this subsection, the effects of the dynamical (vertical motion) and microphysical (IC deposition/sublimation) processes on RHi fluctuations are further investigated through several sensitivity experiments. It is noteworthy that results from the Wno experiment are not shown in the figure below (i.e., Figure 5) because it is difficult to illustrate the special statistical characteristics in figures (almost all samples have RHi = 100%, similar to the corresponding case simulation in Figure 2). Furthermore, unless otherwise specified, the results of the sensitivity experiments are analyzed and compared with those of the REF experiment.
Without the vertical motion (i.e., the Wno experiment), the initial ICs scavenge/release water vapor, and then, in most cases, the RHi is pulled to 100% after the first hour (not shown), except for a few simulations with a much lower initial RHi (e.g., RHi = 60%) and a very small initial qi (i.e., ICs cannot provide enough water vapor). If RHi = 100%, there are no changes in the air parcel (i.e., cirrus evolution is stopped). This, again, suggests that vertical motion (i.e., external forcing) controls the change in RHi, as well as the cirrus evolution. After reducing the amplitude of the vertical motion by half (i.e., the Wamp experiment), the occurrence frequency distribution of W becomes narrow (Figure 5). The occurrence frequency distribution of the ΔT also becomes narrow due to the reduced W. The fluctuation of RHi is mostly determined by the ΔT. Therefore, the points (RHi and W) in the scatter plots (Figure 5) converge more towards the central point (RHi = 100%, and W = 0 m s−1). After doubling the wave frequency of the vertical motion (i.e., the Wfre experiment), the curve of the spectrum analysis moves towards shorter periods (Figure 5). As expected, the ΔH is reduced by half (not shown). Therefore, the occurrence frequency distribution of ΔT becomes narrow, while the occurrence frequency distribution of the W is almost unchanged (Figure 5). Here, the relation between ΔT and the W is not consistent with the approximated formula (δW = 0.23 δT) introduced in Section 2.2. This suggests that the wave spectrum characteristic used in the Wfre experiment is not consistent with reality. Furthermore, the points (RHi and W) in the scatter plot converge more towards the line of RHi = 100% (Figure 5). Although the occurrence frequency of the W is similar to that in the observations, the scatter plot of RHi vs. W is obviously different from the observations. This also suggests that the wave spectrum characteristic used in the Wfre experiment might not conform to reality. Both the Wamp and Wfre experiments show that the RHi is mainly within the range of 75~125% rather than 50~150% (Figure 5). The main reason for the narrow range of RHi is that the fluctuation of the ΔT, which is mostly caused by the W, is weakened. In short, both the amplitude and frequency of the vertical motion determine the change in RHi during cirrus evolution.
Figure 6 shows the results of the sensitivity experiments for the IC deposition/sublimation process. In the scatter plots (both RHi vs. W and RHi vs. NiRi) from the ICnosub experiment, most points are at an RHi < 100% because the ICs can only scavenge water vapor. The scatter plot of RHi vs. NiRi shows that the samples with a higher NiRi (> 100 μm L−1) are mostly from RHi < 75%. The reason is that a higher NiRi usually indicates that most water vapor has been absorbed by the ICs. Because the ICs grew bigger (i.e., a larger Ri) during the simulation without IC sublimation (see the case simulations), it is more obvious in the scatter plot of W vs. NiRi from the ICnosub experiment that a higher NiRi usually goes with a wider range of W. Although the relative occurrence frequency of RHi at around 100% from the ICadH experiment is obviously larger than that from the REF experiments, the three scatter plots from the ICadH experiment do not show any obvious differences. However, the differences between the scatter plots from the ICadL and REF experiments are obvious. Because the effect of IC deposition/sublimation is weakened, the range of RHi becomes wider in the scatter plots of RHi vs. W. The scatter plots of both RHi vs. NiRi and W vs. NiRi show a big increase in the number of samples with a higher NiRi (> 10,000 μm L−1). This can be explained by the fact that homogeneous nucleation occurs (a large number of ICs is produced) more frequently due to a stronger RHi fluctuation. Because the efficiency of IC growth (i.e., the increase in Ri) is reduced, the phenomenon where a higher NiRi (excluding NiRi > 10,000 μm L−1) goes with a wider range of W becomes weak in the scatter plot of W vs. NiRi. As compared to the ICnosub and ICadH experiments, the number of lower NiRi (<10 μm L−1) is obviously increased in the ICadL experiment because smaller ICs could survive longer. In short, it is clear that IC deposition/sublimation can significantly impact RHi as well as cirrus evolution.

4. Discussion

It is necessary to discuss the uncertainty of the observation data. The accuracy of RHi is 10~20% [39], while the precision of the W measurements is about 0.05 m s−1 [57]. Considering errors in measurement, the range of both W and RHi showed in this study might be slightly wider as compared with reality. Note that if the range of observed W and RHi (used for the parcel model initial state) becomes narrow, the range of simulated W and RHi also becomes narrow. The statistical characteristics of W and RHi from the observations could also be reproduced by the REF experiment (not shown). Because the ICs could be shattered during measurement, it is possible for Ni to be overestimated [13,44,58]. Under the same qi, NiRi is decreased with decreasing Ni. An increase in the αd (tunable parameter) can compensate for the impact of a decrease in NiRi (i.e., weaker efficiency of ICs scavenging/releasing water vapor). In short, the uncertainty of the observation data cannot change the conclusion of this study (i.e., the wide range of observed RHi can be well explained by the vertical motion and IC deposition/sublimation).
One should keep in mind that this study was carried out based on the closed adiabatic assumption and that only wave-related cirrus clouds were selected for analysis. In order to examine the possibility of extending the conclusion beyond the closed adiabatic assumption, we also analyzed the observation data for convective cirrus clouds (not only the MACPEX but also another field campaign) and corresponding large-ensemble simulations. The statistical characteristics of the observations could not be well reproduced by the large-ensemble simulations. However, these large-ensemble simulations still suggest that both vertical motion and IC deposition/sublimation are two important mechanisms for determining the RHi fluctuation.

5. Conclusions

This study investigated the possible main mechanisms of the wide range of RHi in cirrus clouds. Considering the current measurement and theoretical limitations of the entrainment and mixing process on cloud edges [28,29,30,31], only the vertical motion and IC deposition/sublimation related to in-cloud RHi (two mechanisms under the closed adiabatic assumption) were investigated through in situ observations and parcel model simulations. The adiabatic closed cloud parcel model used in this study was driven by a prescribed vertical motion. The mostly wave-related cirrus observations from the MACPEX were used to analyze cloud microphysical properties and carry out parcel model simulations.
The results of the parcel model case simulations showed that vertical motion is an active external force that changes the RHi, while IC deposition/sublimation plays a role in mitigating the change in the RHi. These can be clearly illustrated by the cloud microphysics theory. The results from the large-ensemble parcel model simulations almost reproduced the statistical cloud characteristics from the observations. This suggests that vertical motion and IC deposition/sublimation are two of the most important mechanisms involved in controlling the RHi fluctuation and could well explain the wide range of RHi in wave-related cirrus.
There is an additional finding in this study. The comparison of some statistical characteristics from the observation and simulation results could be used to constrain some uncertain factors that determine cloud evolution but are difficult to measure. For instance, the wave frequency of the vertical motion could significantly impact the statistical characteristics in the RHi vs. W scatter plot. Conversely, the statistical characteristics in the RHi vs. W scatter plot could also be used to infer the statistical characteristics of wave frequency. The αd is another example. The comparisons of the scatter plots of the W, RHi, and NiRi show that a very low value (αd = 0.001) for the αd is highly unlikely.

Author Contributions

X.S. designed this study and wrote the Fortran code of the air parcel model. M.Z. carried out the simulations used in this study and created all figures. X.S. and M.Z. wrote the manuscript. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the National Key Research and Development Program of China (grant nos. 2018YFC1507001 and 2017YFA0604001) and the National Natural Science Foundation of China (grant nos. 41775095 and 42075145). The APC was supported by the same funders.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The MAPEX observation data can be downloaded from the Earth Science Project Office (ESPO) https://espoarchive.nasa.gov/archive/browse/ (accessed on 6 December 2022). The Fortran code for the parcel model, the simulation results, and the NCL scripts have been archived in a public repository https://doi.org/10.5281/zenodo.7589993 (accessed on 31 January 2023).

Acknowledgments

This study was conducted at the High-Performance Computing Center of Nanjing University of Information Science and Technology.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Liou, K.-N. Influence of Cirrus Clouds on Weather and Climate Processes: A Global Perspective. Mon. Weather Rev. 1986, 114, 1167–1199. [Google Scholar] [CrossRef]
  2. Wang, P.-H.; Minnis, P.; McCormick, M.P.; Kent, G.S.; Skeens, K.M. A 6-year climatology of cloud occurrence frequency from Stratospheric Aerosol and Gas Experiment II observations (1985-1990). J. Geophys. Res. Atmos. 1996, 101, 29407–29429. [Google Scholar] [CrossRef]
  3. Luo, Z.; Rossow, W.B. Characterizing Tropical Cirrus Life Cycle, Evolution, and Interaction with Upper-Tropospheric Water Vapor Using Lagrangian Trajectory Analysis of Satellite Observations. J. Clim. 2004, 17, 4541–4563. [Google Scholar] [CrossRef]
  4. Gayet, J.-F.; Shcherbakov, V.; Mannstein, H.; Minikin, A.; Schumann, U.; Ström, J.; Petzold, A.; Ovarlez, J.; Immler, F. Microphysical and optical properties of midlatitude cirrus clouds observed in the southern hemisphere during INCA. Q. J. R. Meteorol. Soc. 2006, 132, 2719–2748. [Google Scholar] [CrossRef]
  5. Korolev, A.; Isaac, G.A. Relative Humidity in Liquid, Mixed-Phase, and Ice Clouds. J. Atmos. Sci. 2006, 63, 2865–2880. [Google Scholar] [CrossRef] [Green Version]
  6. Sassen, K.; Wang, Z.; Liu, D. Global distribution of cirrus clouds from CloudSat/Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) measurements. J. Geophys. Res. Atmos. 2008, 113, D00A12. [Google Scholar] [CrossRef]
  7. Bony, S.; Stevens, B.; Frierson, D.M.W.; Jakob, C.; Kageyama, M.; Pincus, R.; Shepherd, T.G.; Sherwood, S.C.; Siebesma, A.P.; Sobel, A.H.; et al. Clouds, circulation and climate sensitivity. Nat. Geosci. 2015, 8, 261–268. [Google Scholar] [CrossRef] [Green Version]
  8. Zhou, C.; Penner, J.E.; Lin, G.; Liu, X.; Wang, M. What controls the low ice number concentration in the upper troposphere. Atmos. Meas. Tech. 2016, 16, 12411–12424. [Google Scholar] [CrossRef] [Green Version]
  9. Barahona, D.; Molod, A.; Kalesse, H. Direct estimation of the global distribution of vertical velocity within cirrus clouds. Sci. Rep. 2017, 7, 6840. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  10. Ovarlez, J.; Gayet, J.-F.; Gierens, K.; Ström, J.; Ovarlez, H.; Auriol, F.; Busen, R.; Schumann, U. Water vapour measurements inside cirrus clouds in Northern and Southern hemispheres during INCA. Geophys. Res. Lett. 2002, 29, 60–61. [Google Scholar] [CrossRef] [Green Version]
  11. Comstock, J.M.; Ackerman, T.P.; Turner, D.D. Evidence of high ice supersaturation in cirrus clouds using ARM Raman lidar measurements. Geophys. Res. Lett. 2004, 31, 293–317. [Google Scholar] [CrossRef] [Green Version]
  12. Krämer, M.; Schiller, C.; Afchine, A.; Bauer, R.; Gensch, I.; Mangold, A.; Schlicht, S.; Spelten, N.; Sitnikov, N.; Borrmann, S.; et al. Ice supersaturations and cirrus cloud crystal numbers. Atmos. Meas. Tech. 2009, 9, 3505–3522. [Google Scholar] [CrossRef] [Green Version]
  13. Jensen, E.J.; Lawson, R.P.; Bergman, J.W.; Pfister, L.; Bui, T.P.; Schmitt, C.G. Physical processes controlling ice concentrations in synoptically forced, midlatitude cirrus. J. Geophys. Res. Atmos. 2013, 118, 5348–5360. [Google Scholar] [CrossRef]
  14. Cziczo, D.J.; Froyd, K.D.; Hoose, C.; Jensen, E.J.; Diao, M.; Zondlo, M.A.; Smith, J.B.; Twohy, C.H.; Murphy, D.M. Clarifying the Dominant Sources and Mechanisms of Cirrus Cloud Formation. Science 2013, 340, 1320–1324. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  15. Spichtinger, P.; Gierens, K.M. Modelling of cirrus clouds—Part 1a: Model description and validation. Atmos. Meas. Tech. 2009, 9, 685–706. [Google Scholar] [CrossRef] [Green Version]
  16. Kärcher, B. Supersaturation Fluctuations in Cirrus Clouds Driven by Colored Noise. J. Atmos. Sci. 2012, 69, 435–443. [Google Scholar] [CrossRef] [Green Version]
  17. Krämer, M.; Rolf, C.; Spelten, N.; Afchine, A.; Fahey, D.; Jensen, E.; Khaykin, S.; Kuhn, T.; Lawson, P.; Lykov, A.; et al. A microphysics guide to cirrus—Part 2: Climatologies of clouds and humidity from observations. Atmos. Meas. Tech. 2020, 20, 12569–12608. [Google Scholar] [CrossRef]
  18. Diao, M.; Zondlo, M.A.; Heymsfield, A.J.; Avallone, L.M.; Paige, M.E.; Beaton, S.P.; Campos, T.; Rogers, D.C. Cloud-scale ice-supersaturated regions spatially correlate with high water vapor heterogeneities. Atmos. Meas. Tech. 2014, 14, 2639–2656. [Google Scholar] [CrossRef] [Green Version]
  19. Murphy, D.M. Rare temperature histories and cirrus ice number density in a parcel and a one-dimensional model. Atmos. Meas. Tech. 2014, 14, 13013–13022. [Google Scholar] [CrossRef] [Green Version]
  20. Muhlbauer, A.; Berry, E.; Comstock, J.M.; Mace, G.G. Perturbed physics ensemble simulations of cirrus on the cloud system-resolving scale. J. Geophys. Res. Atmos. 2014, 119, 4709–4735. [Google Scholar] [CrossRef]
  21. Dinh, T.; Podglajen, A.; Hertzog, A.; Legras, B.; Plougonven, R. Effect of gravity wave temperature fluctuations on homogeneous ice nucleation in the tropical tropopause layer. Atmos. Meas. Tech. 2016, 16, 35–46. [Google Scholar] [CrossRef] [Green Version]
  22. Podglajen, A.; Plougonven, R.; Hertzog, A.; Jensen, E. Impact of gravity waves on the motion and distribution of atmospheric ice particles. Atmos. Meas. Tech. 2018, 18, 10799–10823. [Google Scholar] [CrossRef] [Green Version]
  23. Horner, G.A.; Gryspeerdt, E. The evolution of deep convective systems and their associated cirrus outflows. Atmos. Chem. Phys. Discuss. 2022, 1–22. [Google Scholar] [CrossRef]
  24. Maciel, F.V.; Diao, M.; Patnaude, R. Examination of aerosol indirect effects during cirrus cloud evolution. Atmos. Meas. Tech. 2023, 23, 1103–1129. [Google Scholar] [CrossRef]
  25. Stommel, H. Entrainment of Air into a Cumulus Cloud. J. Meteorol. 1947, 4, 91–94. [Google Scholar] [CrossRef]
  26. Burnet, F.; Brenguier, J.-L. Observational Study of the Entrainment-Mixing Process in Warm Convective Clouds. J. Atmos. Sci. 2007, 64, 1995–2011. [Google Scholar] [CrossRef]
  27. Pinsky, M.; Khain, A. Theoretical Analysis of the Entrainment–Mixing Process at Cloud Boundaries. Part II: Motion of Cloud Interface. J. Atmos. Sci. 2019, 76, 2599–2616. [Google Scholar] [CrossRef]
  28. Lu, C.; Liu, Y.; Niu, S. Examination of turbulent entrainment-mixing mechanisms using a combined approach. J. Geophys. Res. Atmos. 2011, 116, D20. [Google Scholar] [CrossRef]
  29. Yum, S.S.; Wang, J.; Liu, Y.; Senum, G.; Springston, S.; McGraw, R.; Yeom, J.M. Cloud microphysical relationships and their implication on entrainment and mixing mechanism for the stratocumulus clouds measured during the VOCALS project. J. Geophys. Res. Atmos. 2015, 120, 5047–5069. [Google Scholar] [CrossRef]
  30. Ghate, V.P.; Cadeddu, M.P. Drizzle and Turbulence Below Closed Cellular Marine Stratocumulus Clouds. J. Geophys. Res. Atmos. 2019, 124, 5724–5737. [Google Scholar] [CrossRef] [Green Version]
  31. Desai, N.; Liu, Y.; Glienke, S.; Shaw, R.A.; Lu, C.; Wang, J.; Gao, S. Vertical Variation of Turbulent Entrainment Mixing Processes in Marine Stratocumulus Clouds Using High-Resolution Digital Holography. J. Geophys. Res. Atmos. 2021, 126, e2020JD033527. [Google Scholar] [CrossRef]
  32. Pruppacher, H.; Klett, J. Microphysics of Clouds and Precipitation, 2nd ed.; Kluwer Academic Publishers: New York, NY, USA, 1996; pp. 502–561. [Google Scholar] [CrossRef]
  33. Haag, W. The impact of aerosols and gravity waves on cirrus clouds at midlatitudes. J. Geophys. Res. Atmos. 2004, 109, D12. [Google Scholar] [CrossRef] [Green Version]
  34. Jensen, E.; Pfister, L.; Bui, T.; Weinheimer, A.; Weinstock, E.; Smith, J.; Pittman, J.; Baumgardner, D.; Lawson, P.; McGill, M.J. Formation of a tropopause cirrus layer observed over Florida during CRYSTAL-FACE. J. Geophys. Res. Atmos. 2005, 110, D3. [Google Scholar] [CrossRef] [Green Version]
  35. Schiller, C.; Krämer, M.; Afchine, A.; Spelten, N.; Sitnikov, N. Ice water content of Arctic, midlatitude, and tropical cirrus. J. Geophys. Res. Atmos. 2008, 113, D24. [Google Scholar] [CrossRef] [Green Version]
  36. Kärcher, B.; Burkhardt, U. A cirrus cloud scheme for general circulation models. Q. J. R. Meteorol. Soc. 2008, 134, 1439–1461. [Google Scholar] [CrossRef] [Green Version]
  37. Barahona, D.; Nenes, A. Dynamical states of low temperature cirrus. Atmos. Meas. Tech. 2011, 11, 3757–3771. [Google Scholar] [CrossRef] [Green Version]
  38. Shi, X.; Liu, X. Effect of cloud-scale vertical velocity on the contribution of homogeneous nucleation to cirrus formation and radiative forcing. Geophys. Res. Lett. 2016, 43, 6588–6595. [Google Scholar] [CrossRef] [Green Version]
  39. Krämer, M.; Rolf, C.; Luebke, A.; Afchine, A.; Spelten, N.; Costa, A.; Meyer, J.; Zöger, M.; Smith, J.; Herman, R.L.; et al. A microphysics guide to cirrus clouds—Part 1: Cirrus types. Atmos. Meas. Tech. 2016, 16, 3463–3483. [Google Scholar] [CrossRef] [Green Version]
  40. Jensen, E.J.; Diskin, G.S.; DiGangi, J.; Woods, S.; Lawson, R.P.; Bui, T.V. Homogeneous Freezing Events Sampled in the Tropical Tropopause Layer. J. Geophys. Res. Atmos. 2022, 127, e2022JD036535. [Google Scholar] [CrossRef]
  41. Strandgren, J. The Life Cycle of Anvil Cirrus Clouds from a Combination of Passive and Active Satellite Remote Sensing. Ph.D. Thesis, The Ludwig Maximilian University, München, Germany, 6 September 2018. [Google Scholar] [CrossRef]
  42. Rollins, A.W.; Thornberry, T.D.; Gao, R.S.; Smith, J.B.; Sayres, D.S.; Sargent, M.R.; Schiller, C.; Krämer, M.; Spelten, N.; Hurst, D.F.; et al. Evaluation of UT/LS hygrometer accuracy by intercomparison during the NASA MACPEX mission. J. Geophys. Res. Atmos. 2014, 119, 1915–1935. [Google Scholar] [CrossRef] [PubMed]
  43. Xu, Z.; Mace, G.G. Ice Particle Mass–Dimensional Relationship Retrieval and Uncertainty Evaluation Using the Optimal Estimation Methodology Applied to the MACPEX Data. J. Appl. Meteorol. Clim. 2017, 56, 767–788. [Google Scholar] [CrossRef]
  44. Scott, S.G.; Bui, T.P.; Chan, K.R.; Bowen, S.W. The Meteorological Measurement System on the NASA ER-2 Aircraft. J. Atmos. Ocean. Technol. 1990, 7, 525–540. [Google Scholar] [CrossRef]
  45. Weinstock, E.M.; Hintsa, E.J.; Dessler, A.E.; Oliver, J.F.; Hazen, N.L.; Demusz, J.N.; Allen, N.T.; Lapson, L.B.; Anderson, J.G. New fast response photofragment fluorescence hygrometer for use on the NASA ER-2 and the Perseus remotely piloted aircraft. Rev. Sci. Instrum. 1994, 65, 3544–3554. [Google Scholar] [CrossRef]
  46. Murphy, D.M.; Koop, T. Review of the vapour pressures of ice and supercooled water for atmospheric applications. Q. J. R. Meteorol. Soc. 2005, 131, 1539–1565. [Google Scholar] [CrossRef]
  47. Lawson, R.P.; O’Connor, D.; Zmarzly, P.; Weaver, K.; Baker, B.; Mo, Q.; Jonsson, H. The 2D-S (Stereo) Probe: Design and Preliminary Tests of a New Airborne, High-Speed, High-Resolution Particle Imaging Probe. J. Atmos. Ocean. Technol. 2006, 23, 1462–1477. [Google Scholar] [CrossRef] [Green Version]
  48. Khvorostyanov, V.I.; Sassen, K. Microphysical processes in cirrus and their impact on radiation: A mesoscale modeling perspective. In Cirrus, 1st ed.; Lynch, D.K., Sassen, K., Starr, D.O.C., Stephens, G., Eds.; Oxford University Press: New York, NY, USA, 2002; Chapter 19; pp. 397–432. [Google Scholar] [CrossRef]
  49. Wang, M.; Liu, X.; Zhang, K.; Comstock, J.M. Aerosol effects on cirrus through ice nucleation in the Community Atmosphere Model CAM5 with a statistical cirrus scheme. J. Adv. Model. Earth Syst. 2014, 6, 756–776. [Google Scholar] [CrossRef]
  50. Kärcher, B.; Podglajen, A. A Stochastic Representation of Temperature Fluctuations Induced by Mesoscale Gravity Waves. J. Geophys. Res. Atmos. 2019, 124, 11506–11529. [Google Scholar] [CrossRef] [Green Version]
  51. Lu, X.; Chu, X.; Li, H.; Chen, C.; Smith, J.A.; Vadas, S.L. Statistical characterization of high-to-medium frequency mesoscale gravity waves by lidar-measured vertical winds and temperatures in the MLT. J. Atmos. Sol. -Terr. Phys. 2017, 162, 3–15. [Google Scholar] [CrossRef]
  52. Kalesse, H.; Kollias, P. Climatology of High Cloud Dynamics Using Profiling ARM Doppler Radar Observations. J. Clim. 2013, 26, 6340–6359. [Google Scholar] [CrossRef]
  53. Gierens, K.M.; Monier, M.; Gayet, J. The deposition coefficient and its role for cirrus clouds. J. Geophys. Res. Atmos. 2003, 108, D2. [Google Scholar] [CrossRef]
  54. Skrotzki, J.; Connolly, P.; Schnaiter, M.; Saathoff, H.; Möhler, O.; Wagner, R.; Niemand, M.; Ebert, V.; Leisner, T. The accommodation coefficient of water molecules on ice-cirrus cloud studies at the AIDA simulation chamber. Atmos. Meas. Tech. 2013, 13, 4451–4466. [Google Scholar] [CrossRef] [Green Version]
  55. Korolev, A.V.; Mazin, I.P. Supersaturation of Water Vapor in Clouds. J. Atmos. Sci. 2003, 60, 2957–2974. [Google Scholar] [CrossRef]
  56. Zhang, W.; Shi, X.; Lu, C. Impacts of the ice-particle size distribution shape parameter on climate simulations with the Community Atmosphere Model Version 6 (CAM6). Geosci. Model Dev. 2022, 15, 7751–7766. [Google Scholar] [CrossRef]
  57. Jensen, E.J.; Ueyama, R.; Pfister, L.; Bui, T.V.; Alexander, M.J.; Podglajen, A.; Hertzog, A.; Woods, S.; Lawson, R.P.; Kim, J.; et al. High-frequency gravity waves and homogeneous ice nucleation in tropical tropopause layer cirrus. Geophys. Res. Lett. 2016, 43, 6629–6635. [Google Scholar] [CrossRef]
  58. Mascio, J.; Mace, G.G. Quantifying uncertainties in radar forward models through a comparison between CloudSat and SPartICus reflectivity factors. J. Geophys. Res. Atmos. 2017, 122, 1665–1684. [Google Scholar] [CrossRef]
Atmosphere 14 00583 g001 550
Figure 1. The statistical characteristics of the 50 prescribed W time series used to drive the parcel model (first row). They are the occurrence frequencies of W and ΔT (left) and the spectrum analysis of the W time series (right). The second and third rows present two cases (Wave1 and Wave2, respectively). The left columns show the time series of W (gray solid line), the ΔH (black solid line), and ΔT (red dotted line). The right columns show the spectrum analyses of the corresponding W time series. The ΔT and ΔH were calculated using the parcel model, with the initial conditions of T = 225 K and P = 250 hPa.
Figure 1. The statistical characteristics of the 50 prescribed W time series used to drive the parcel model (first row). They are the occurrence frequencies of W and ΔT (left) and the spectrum analysis of the W time series (right). The second and third rows present two cases (Wave1 and Wave2, respectively). The left columns show the time series of W (gray solid line), the ΔH (black solid line), and ΔT (red dotted line). The right columns show the spectrum analyses of the corresponding W time series. The ΔT and ΔH were calculated using the parcel model, with the initial conditions of T = 225 K and P = 250 hPa.
Atmosphere 14 00583 g001
Atmosphere 14 00583 g002 550
Figure 2. Cirrus evolutions from parcel model simulations using the W time series shown in Figure 1 (Wave1 or Wave2). Shown are RHi (blue thick solid line), W (gray thin solid line), Ni (orange thin solid line), qi (green thin solid line). Black dotted line represents the line of W = 0, also the line of RHi = 100%. The shaded figures indicate the Ni contribution from each radius bin. A total of 50 bins were used in this study. Experiment names and the corresponding W time series are marked in the upper left corners. Note that the ranges of the Ni-axis and RHi-axis from the ICadL experiment are different from the others.
Figure 2. Cirrus evolutions from parcel model simulations using the W time series shown in Figure 1 (Wave1 or Wave2). Shown are RHi (blue thick solid line), W (gray thin solid line), Ni (orange thin solid line), qi (green thin solid line). Black dotted line represents the line of W = 0, also the line of RHi = 100%. The shaded figures indicate the Ni contribution from each radius bin. A total of 50 bins were used in this study. Experiment names and the corresponding W time series are marked in the upper left corners. Note that the ranges of the Ni-axis and RHi-axis from the ICadL experiment are different from the others.
Atmosphere 14 00583 g002
Atmosphere 14 00583 g003 550
Figure 3. The scatter plots of W, RHi, and NiRi from observations (upper panel) and the REF experiment (lower panel). The dotted lines indicate RHi = 100% or W = 0 m s−1. Note that the 1000 samples selected for parcel model simulations are marked by red dots.
Figure 3. The scatter plots of W, RHi, and NiRi from observations (upper panel) and the REF experiment (lower panel). The dotted lines indicate RHi = 100% or W = 0 m s−1. Note that the 1000 samples selected for parcel model simulations are marked by red dots.
Atmosphere 14 00583 g003
Atmosphere 14 00583 g004 550
Figure 4. The occurrence frequencies of NiRi and RHi from observations (black lines) and the REF experiment (red lines). The dotted line indicates RHi = 100%.
Figure 4. The occurrence frequencies of NiRi and RHi from observations (black lines) and the REF experiment (red lines). The dotted line indicates RHi = 100%.
Atmosphere 14 00583 g004
Atmosphere 14 00583 g005 550
Figure 5. The occurrence frequency of RHi from the REF, Wamp, and Wfre experiments (upper panel). The lower panel shows the spectrum analysis of the W time series (left), the distributions of W and ΔT (middle), and the scatter plots of RHi vs. W (right) from the Wamp (second row) and Wfre (third row) experiments.
Figure 5. The occurrence frequency of RHi from the REF, Wamp, and Wfre experiments (upper panel). The lower panel shows the spectrum analysis of the W time series (left), the distributions of W and ΔT (middle), and the scatter plots of RHi vs. W (right) from the Wamp (second row) and Wfre (third row) experiments.
Atmosphere 14 00583 g005
Atmosphere 14 00583 g006 550
Figure 6. The occurrence frequency of RHi from the Ref, ICnosub, ICadH, and ICadL experiments (upper panel). The lower panel shows scatter plots similar to those in Figure 4 but for the ICnosub (second row), ICadH (third row), and ICadL (fourth row) experiments.
Figure 6. The occurrence frequency of RHi from the Ref, ICnosub, ICadH, and ICadL experiments (upper panel). The lower panel shows scatter plots similar to those in Figure 4 but for the ICnosub (second row), ICadH (third row), and ICadL (fourth row) experiments.
Atmosphere 14 00583 g006
Table
Table 1. The experiments carried out in this study.
Table 1. The experiments carried out in this study.
ExperimentDescription
REFReference experiment. The αd is 0.05.
Sensitivity experiments for vertical motion
WampSame as REF, but the amplitude of the W time series is reduced to half.
WfreSame as REF, but the frequency of the W time series is doubled.
WnoSame as REF, but W has a constant value of 0.
Sensitivity experiments for the IC deposition/sublimation process.
ICnosubSame as REF, but the sublimation process is not allowed.
ICadHSame as REF, but the αd is set to 1.0.
ICadLSame as REF, but the αd is set to 0.001.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Zhao, M.; Shi, X. A Study on the Wide Range of Relative Humidity in Cirrus Clouds Using Large-Ensemble Parcel Model Simulations. Atmosphere 2023, 14, 583. https://doi.org/10.3390/atmos14030583

AMA Style

Zhao M, Shi X. A Study on the Wide Range of Relative Humidity in Cirrus Clouds Using Large-Ensemble Parcel Model Simulations. Atmosphere. 2023; 14(3):583. https://doi.org/10.3390/atmos14030583

Chicago/Turabian Style

Zhao, Miao, and Xiangjun Shi. 2023. "A Study on the Wide Range of Relative Humidity in Cirrus Clouds Using Large-Ensemble Parcel Model Simulations" Atmosphere 14, no. 3: 583. https://doi.org/10.3390/atmos14030583

APA Style

Zhao, M., & Shi, X. (2023). A Study on the Wide Range of Relative Humidity in Cirrus Clouds Using Large-Ensemble Parcel Model Simulations. Atmosphere, 14(3), 583. https://doi.org/10.3390/atmos14030583

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop