Cirrus Formation Research Articles

Impact of the Mount Pinatubo eruption on cirrus clouds formed by homogeneous freezing in the ECHAM4 GCM

Volcanic emissions may have the potential to alter cirrus cloud properties. Here we conduct different sensitivity studies with the ECHAM4 general circulation model for 2.5 years after the Mount Pinatubo eruption (July 1991 to December 1993) and compare homogeneous cirrus formation caused by sedimenting sulfate particles produced in the eruption plume with homogeneous cirrus formation in the undisturbed atmosphere. In the first scenario, the sulfate aerosol mass from Pinatubo is added to the background aerosol concentration assuming a monomodal aerosol. Here the aerosol concentration increases by up to 3000 cm−3 in 1992, which can be regarded as an upper limit more representative for the first months after the eruption. The ice crystal number concentration increases by up to 1 cm−3 near the tropical tropopause more than doubling the pre‐existing concentration in this region one year after the eruption. In the second, more realistic, scenario the Pinatubo aerosol is added to a bimodal background size distribution as a separate large particle mode. Here the aerosol number concentration increases by 10–25 cm−3, which can be regarded as a lower bound more representative for what has been observed in 1992. Then the ice crystal number increases at most 50% in the tropics in 1992. Satellite observations show an increase in ice water path starting in 1992 that could be related to either the Mount Pinatubo eruption or the El Niño event or both but a decrease in total cloud cover. While there is no trend on cloud microphysical or optical properties in our second scenario, the first scenario shows a pronounced increase in ice water path and a noticeable impact on cloud radiative forcing.

A parameterization of cirrus cloud formation: Heterogeneous freezing

A physically based parameterization of cirrus cloud formation by heterogeneous freezing is developed along with a novel method to compute associated nucleation rates. The analysis is restricted to immersion freezing, possibly the dominant pathway for heterogeneous cirrus formation under cold (<235 K) conditions. The size of ice nuclei (IN) immersed in a liquid particle does not significantly affect the heterogeneous freezing threshold (the saturation ratio over ice where ice formation is initiated) of the mixed particle. If perfect IN were present at cirrus altitudes, almost all of them would freeze near ice saturation, even in slow updrafts. If only one type of less potent IN with freezing thresholds >1.3–1.4 triggers cirrus formation, cloud properties are not very susceptible to changes of IN properties, as in the case of homogeneous freezing. In contrast, much stronger indirect aerosol effects on cirrus clouds are possible if at least two types of IN with distinct freezing thresholds compete during the freezing process, most likely leading to a suppression of ice crystal concentrations.

On the transition between heterogeneous and homogeneous freezing

Abstract. Box model simulations of an uplifting and adiabatically cooling cloud of aerosol have been performed in order to study the transition between cirrus formation dominated by homogeneous nucleation of ice to that dominated by heterogeneous nucleation. The aerosol was assumed to consist of an internal mixture of sulfuric acid solution droplets with inclusions of soot. The parametrisation of De Mott et al. (1997) was used to simulate the heterogeneous nucleation of ice in such droplets with soot inclusions. The simulations show that the transition from heterogeneous to homogeneous nucleation occurs over a narrow range of soot concentration. Thus it seems to be possible to fix critical concentrations of heterogeneous ice nuclei which must be exceeded if heterogeneous freezing dominates cirrus formation. A formula has been derived that allows to compute the critical concentrations of heterogeneous ice nuclei as a function of temperature, updraft speed, ambient pressure, and supersaturation at which heterogeneous freezing occurs. Generally, homogeneous nucleation dominates in regions with updrafts stronger than 20 cm s -1, with the exception of heavily polluted areas which could be common in the northern hemisphere due to air traffic, where updrafts of the order 1 m s-1 may be necessary to render heterogeneous nucleation unimportant. According to the present results it cannot be excluded that heterogeneous nucleation plays a more important role for cirrus formation in the northern midlatitudes than anywhere else. A possible consequence of these results is that air pollution may lead to a higher coverage of cirrus clouds, but then these clouds will be optically thinner than clouds formed by homogeneous freezing, with the exception of regions where condensation trails are frequent.

Potential impact of subsonic and supersonic aircraft exhaust on water vapor in the lower stratosphere assessed via a trajectory model

We employ a trajectory model to assess the impact on the stratosphere of water vapor present in the exhaust of subsonic and a proposed fleet of supersonic aircraft. Air parcels into which water vapor from aircraft exhaust has been injected are run through a 6‐year simulation in the trajectory model using meteorological data from the UKMO analyses with emissions dictated by the standard 2015 emissions scenario. For the subsonic aircraft, our results suggest maximum enhancements of ∼150 ppbv just above the Northern Hemisphere tropopause and of much less than 50 ppbv in most other regions. Inserting the perturbed water vapor profiles into a radiative transfer model, but not considering the impact of additional cirrus formation resulting from emissions by subsonic aircraft, we find that the impact of subsonic water vapor emissions on the radiative balance is negligible. For the supersonic case, our results show maximum enhancements of ∼1.5 ppmv in the tropical stratosphere near 20 km. Much of the remaining stratosphere between 12 and 25 km sees enhancements of greater than 0.1 ppmv, although enhancements above 35 km are generally less than 50 ppbv, in contrast to previous 2‐D and 3‐D model studies. Radiative calculations based upon these projected water vapor perturbations indicate they may cause a nonnegligible impact on tropical temperature profiles. Since our trajectory model includes no chemistry and our radiative calculations use the most extreme water vapor perturbations, our results should be viewed as upper limits on the potential impacts.

Cirrus Cloud Properties from a Cloud-Resolving Model Simulation Compared to Cloud Radar Observations

Abstract Cloud radar data collected at the Atmospheric Radiation Measurement (ARM) Program's Southern Great Plains site were used to evaluate the properties of cirrus clouds that occurred in a cloud-resolving model (CRM) simulation of the 29-day summer 1997 intensive observation period (IOP). The simulation was “forced” by the large-scale advective temperature and water vapor tendencies, horizontal wind velocity, and turbulent surface fluxes observed at the Southern Great Plains site. The large-scale advective condensate tendency was not observed. The correlation of CRM cirrus amount with Geostationary Operational Environmental Satellite (GOES) high cloud amount was 0.70 for the subperiods during which cirrus formation and decay occurred primarily locally, but only 0.30 for the entire IOP. This suggests that neglecting condensate advection has a detrimental impact on the ability of a model (CRM or single-column model) to properly simulate cirrus cloud occurrence. The occurrence, vertical location, and thi...

Ice nucleation by kerosene soot under upper tropospheric conditions

Quasi-elastic neutron scattering (QENS) and neutron diffraction (ND) experiments were conducted to characterize the dynamics and structure of water/ice confined in a well-defined kerosene-soot pores network at 40, 60, and 100% relative humidity and at temperatures from 283 to 204 K. The QENS spectra indicate that about 35% of the adsorbed water is in an amorphous solid-like state at T = 283 K and that an appreciable amount of liquid water exists in the soot supermicropores down to the lowest temperature investigated, i.e., 204 K. The depression in freezing temperature is related to the pore dimensions. At T [Formula: see text] 260 K ND shows a mixture of amorphous ice, probably located in the soot mesopores coexisting with ice Ih at the surface of the soot primary particles. The results are discussed within the context of aircraft-generated soot aerosols in the upper troposphere as candidates for heterogeneous ice nucleation responsible for contrails and cirrus formation. PACS Nos.: 82.33Tb, 92.60Mt, 81.05Rm, 61.12-q

Reconsideration of the morphogenesis in the marine hypotrichous ciliate, Aspidisca leptaspis Fresenius, 1865 (Protozoa, Ciliophora)

Morphogenetic events during division of the marine hypotrichous ciliate, Aspidisca leptaspis were investigated. The oral primordium of the opisthe develops hypoapokinetally in a sub-surface pouch posterior to the rear parental adoral zone, as in congeners. The parental adoral zones of membranelles are inherited by the proter. The proter's buccal cirrus develops from an “extra” anlage, which is formed de novo on the cell surface at an early stage of morphogenesis. Basal bodies in the oral primordium seem not to contribute to the formation of the buccal cirrus of the opisthe, which originates independently within the pouch opposite to the oral primordium and then migrates onto the cell surface. The paroral membrane derives from the posterior end of the oral primordium. Five frontoventral-transverse cirral anlagen are also formed de novo, initially as primary primordia, and develop into 3:3:2:2:2 cirri, respectively, in both dividers. Each of the four dorsal kineties evolves from an intrakinetal anlage in both proter and opisthe. Replication bands progress from the centre to the two ends of the macronucleus during the first half of the morphogenetic process, before this nucleus shortens and divides. The present investigation has clarified division processes within this species and confirmed some morphogenetic differences within the genus.

The Implications of Tropical Rossby Waves for Tropical Tropopause Cirrus Formation and for the Equatorial Upwelling of the Brewer–Dobson Circulation

This study puts forward a mechanism for the observed upwelling in the tropical upper troposphere and lower stratosphere. In this hypothesis, the tropical upwelling is driven by momentum transport by Rossby waves that are generated by tropical convection. To test this hypothesis, model runs are conducted using an axisymmetric, global, primitive equation model. In these runs, the effect of Rossby waves is included by driving the model with observed fields of large-scale eddy momentum flux convergence. The resulting overturning circulation includes both meridional flow from the intertropical convergence zone (ITCZ) to the equator and rising motion in the tropical tropopause transition layer (TTL). This circulation therefore helps to explain the transport of moisture from the lower portion of the TTL in the ITCZ to the equatorial cold-point tropopause, where tropopause cirrus layers frequently occur.

A Parameterization of cirrus cloud formation: Homogeneous freezing including effects of aerosol size

In a previous study, we have derived an approximate, physically based parameterization of cirrus cloud formation by homogeneous freezing, applying to a wide class of supercooled aerosols in the upper troposphere and tropopause region. In this study, the parameterization scheme is extended to include the effects of aerosol size on the freezing process in adiabatically rising air parcels. Aerosol size effects become important when the timescale of the freezing event is fast compared to the timescale of depositional growth of the pristine ice particles. The generalized parameterization scheme is validated with parcel model simulations and can directly be applied in models that do not explicitly resolve the ice nucleation process, such as cloud‐resolving models, weather forecast models, and climate models. The relationship between aerosol and ice crystal number concentrations in cirrus clouds formed by homogeneous freezing is discussed. This relationship is much weaker than in liquid water clouds. It is shown that even freezing of enhanced levels of sulfate aerosol originating from strong volcanic eruptions is unlikely to exert a sensible influence on cirrus formation.

Effect of Mount Pinatubo H2SO4/H2O aerosol on ice nucleation in the upper troposphere using a global chemistry and transport model

A 2‐year simulation of Mount Pinatubo volcanic aerosol is performed using a global chemistry and transport model. The model is driven by meteorological fields from the NASA Goddard Data Assimilation Office (DAO) general circulation model (GCM) with a horizontal resolution of 2° latitude by 2.5° longitude and 46 vertical levels. The model reproduces the equatorial aerosol reservoir bounded between 20°S and 30°N several months after the eruption and the detrainment of sulfate aerosol from the equatorial reservoir occurring in connection with anticyclonic circulation systems. Our model results are generally consistent with the dispersal characteristics of sulfate aerosol derived from the Stratospheric Aerosol and Gas Experiment II (SAGE II) aerosol extinction observations. On the basis of global distribution of Pinatubo sulfate, the homogeneous ice nucleation rate of H2SO4/H2O aerosol (number of aerosol particles freezing per cm3 of air mass and per second) in the lower stratospheric and upper tropospheric regions is calculated and compared with those of natural and anthropogenic sulfate from sources injected near Earth's surface. Large ice nucleation rates from Pinatubo H2SO4/H2O aerosol are found in the region near the equatorial tropopause and generally in the regions along the tropopause in the bottom of the main Pinatubo aerosol layer during the first year following the eruption. Ice nucleation rates of Mount Pinatubo H2SO4/H2O aerosols are much larger than those of sulfate aerosols from the surface sources in many regions of the upper troposphere during the first year and could still be comparable to rates from these background sources during the second year after the eruption. Our results suggest that the Pinatubo sulfate aerosol could have influenced cirrus formation and evolution globally through homogeneous ice nucleation, although the exact impact also requires a more complete calculation that includes cirrus cloud dynamics and microphysics. Available satellite observations support the above assertion.

First interactive simulations of cirrus clouds formed by homogeneous freezing in the ECHAM general circulation model

A physically based parameterization for the nucleation and initial growth of ice crystals in cirrus clouds at low (<238 K) temperatures prevailing in the upper troposphere and in the tropopause region is implemented in the ECHAM general circulation model (GCM). With this model we performed the first interactive simulations of cirrus clouds in a global climate model. As homogeneous ice crystal nucleation requires relative humidities with respect to ice of 140–170% at these temperatures, we abandoned the saturation adjustment scheme previously used to compute the properties of cirrus clouds in GCMs. The comparison with relative humidity data taken on commercial aircraft showed that ECHAM reproduces the frequency distribution of supersaturation with respect to ice in cloud‐free regions well especially if high values of the subgrid‐scale vertical velocity are assumed. On the other hand, the higher frequency of occurrence of high supersaturations in the model may point to heterogeneous freezing as the missing mechanism for ice formation. We found that homogeneous freezing may be limited by the number of hygroscopic aerosols present. However, as the ice water content remains the same for the natural and present‐day aerosol scenarios the longwave radiation does not change. Thus anthropogenic aerosol and precursor emissions have only a small effect on cirrus formed by homogeneous freezing. Aircraft emissions of sulfate aerosols are not likely to be important for cirrus formation. Aircraft soot emissions may be important if the soot particles nucleate ice more efficiently than by homogeneous freezing.

A parameterization of cirrus cloud formation: Homogeneous freezing of supercooled aerosols

The nucleation and initial growth of ice crystals in cirrus clouds at low (<235 K) temperatures prevailing in the upper troposphere and in the tropopause region is theoretically considered. The analysis explains the dependence of the number density of ice crystals on the vertical velocity and temperature seen in numerical simulations of cirrus formation when the timescale of depositional growth of the pristine ice particles is fast compared to the timescale of the freezing event. In such cases, applicable in many situations, the number of crystals formed via homogeneous freezing of aqueous solution droplets is rather insensitive to details of the aerosol size distribution, but increases rapidly with updraft velocity and decreases with temperature. The derived parameterization is validated with parcel model simulations, and its applicability for use in climate models is discussed. The potential role of aerosol size and heterogeneous freezing processes in altering the predicted cirrus properties is briefly addressed.

Modeling with explicit spectral water and ice microphysics of a two‐layer cloud system of altostratus and cirrus observed during the FIRE Arctic Clouds Experiment

A one‐dimensional version of a cloud model with an explicit microphysics scheme is used to simulate a case study of middle and upper level cloud formation and evolution that was observed during the FIRE Arctic Clouds Experiment. In the simulations, the midlevel altostratus cloud is initially liquid phase, then partially freezes, and exists in mixed phase for several hours with a relative equilibrium between the rate of drop production by condensation and their depletion by freezing. The dominant mode of cirrus formation was periodic homogeneous freezing of deliquescent submicron haze particles. These crystal layers form near the tropopause and, subsequently, precipitate into the middle troposphere, causing seeding of the underlying altostratus cloud. Sensitivity tests are conducted varying the initial humidity and nucleation schemes.

A Comparison of Surface Observations and ECHAM4-GCM Experiments and Its Relevance to the Indirect Aerosol Effect

The observations of solar irradiance at the surface, total cloud cover, and precipitation rates have been used to evaluate aerosol–cloud interactions in a GCM. Records from Germany and the United States were available for the time period from 1985 to 1990 and 1960 to 1990. The model used here is the European Centre for Medium-Range Weather Forecasts–Deutsches Klimarechenzentrum: Hamburg (ECHAM4) GCM as run for a 5-yr period with a fully coupled sulfur chemistry–cloud scheme by . Two experiments—one with an annual mean sulfate load of 0.36 Tg S for the preindustrial simulation and one with 1.05 Tg S for the present day simulation were studied. The goal was to confirm indirectly the existence of the indirect aerosol effect by finding indices for a better agreement of observations with the present-day experiment as compared with the preindustrial experiment. The authors were able to draw such a conclusion only for the German data but not for the United States. The model correctly predicts the annual mean total cloud cover in Germany and the United States, whereas global solar radiation is underestimated by 13 W m−2. This deficiency stems from cloudy conditions. Clouds are either optically too thick or the vertical distribution of clouds is erroneous. This is confirmed by the modeled overcast solar irradiance, which is 27 W m−2 lower than observed, whereas, for the clear sky, model and observations agree. Precipitation rates are underestimated by 42% in the United States. The seasonal cycle of the precipitation rate is incorrect in all U.S. regions. The modeled cloud cover is too low over the central United States in July and August, and consequently the solar irradiance exceeds the observations during these months. The opposite occurs in winter, when the model overestimates the cloud cover and thus underestimates solar irradiance. The nonseasonality of vegetation and soil parameters is suggested as a possible cause for these deficiencies. The convective precipitation formation might also contribute to these discrepancies. On the other hand, this drying out effect of the inner continent is not as pronounced in coastal regions, and, in particular, the comparisons for the German grid box provide indications for the validity of the indirect aerosol effect. The modeled annual cloud cover and solar radiation cycles for the present-day aerosol load are in better agreement with observations. Furthermore, the model shows an interesting shift from low-cloud reduction to cirrus formation in spring as a consequence of the indirect aerosol effect, a result that is confirmed by observational data.

Development of a Cirrus Parameterization Scheme: Performance Studies in HIRLAM

Abstract A parameterization scheme for the formation and maintenance of cirrus is presented. The technique combines the description of the microphysical processes specific to ice with the observational experience of cirrus behavior. The scheme is a further development of the cloud parameterization approach presented in Sundqvist et al The performance of this cirrus scheme is tested with the High-Resolution Limited Area Model (HIRLAM). Ice crystals are assumed to form through homogeneous freezing of solution droplets, at temperatures below −38°C, provided that the relative humidity exceeds a temperature-dependent threshold for the relative humidity with respect to ice. The onset of ice crystal nucleation is followed by the relaxation of the cloud water and relative humidity toward the quasi-steady state, forced by the large-scale convergence of water vapor. Ice crystals, horizontally advected from the neighboring grid boxes, enable spreading of cirrus whenever the relative humidity exceeds saturation. Perf...

Upper‐tropospheric relative humidity observations and implications for cirrus ice nucleation

Relative humidity (RH) measurements acquired in orographic wave cloud and cirrus environments are used to investigate the temperature‐dependent RH required to nucleate ice crystals in the upper troposphere, RHnuc(T). High ice‐supersaturations in clear air—conducive to the maintenance of aircraft contrails yet below RHnuc and therefore insufficient for cirrus formation—are not uncommon. Earlier findings are supported that RHnuc in mid‐latitude, continental environments decreases from water‐saturation at temperatures above −39°C to 75% RH at −55°C. Uncertainty in determining RHnuc below −55°C results in part from size detection limitations of the microphysical instrumentation, but analysis of data from the SUCCESS experiment indicates that RHnuc below −55°C is between 70 and 88%. A small amount of data acquired off‐shore suggests the possibility that RHnuc may also depend on properties of the aerosols.

In Situ Observations of the Microphysical Properties of Young Cirrus Clouds**

Abstract In situ measurements made in cold (−35° to −60°C) cirrus clouds over southern Germany in March 1994 are presented. The clouds appeared to be in an early stage of their life cycle and their properties in many ways resemble those reported for ice fogs. Crystal concentrations were high (median 2.5 cm−3, STP) and sizes small with a diameter of mean mass of typically 16 μm. The cloud on 18 March presents an interesting case for modeling studies of cirrus formation. On that particular day, the bulk properties of the cloud appeared to be connected to wave structures in the vertical wind field consistent with the Brunt–Vaisala frequency, which gave a corresponding wavelength of 40–50 km. Furthermore, analyses of potential temperature and vertical wind suggested that the vertical displacement producing these clouds was less than 100 m. Size distribution measurements of interstitial particles and crystal residues (particles remaining after evaporation of the crystals) show that small aerosol particles (dia...

On the formation and persistence of subvisible cirrus clouds near the tropical tropopause

We have used a detailed cirrus cloud model to evaluate the physical processes responsible for the formation and persistence of subvisible cirrus near the tropical tropopause and the apparent absence of these clouds at midlatitudes. We find that two distinct formation mechanisms are viable. Energetic tropical cumulonimbus clouds transport large amounts of ice water to the upper troposphere and generate extensive cirrus outflow anvils. Ice crystals with radii larger than 10 – 20 μm should precipitate out of these anvils within a few hours, leaving behind an optically thin layer of small ice crystals (τvis ≃ 0.01 – 0.2, depending upon the initial ice crystal size distribution). Given the long lifetimes of the clouds, wind shear is probably responsible for the observed cloud thickness ≤1 km. Ice crystals can also be generated in situ by slow, synoptic scale uplift of a humid layer. Given the very low temperatures at the tropical tropopause (≃−85°C), synoptic‐scale uplift can generate the moderate ice supersaturations (less than 10%) required for homogeneous freezing of sulfuric acid aerosols. In addition, simulations suggest that relatively large ice crystal number densities should be generated (more than 0.5 cm−3). The numerous crystals cannot grow larger than about 10–20 μm given the available vapor, and their low fall velocities will allow them to remain in the narrow supersaturated region for at least a day. The absorption of infrared radiation in the thin cirrus results in heating rates on the order of a few K per day. If this energy drives local parcel temperature change, the cirrus will dissipate within several hours. However, if the absorbed radiative energy drives lifting of the cloud layer, the vertical wind speed will be about 0.2 cm‐s−1, and the cloud may persist for days with very little change in optical or microphysical properties. The fact that these clouds form most frequently over the tropical western Pacific is probably related (through the nucleation physics) to the very low tropopause temperatures in this region. Simulations using midlatitude tropopause temperatures near −65°C suggest that at the higher temperatures, fewer ice crystals nucleate, resulting in more rapid crystal growth and cloud dissipation by precipitation. Hence, the lifetime of thin cirrus formed near the midlatitude tropopause should be limited to a few hours after the synoptic‐scale system that initiated cloud formation has passed.

Dehydration of the upper troposphere and lower stratosphere by subvisible cirrus clouds near the tropical tropopause

The extreme dryness of the lower stratosphere is believed to be caused by freeze‐drying of air as it enters the stratosphere through the cold tropical tropopause. Previous investigations have been focused on dehydration occurring at the tops of deep convective cloud systems. However, recent observations of a ubiquitous stratiform cirrus cloud layer near the tropical tropopause suggest the possibility of dehydration as air is slowly lifted by large‐scale motions. In this study, we have evaluated this possibility using a detailed ice cloud model. Simulations of ice cloud formation in the temperature minima of gravity waves (wave periods of 1–2 hours) indicate that large numbers of ice crystals will likely form due to the low temperatures and rapid cooling. As a result, the crystals do not grow larger than about 10 µm, fallspeeds are no greater than a few cm‐s−1, and little or no precipitation or dehydration occurs. However, ice clouds formed by large‐scale vertical motions (with lifetimes of a day or more) should have fewer crystals and more time for crystal sedimentation to occur, resulting in water vapor depletions as large as 1 ppmv near the tropopause. We suggest that gradual lifting near the tropical tropopause, accompanied by formation of thin cirrus, may account for the dehydration.

Relative Humidity and Temperature Influences on Cirrus Formation and Evolution: Observations from Wave Clouds and FIRE II

Measurements in orographic wave clouds. and in cirrus sampled during FIRE II, are used to investigate ice nucleation in the upper troposphere. The dynamically and microphysically simpler quasi-steady-state wave clouds provide relatively ideal conditions for observing characteristics of ice nucleation. Conclusions from the wave cloud study are applied to help understand the formation and evolution of ice in the cirrus clouds observed during FIRE II. The wave cloud study extends analyses reported by Heymsfield and Miloshevich down to −56°C, in part by using an improved droplet size spectrometer with a detection threshold of 0.4 µm am a Video Ice Particle Sampler with a detection threshold of 5–10 µm. The measurements show a rapid transition from solution droplets to ice crystals characteristic of homogeneous ice nucleation throughout the temperature range from −35° to −56°C. The temperature dependence of the relative humidity and the droplet sizes when ice nucleation occurs is consistent with theoretical and experimental expectations for homogeneous freezing. An expression is given for the peak RH with respect to water in the wave clouds (RHhn), which decreases from 100% above −39°C to 73% at −56°C; RHhn represents the condition required for ice nucleation in the wave clouds and is shown to be more consistent with the homogeneous freezing of sulfuric acid solution droplets than ammonium sulfate solution droplets. Aircraft measurements made in cirrus during FIRE II show highly ice-supersaturated regions in clear air, placing a lower bound on the RH required for cirrus formation approximately equal to (RHhn–10)%. Measurements from a balloonborne Formvar ice crystal replicator are reported that show the vertical structure of cirrus generally consists of three microphysically distinct regions: a highly ice-supersaturated ice production region near cloud top, an ice-supersaturated ice crystal growth region, and a sublimation region near cloud base formed by fallout of ice into ice-subsaturated air. A negative feedback is observed, and studied numerically, between ice crystal concentration and ice-supersaturation; the RH condition for new ice nucleation in cirrus is most likely to occur near cloud top when existing ice concentrations are low. A second relationship, wherein the ice production rate depends on the RH, leads the authors to propose a conceptual model of the formation and evolution of circus clouds.