Clouds, Storms, and Global Climate - The RAMS Homepage · Chapter 12 Clouds, Storms, and Global Climate 761 calculated that 1500 to 2500 cumulonimbus “hot towers”, having nearly

Chapter 12

Clouds, Storms, and GlobalClimate

12.1. INTRODUCTION

Throughout this book we have examined clouds and storms from both a detailedobservational view and from the perspective of models that resolve cloudphysics and cloud dynamics explicitly. We have also examined the impacts ofclouds and storms on a locale level including rainfall and severe weather. In thischapter we step back and view clouds and storms from a global perspectiveand consider their impacts on the global radiation budget, on the energeticsof the tropical atmosphere, their impacts on the global hydrological cycle, andhow clouds and storms transport pollutants out of the boundary layer and intothe upper troposphere. We also examine how aerosol pollution interacts withclouds and storms to potentially alter the climate and how clouds respond to avarying climate. We end by considering how clouds are represented in generalcirculation models.

12.2. CLOUDS AND THE GLOBAL RADIATION BUDGET

The moderate climate of planet earth is largely a consequence of thehydrological cycle and the associated presence of clouds. High clouds can beviewed as greenhouse warming agents in that they reduce outgoing longwaveradiation flux. The longwave cloud radiative forcing (LWCRF) or longwaveradiative flux at the top of the atmosphere (TOA) by high clouds largely dependson their cloud top temperature. Shortwave cloud radiative forcing (SWCRF)or outgoing shortwave radiative flux at TOA is usually negative, since cloudsreflect solar radiation. The amount of reflected sunlight depends on the liquid-water paths and ice water paths of the clouds as well as particle sizes and liquidor ice phases. The sum of LWCRF and SWCRF at TOA is the net cloud radiativeforcing (NCRF). High thin clouds tend to have a positive NCRF as their albedosare generally low such that SWCRF does not offset LWCRF. Thick tropicalcirrus clouds, on the other hand, particularly the remnants of thick stratiform-anvil clouds of MCSs, exhibit a negative NCRF owing to the high albedo of theoptically thicker clouds.

759

760 PART III Clouds, Storms and Global Climate

Low clouds such as boundary layer stratocumuli and trade-wind cumulicontribute to a NCRF that is negative since they exhibit high albedo andowing to their warm cloud tops, they emit longwave radiation similar to thesurface of the earth. NCRF for tropical deep convective clouds is nearly zeroas LWCRF and SWCRF nearly cancel each other (Bretherton and Hartmann,2009). Likewise, middle-level stratus clouds tend to exhibit near zero NCRF,again owing to the canceling affects of LWCRF and SWCRF. The net globally-averaged NCRF is negative largely due to the high coverage and albedo ofmarine boundary layer stratocumuli and trade-wind cumuli.

Because clouds play such an important role in regulating the radiativebudget of the planet, there is considerable interest and debate regardinghow clouds might change with a warming planet. Will they reinforce, say,greenhouse gas warming, or provide a negative feedback? Ramanathan andCollins (1991) argue that increases in upper level optically thick anvil clouds,as sea surface temperatures (SST’s) rise, will increase planetary albedo orproduce a negative NCRF, and limit further rise in SST’s, and thus serve asa “thermostatic” to the climate system. Lindzen (1990) and Lindzen et al.(2001) argue and provide evidence that in a warming climate, convective cloudswill increase in coverage and intensity in the tropics which will result inenhanced compensating subsidence and thus warming and drying of the uppertroposphere. The warming and drying in the upper troposphere will permitlarger amounts of longwave radiation to escape to space, yielding a negativeNCRF. This is referred to as an “infrared iris” effect. Neither cloud resolvingmodeling (Tompkins and Craig, 1999) nor satellite-based observations overhigher SST regions (Hartmann and Michelsen, 2002; Rapp et al., 2005) supportthe iris hypothesis. It must be recognized that large scale circulations have amajor control over cloud properties. Any major changes in circulations likethe Hadley cell, Walker circulations, or shift in middle latitude storm tracksassociated with changes in global climate, will have a major impact on cloudproperties and, as a consequence, NCRF (e.g. Seager et al., 2007; Vecchi andSoden, 2007). Likewise changes in areal coverage and strength of subtropicalhigh pressure regions (possibly in response to alterations in Hadley cells) willhave a major impact on marine stratocumulus coverage and optical thicknessand hence NCRF. Clearly the impact of clouds on the radiative properties ofthe climate system, puts a major burden on general circulation models (GCMs)to correctly represent the detailed macrostructure and microstructure of clouds,their interactions with their immediate environments, and the need to representregional circulations that have a major control on cloud properties correctly.

12.3. HOT TOWERS AND TROPICAL CIRCULATIONS

Since the early studies of tropical convection by Riehl and Malkus (1958) over50 years ago, it is generally believed that latent heating in deep convectiveclouds fuels the equatorial, upward branch of the tropical Hadley cell. They

Chapter 12 Clouds, Storms, and Global Climate 761

calculated that 1500 to 2500 cumulonimbus “hot towers”, having nearlyundiluted updrafts with speeds of 2 to 4 m s−1. would supply enough latentheat release to drive the Hadley circulation. Riehl and Simpson (1979) repeatedthe energy budget calculations and again concluded that undiluted hot towerswere essential for driving the Hadley cell. Fierro et al. (2009) performed threedimensional simulations and observational analysis of a tropical squall line.They performed parcel trajectory analysis using both Doppler radar observedand simulated storm velocities and interpreted the results relative to the hottower hypothesis. The original Riehl and Malkus (Riehl and Malkus, 1958;Riehl and Simpson, 1979) hypothesis was based on the assumption that muchof the transport of boundary layer air into the upper troposphere was achievedby undiluted protected cores within the stronger/deeper cumulonimbi. As wenoted in Chapter 8, there is little evidence supporting the prevalence of undiluteprotected cores ascending from the boundary layer to the tropopause in thetropical marine atmosphere. Fierro et al. ’s analysis supported Zipser’s (2003)conclusion that the hot tower hypothesis should be modified to include theprevalence of heavily entrained cumulus towers, particularly below about 5 kmaltitude. Fierro et al. concluded that ice-phase related latent heating invigoratedrising parcels giving rise to a secondary maximum in updrafts at 10-11 kmlevels. Thus the latent heating associated with the ice phase offset the effectsof dilution by entrainment in the lower troposphere. They suggested that in thefuture the concept of hot towers should be redefined as “any deep convectivetower rooted on the boundary layer and topping the upper troposphere”.

The Riehl and Malkus (1958) studies recognized that much of the convectionin the tropics is organized into clusters of clouds which we now call MCSs.We have seen in Chapters 8 and 9, as well as in studies by Tao et al. (2002)and Johnson et al. (2007), that the heating profiles differ appreciably betweenisolated upright ordinary convection or isolated cumulonimbi, and MCSs.Cotton et al. (1995) estimated that the global annual contribution of MCSsto total precipitation is roughly a factor of 5 greater than that from ordinarycumulonimbi (hot towers). Because latent heating scales with precipitation,this suggests that heating associated with MCSs dominates that from ordinaryupright convection. Modeling and analysis studies by Mapes and Houze (1995),Donner et al. (2001) and Schumacher et al. (2004) all suggest that the elevatedheating associated with the stratiform-anvil circulations of MCSs has a majorimpact on tropical circulations. Donner et al. (2001) concluded that heatingprofiles associated with the mesoscale circulations of MCSs produced strongerHadley and Walker circulations, warmer upper tropospheric Tropics, andmoister Tropics than that produced by ordinary upright convection(hot towers)in simulations with the GFDL GCM. As noted by Bellon and Sobel (2010) thecharacter of the Hadley cell and the simulated intertropical convergence zone(ITCZ) depends strongly on how convection is parameterized. It is thereforeimportant to represent the relative contributions of cumulus congestus, isolated


cumulonimbi, versus MCSs, versus TCs, correctly in global simulations of theEarth’s climate.

12.4. CLOUDS AND THE GLOBAL HYDROLOGICAL CYCLE

We have seen that clouds associated with extratropical cyclones, cumulonimbi,and MCSs are major contributors to rainfall and therefore to regional hydrology.We now ask how clouds and cloud systems may respond to a changing globalclimate, particularly a warming climate in response to warming induced byanthropogenically-produced greenhouse gases, and how those responses mayalter the global hydrological cycle? This is not a simple question to answer asit requires greater skill in observing the earth’s components to the hydrologicalcycle than is presently possible, and it places major demands on global climatemodels which can only represent the major cloud contributions to globalhydrology through rather crude parameterizations.

Stephens and Ellis (2008) analyzed simulated model output data fromcoupled atmosphere/ocean models used in the IPCC Fourth Assessment Reportto diagnose the simulated response of the global hydrological cycle to a 1%increase in CO2 per year until its level doubled. As found in previous modelingstudies (e.g. Trenberth et al., 2007), atmospheric water vapor increases inresponse to the applied heating at roughly 7% K−1. This is largely a response toincreases in SST’s and following the Clausius-Clapeyron relation for saturatedair (see Chapter 2) the water vapor content of the atmosphere must increase.Many studies have shown that changes in column-mean water vapor contentfollow the expected behavior of the Clausius-Clapeyron equation (Stephens,1990; Wentz and Schabel, 2000; Trenberth et al., 2005). Of course there arelarge regional variations in column-integrated water vapor in response to globalwarming, largely due to changes in global circulations. For example, strongerwinds over the southern oceans poleward of 40 degrees south in response toglobal warming, lead to regionally greatly enhanced water vapor fluxes.

One would expect that precipitation should increase at the same globally-averaged rate (e.g. Wentz et al., 2007). But Stephens and Ellis (2008) show thatglobal precipitation increases at only roughly 1% K−1. Figure 12.1 illustrateshow the model-predicted changes in precipitation rate are much less than wouldbe expected by the increase in column-integrated water vapor content. That is,the globally-averaged change in precipitation efficiency is defined as

ε =W

P

1P

1W, (12.1)

where W and P are global mean values of column water vapor and precipitation,respectively, and δP and δW are the increased precipitation and column watervapor related to global warming. They show that ε is much less than one. This isbecause increases in global precipitation tend to track increases in radiative heatloss (see Figure 12.2). Thus as water vapor increases, the atmosphere cannot


Column water vapor

Precipitation rate

Δ Ts (K)

FIGURE 12.1 The relative changes in column water vapor amount and precipitation rate,expressed as percentage changes, as functions of global temperature change derived from theAR4 models. The change in column water vapor derived assuming the CC relationship correspondsto an increase of 7.4% K−1. The sensitivity of global precipitation rate changes to temperaturechanges is approximately 2.3% K−1. The discrepancy between these two sensitivities indicatesthat the ratio of precipitation sensitivity to water vapor sensitivity in these models must be muchless than unity. Note: Not all models had both column water vapor and precipitation data. (FromStephens and Ellis (2008).)

admit radiation at a large enough rate to permit precipitation increases at thesame rate as water vapor increases. The change in the increase of clear skyemission associated with increases in column-integrated water vapor dominatesthe change in the energy balance of the atmosphere and ε. Nonetheless, anegative cloud radiative feedback occurs through reductions in cloud amountin the middle troposphere. The reduction of middle troposphere cloud amountexposes the warmer atmosphere below to high clouds resulting in a net warmingof the atmospheric column by clouds. This leads to a negative feedback onprecipitation. Unfortunately, as will be discussed later, we are not confidentthat GCMs can represent changes in cloudiness in response to global warming,realistically. This could account for the fact that observationally-based studies(Gu et al., 2007; Allan and Soden, 2007; Zhang et al., 2007; Wentz et al., 2007)infer ε values closer to unity. Stephens and Ellis (2008) however point out thelimitations of these observational studies. Thus we return to our introductorycomments that such analysis requires greater skill in observing the earth’scomponents to the hydrological cycle than is presently possible and it placesmajor demands on global climate models which can only represent the majorcloud contributions to global hydrology through rather crude parameterizations.


–Δ Rnet (W m–2)

L Δ

P (

W m

–2)

FIGURE 12.2 The relationship between changes in latent heating (L1P) vs. changes inatmospheric column cooling (1Rnet) for the AR4 models. The dotted line represents thelinear relationship between the two quantities, and the offset between that line and the solid linerepresenting a one-to-one correspondence reflects the contribution of sensible heating to the energybalance. (From Stephens and Ellis (2008).)

12.5. CLOUD VENTING

“Cloud venting” refers to the process of transporting gaseous matter andaerosols from the lower troposphere into the middle and upper troposphere(Ching, 1982). In this section we review global estimates of cloud venting. Forthe most part cloud venting studies have been restricted to a few observationalcases and the use of parameterized or cloud resolving models to make estimatesof venting by convective clouds. The only global estimates that we are awareof were made by Cotton et al. (1995). They used the archived results of cloudresolving model simulations using RAMS to make estimates of cloud ventingrates for various cloud types. They found that the characteristic extratropicalcyclone exhibited the highest boundary layer mass flux of all the cloud systemsconsidered. In terms of annual boundary layer mass flux, extratropical cyclonesstill contribute the most. Owing to their great numbers, MCSs and ordinarythunderstorms rank second and third, respectively, to the total boundary layer airbeing vented. This is followed by TCs and MCCs which, while they vent largeamounts of boundary layer air per storm event, are much fewer in numbers. Theestimated total annual boundary layer mass flux of 4.95 × 1019 kg by all thesecloud systems represents a venting of the entire boundary layer about 90 timesper year.


12.6. AEROSOL POLLUTION IMPACTS ON GLOBAL CLIMATE

As we have seen in earlier chapters, aerosol pollution can impact clouds ina number of ways, and if the impact on clouds occurs over a large enougharea, aerosol pollution can impact global climate. Aerosols can impact climatedirectly by absorbing and reflecting solar radiation. They can also impactclimate indirectly by increasing the albedo of clouds, the so-called Twomeyeffect, and also by altering the precipitation process which can potentiallyimpact cloud lifetimes and optical thicknesses. The latter is sometimes referredto as the second indirect effect or Albrecht effect. While GCMs treat theseprocesses as if they are two independent processes, they are, in fact, intimatelycoupled. We have seen that changes in precipitation rates associated withpollution aerosols can result in very nonlinear responses including invigorationof entrainment processes, changes in boundary layer stability, and cloud andstorm propagation changes by alteration of low-level cold-pools. Virtually noneof these nonlinear responses are represented in global climate models.

Rough estimates of the global mean magnitude of the Twomey (or cloudalbedo) effect since pre-industrial times lies between −0.5 and −1.9 W m−2

from different climate models, whereas the cloud lifetime effect is estimatedto be between −0.3 and −1.4 W m−2. Kristjansson (2002) and Williams et al.(2001) concluded that the Twomey effect at the top-of-the atmosphere is fourtimes as important as the cloud lifetime effect whereas Ghan et al. (2001) andQuaas et al. (2004) simulated a cloud lifetime effect that is larger than theTwomey effect.

A few GCMs have been used to examine the impacts of aerosol acting asice nuclei (IN). Lohmann (2002) found that increases in contact IN result inmore frequent glaciation of clouds and increase the amount of precipitation viathe ice phase. This effect can offset, in part at least, the solar indirect aerosoleffect (Twomey effect) on water clouds and oppose the suppression of drizzleby enhanced CCN concentrations.

Several GCMs have simulated changes in the general circulation, whichthen affects precipitation over large areas (Rotstayn et al., 2000; Williamset al., 2001; Rotstayn and Lohmann, 2002). These models were coupled toan ocean mixed-layer model so that enhanced cloud albedo produced lowerocean surface temperatures in the northern hemisphere. In addition, suppressedrainfall resulted in more extensive cloud cover that also contributed to coolerocean surfaces. The models responded by shifting the Intertropical ConvergenceZone (ITCZ) southward, which enhanced precipitation in tropical regions of thesouthern hemisphere (Rotstayn et al., 2000) and drying in the Sahel zone inAfrica (Rotstayn and Lohmann, 2002). The latter response is consistent withthe observed reduction in rainfall in the Sahel zone during the 20th century(Giannini et al., 2003). Williams et al. (2001) also found a similar responseto both the direct and indirect effects of pollutant aerosols; in addition, theyreported a reduction in precipitation associated with the Indian monsoon duringJune, July, and August. The cooling in their model also resulted in expanded


sea-ice coverage in the Arctic Ocean in summer. This was in response to thesouthward displacement of storm tracks associated with the shift of the ITCZsouthward. Thus, the greatest impacts of enhanced aerosol concentrations wereover the north Polar regions and secondarily around 40 ◦N. However, one shouldnot interpret the results of these simulations as being quantitative forecasts of theeffects of aerosols on patterns and amounts of regional precipitation. As notedpreviously, there are many uncertaintiesin the distribution and concentrationsof aerosols in the past, and even in the present. In addition, there are manysimplifications in the models that limit their ability to realistically simulatethe indirect effects of aerosols. However, these model simulations demonstratethe potential effects of direct and indirect aerosol forcing on clouds andprecipitation in regions well beyond those directly influenced by changes inradiation produced by the aerosol.

We have seen that greenhouse warming, as a result of enhanced CO2concentrations, is only significant when greater amounts of water vapor areevaporated into the air principally over the oceans but also over land. RecentGCM simulations of greenhouse warming and direct and indirect aerosol effects(Liepert et al., 2004) suggest that the indirect and direct cooling effects ofaerosols reduce surface latent and sensible heat transfer and, as a consequence,act to lower water vapor amounts in the troposphere, and thereby substantiallyweaken the impacts of greenhouse gas warming. This is important sincemost investigators compare top of the atmosphere radiative differences forgreenhouse gas warming and aerosol direct and indirect effects separately. Sincegreenhouse warming depends on enhancement of the water vapor content ofthe atmosphere, and aerosol direct and indirect cooling reduces it, the potentialinfluence of aerosols on climate could be far more significant than previouslythought.

12.7. REPRESENTING CLOUDS IN GENERAL CIRCULATIONMODELS

12.7.1. Introduction

Representing clouds in general circulation models (GCMs) is a major challengeas they typically have a limited number of vertical levels (typically 12-15)which provides insufficient resolution to represent cirrus, middle troposphericstratus, and boundary layer clouds well. Moreover, their horizontal resolutionis generally about 100 km which means all forms of convective clouds mustbe parameterized in some way. In the first edition to this book we devotedan entire chapter to discussing large scale cloud diagnostics and cumulusparameterization. Since that time there has been a growing awareness thatcumulus parameterization schemes do not represent the behavior of manytropical weather systems properly, such as equatorially trapped waves which inGCMs propagate more like dry waves rather than slower convectively-coupledwaves. Moreover, GCMs do a rather poor job of representing intraseasonal


variability associated with monsoons which have strong deep convective forcingand convectively-coupled wave systems like the Madden-Julian oscillation.

12.7.2. Representing Boundary Layer Clouds

GCMs have been shown to be highly sensitive to the representation orparameterization of boundary layer clouds, such as marine stratocumulus cloudsand trade-wind cumuli (Bony and Dufresne, 2005). Large scale controls onboundary layer clouds include subsidence, inversion strength, and boundarylayer humidity. These are features that a GCM can crudely represent althoughinversion strength is a feature not represented well owing to their limited verticalresolution. As described in Chapter 6 the approaches to modeling boundarylayer clouds includes one-dimensional layer-averaged or mixed-layer models,entity-type or plume models, higher ordered closure models, and large-eddysimulation models. LES models require grid spacings of 50 m or finer and thusare too costly to represent boundary layer clouds in GCMs.

Most GCMs use mixed-layer models to represent stratocumuli and somecombination of plume or mass flux models and mixed layer models to simulateboth stratocumuli and trade-wind cumuli. These models are limited in theirapplication to drizzling boundary layer clouds and their impacts on boundarylayer stability and cloud organization and coverage. They are also limited intheir ability to represent aerosol influences on the cloudy boundary layer exceptfor the pure Twomey hypothesis. As noted previously, the distinction betweenthe first and second indirect aerosol effect is quite artificial and once clouddroplet concentrations are modified by varying aerosol amounts, feedbacksthrough entrainment and drizzle impacts on boundary layer stability, andcloud organization (coverage) have major impacts on cloud properties. Thesenonlinear dynamic consequences of aerosol variability are not represented well,if at all, in GCMs.

Only a few GCMs use higher-order closure models to represent boundarylayer clouds including stratocumuli and trade-wind cumuli. These modelsrequire higher vertical resolution and higher temporal resolution than mixedlayer models and thus are computationally quite expensive. They do offer thepotential for making a smooth transition from solid stratus to cumulus regimeswithout adjusting parameters, although they are unable to provide predictions ofchanges in cloud organization such as open-cell versus closed cell organization.Using a PDF approach, subgrid quantities, such as vertical velocity and LWP,are determined from prescribed basis functions in which various moments of thebasis functions are predicted in the models (Pincus and Klein, 2000; Golaz et al.,2002a,b; Larson et al., 2005). The prediction of PDFs of vertical velocity alsoprovides information for use in the activation of CCN to form cloud droplets.

In spite of their importance to global climate, GCMs still do not representboundary layer clouds well, especially in a drizzling boundary layer.


12.7.3. Representing Middle and High Clouds

GCMs represent layer clouds in the middle and upper troposphere throughexplicitly-resolved vertical motions and large-scale moistening. Unfortunatelythese cloud systems are often only a few hundred meters in depth, whilethe limited vertical resolution of GCMs smear these features over depths ofseveral kilometers. Moreover, these clouds often exhibit embedded convectionor turbulence which must be represented by some form of turbulence closuremodel either mixed layer models or higher-order closure models. Unfortunatelythese models have not benefited from the large observational data base thatis available for boundary layer clouds. Much of the effort in representingmiddle and high clouds in GCMs has focused on representing the cloudmicrophysics of mixed-phase and ice-phase clouds. This is often done withoutrepresenting the cloud-scale updrafts in those clouds which is importantin determining the nucleation of cloud particles, the sedimentation rates ofhydrometeors, and the coverage of clouds. Most of the GCMs represent warmcloud microphysics following a Kessler-type of approach to bulk microphysicsparameterization such as overviewed in Chapter 4. For ice-phase clouds, a 1970sera cloud microphysics parameterization is often followed in which, at a certaintemperature, the cloud is immediately converted into an ice saturated cloudwhere non-precipitating “cloud ice” is represented. Thus the precipitation shaftsor fall streaks that are characteristic of higher-level clouds is not represented.Moreover, the affects of varying aerosol in the middle and upper troposphere andtheir impacts of cloud radiative and hydrological properties are not representedin current generation GCMs.

12.7.4. Representing deep convective clouds

While deep convective clouds may not play a major role in controlling net cloudradiative forcing, they still are major contributors in driving tropical circulationsincluding the Walker and Hadley circulations. In the first edition of this book wedevoted an entire chapter to discussing convective parameterization schemes,including their detailed formulations and the fundamental theories and theconcepts they are based on. While there certainly have been major advancesin the formulation of convective parameterization schemes in the intervening 20years, as noted by Grabowski and Petch (2009), it has become recognized thatGCMs using convective parameterization schemes have major shortcomingsincluding:

• They typically do not represent the transition from ordinary uprightconvection to MCSs properly, if at all.• They do a poor job of simulating intraseasonal variability including

monsoonal circulations and the Madden-Julian oscillation.• They misrepresent the phase of the diurnal cycle of warm season

precipitation over the continents.


• They misrepresent the frequency and intensity of convective precipitation.• Changes in cloud-scale processes are immediately felt on the global scale

whereas in the real world responses occur on mesoscales and regional scaleswith only the residual imbalances on these scales felt by the global scales.

The latter point may be more a result of GCM resolution than convectiveparameterization schemes.

What are the options that can overcome these deficiencies of convectiveparameterization schemes? One approach is to use a cloud resolving model(CRM) embedded in GCM grid points, or what has been called “superparameterization schemes” (Grabowski and Smolarkiewicz, 1999; Grabowski,2001; Randall et al., 2003; Wyant et al., 2006), or the multiscale modelingframework (MMF) by Khairoutdinov et al. (2007). We should note that theuse of the term CRM is an abuse of the concept as these embedded modelstypically have 4-5 km grid spacings and are two-dimensional. As noted inearlier chapters, a true cloud-resolving model should have resolution of about100 m and be three-dimensional. We will use the term cloud-representing-models as CRM. According to Grabowski and Petch (2009) this approach isabout 2 to 3 orders of magnitude more computationally demanding than use ofconvective parameterization schemes. It is still not practical for use in longerterm climate simulations. One problem with this approach is that there is not anatural continuum of cloud responses from upright convection to the mesoscaleto global scales so that the coupling between the convective scales and globalscales is artificial (see Khairoutdinov et al., 2007). MCSs that form in a givengrid cell in the GCM cannot propagate into neighboring cells and, moreover,the deck is reshuffled at the end of each time step such that at the next time stepthe represented convection must undergo evolution from scratch from uprightconvection to MCSs. In other words, there is no grid point memory of previousconvective organization.

A second approach is to use a nonhydrostatic GCM with high enoughresolution to represent deep convection explicitly. We will refer to this as aGCRM. These are normally referred to as cloud-resolving GCMs but, sincethey use grid spacings of roughly 7 km and occasionally 3 km (see Collinsand Satoh, 2009), we will use the term cloud-representing GCMs in the spiritof the discussion in the previous section. The Earth Simulator (called NICAM)developed at Japan’s Frontier Research Center (Miura et al., 2007) is an exampleof a GCRM. NICAM has been able to reproduce a Madden-Julian Oscillationevent (Miura et al., 2007) and is purported to produce global climatologies closeto those observed (Iga et al., 2007). It is the experience of the lead author(Cotton) with running RAMS in realtime mesoscale forecasts over Coloradowith 3 km grid spacing for about 10 years, that during the convective season,convection is delayed until CAPE is large enough to support resolved convectionand the subsequent simulated storms are more vigorous than observed (toolittle entrainment) and produce too much precipitation. If CAPE does not buildup, then no convection forms and an under-prediction of precipitation results.


NCAM exhibits a similar over-prediction bias in precipitation in the tropics.Furthermore, according to Grabowski and Petch (2009) the use of a GCRMin NICAM is roughly 6 orders of magnitude more computationally demandingthan a conventional GCM.

Clearly we still have a long way to go in representing clouds properly inclimate models used for decadal and centuries-long simulations.

REFERENCES

Allan, R. P., and Soden, B. J. (2007). Large discrepancy between observedand simulated precipitation trends in the ascending and descendingbranches of the tropical circulation. Geophys. Res. Lett. 34, L18705,doi:10.1029/2007GL031460.

Bellon, G., and Sobel, A. H. (2010). Multiple equilibria of the Hadleycirculation in an intermediate-complexity axisymmetric model. J. Climate23, 1760–1778.

Bony, S., and Dufresne, J.-L. (2005). Marine boundary layer clouds at the heartof tropical cloud feedback uncertainties in climate models. Geophys. Res.Lett. 32, L20806, doi:10.1029/2005GL023851.

Bretherton, C. S., and Hartmann, D. L. (2009). Large-scale controlson cloudiness. In “Clouds in the Perturbed Climate System: TheirRelationship to Energy Blance, Atmospheric Dynamics, and Precipitation”(J. Heintzenberg and R. J. Charlson, Eds.), pp. 217–234. The MIT Press,Cambridge, MA.

Ching, J. (1982). The role of convective clouds in venting ozone from the mixedlayer. In “3rd Joint Conf. Applications of Air Pollution Meteorology, 12-15Jan 1982”, San Antonio, TX, Amer. Meteor. Soc. Preprints.

Collins, W. D., and Satoh, M. (2009). Simulating global clouds: Past,present and future. In “Clouds in the Perturbed Climate System: TheirRelationship to Energy Blance, Atmospheric Dynamics, and Precipitation”(J. Heintzenberg and R. J. Charlson, Eds.), pp. 469–486. The MIT Press,Cambridge, MA.

Cotton, W. R., Alexander, G. D., Hertenstein, R., Walko, R. L., McAnelly, R.L., and Nicholls, M. (1995). Cloud Venting–Areview and some new globalannual estimates. Earth Sci. Rev. 39, 169–206.

Donner, L. J., Seman, C. J., Hemler, R. S., and Fan, S. (2001). A cumulusparameterization including mass fluxes, convective vertical velocities, andmesoscale effects: Thermodynamic and hydrological aspects in a generalcirculation model. J. Climate 14, 3444–3463.

Fierro, Alexandre O., Simpson, Joanne, LeMone, Margaret A., Straka, JerryM., and Smull, Bradley F. (2009). On how hot towers fuel the Hadley cell:An observational and modeling study of line-organized convection in theequatorial trough from TOGA COARE. J. Atmos. Sci. 66, 2730–2746.


Ghan, S. J., Easter, R. C., Chapman, E. G., Abdul-Razzak, H., Zhang, Y., Leung,L. R., Laulainen, N., Saylor, R. D., and Zaveri, R. A. (2001). A physicallybased estimate of forcing by anthropogenic sulfate aerosol. J. Geophys. Res.106, 5279–5293.

Giannini, A., Saravanan, R., and Chang, P. (2003). Oceanic forcing of Sahelrainfall on interannual to interdecadal time scales. Science 302, 1027–1030,doi:10.1126/science.1089357.

Golaz, J.-C., Larson, V. E., and Cotton, W. R. (2002a). A PDF-based modelfor boundary layer. Part I: Method and model description. J. Atmos. Sci. 59,3540–3551.

Golaz, J.-C., Larson, V. E., and Cotton, W. R. (2002b). A PDF-based model forboundary layer clouds. Part II: Model results. J. Atmos. Sci. 59, 3552–3571.

Grabowski, W. W., and Smolarkiewicz, P. K. (1999). CRCP: A cloudresolving convection parameterization for modeling the tropical convectingatmosphere. Physica D 133, 171–178.

Grabowski, W. W., and Petch, J. C. (2009). Deep convective clouds. In“Clouds in the Perturbed Climate System: Their Relationship to EnergyBlance, Atmospheric Dynamics, and Precipitation” (J. Heintzenberg and R.J. Charlson, Eds.), pp. 197–215. The MIT Press, Cambridge, MA.

Grabowski, W. (2001). Coupling cloud processes with the large-scale dynamicsusing the cloud-resolving convective parameterization (CRCP). J. Atmos.Sci. 58, 978–997.

Gu, G., Adler, R. F., Huffman, G. J., and Curtis, S. (2007). Tropical rainfallvariability on interannual-to-interdecadal and longer time scales derivedfrom the GPCP monthly product. J. Climate 20, 4033–4046.

Hartmann, D. L., and Michelsen, M. L. (2002). No evidence for iris. Bull. Amer.Meteor. Soc. 83, 249–254.

Iga, S., Tomita, H., Tsushima, Y., and Satoh, M. (2007). Climatology of anonhydrostatic global model with explicit cloud processes. Geophys. Res.Lett. 34, L22814.

Johnson, D. E., Tao, W.-K., Simpson, J., and Sui, C.-H. (2007). A study ofthe response of deep tropical clouds to large-scale thermodynamic forecings.Part I: Modeling strategies and simulations of TOGA COARE convectivesystems. J. Atmos. Sci. 59, 3492–3518.

Khairoutdinov, M., DeMott, C., and Randall, D. (2007). Evaluation ofthe simulated interannual and subseasonal variability in an AMIP-stylesimulation using the CSU multi-sclae modeling framework. J. Climate 21,413.

Kristjansson, J. E. (2002). Studies of the aerosol indirect effect from sulfate andblack carbon aerosols. J. Geophys. Res. 107, doi:10.1029/2001JD000887.

Larson, V. E., Golaz, J.-C., Jiang, H., and Cotton, W. R. (2005). Supplying localmicrophysics parameterizations with information about subgrid variability:Latin hypercube sampling. J. Atmos. Sci. 62, 4010–4026.


Liepert, B. G., Feichter, J., Lohmann, U., and Roeckner, E. (2004). Can aerosolsspin down the watercycle in a warmer and moister world? Geophys. Res.Lett. 31, doi:10.1029/2003GL019060.

Lindzen, R. S. (1990). Some coolness concerning global warming. Bull. Amer.Meteor. Soc. 71, 288–299.

Lindzen, R. S., Chou, M.-D., and Hou, A. Y. (2001). Does the earth have anadaptive infrared iris? Bull. Amer. Meteor. Soc. 82, 417–432.

Lohmann, U. (2002). A glaciation indirect aerosol effect caused by sootaerosols. Geophys. Res. Lett. 29, doi:10.1029/2001GL014357.

Mapes, B. E., and Houze Jr., R. A. (1995). Diabatic divergence profiles inwestern Pacific mesoscale convective systems. J. Atmos. Sci. 52, 1807–1828.

Miura, H., Satoh, M., Nasuno, T., Noda, A. T., and Oouchi, K. (2007). AMadden-Julian Oscillation event simulation by a global cloud-resolvingmodel. Science 318, 1763–1765.

Pincus, R., and Klein, S. A. (2000). Unresolved spatial variability andmicrophysical process rates in large-scale models. J. Geophys. Res. 105,27059–27065.

Quaas, J., Boucher, O., and Breon, F.-M. (2004). Aerosol indirecteffects in POLDER satellite data and the Laboratoire de MeteorologieDynamique-Zoom (LMDZ) general circulation model. J. Geophys. Res. 10,doi:10.1029/2003JD004317.

Ramanathan, V., and Collins, W. (1991). Thermodynamic regulation of oceanwarming by cirrus clouds deduced from observationsl of the 1987 El Nino.Nature 351, 27–32, doi:10.1038/351027a0.

Randall, D., Khairoutdinov, M., Arakawa, A., and Grabowski, W. (2003).Breaking the cloud parameterization deadlock. Bull. Amer. Meteor. Soc. 84,1547–1564, doi:10.1175/BAMS-84-11-1547.

Rapp, A. D., Kummerow, C., Berg, W., and Griffith, B. (2005). An evaluationof the proposed mechanism of the adaptive infrared Iris hypothesis usingTRIMM VIRS and PR measurements. J. Climate 18, 4185–4194.

Riehl, H., and Malkus, J. S. (1958). On the heat balance in the equatorial troughzone. Geophysica 6, 503–538.

Riehl, H., and Simpson, J. S. (1979). On the heat balance in the equatorial troughzone, revisited. Contrib. Atmos. Phys. 52, 287–305.

Rotstayn, L. D., and Lohmann, U. (2002). Tropical rainfall trends and theindirect aerosol effect. J. Clim. 15, 2103–2116.

Rotstayn, L. D., Ryan, B. F., and Penner, J. E. (2000). Precipitation changes in aGCM resulting from the indirect effects of anthropogenic aerosols. Geophys.Res. Lett. 27, 3045–3048.

Schumacher, C., Houze Jr., R. A., and Kraucunas, I. (2004). The tropicaldynamical response to latent heating estimates derived from the TRMMprecipitation radar. J. Atmos. Sci. 61, 1341–1358.

Seager, R., and coauthors, (2007). Model projections of an imminent transitionto a more arid climate in southwestern North America. Science 316,1181–1184, doi:10.1126/science.1139601.


Stephens, G. L. (1990). On the relationship between water vapor over the oceansand sea surface temperature. J. Climate 3, 634–645.

Stephens, G. L., and Ellis, T. D. (2008). Controls of global-mean precipitationincreases in global warming GCM experiments. J. Climate 21, 6141–6155.

Tao, W.-K., Lang, S., Olson, W. S., Meneghini, R., Yang, S., Simpson, J.,Kummerow, C., Smith, E., and Halverson, J. (2002). Retrieved verticalprofiles of latent heat release using TRMM rainfall products for February1988. J. Appl. Meteor. 40, 957–982.

Tompkins, A. M., and Craig, G. C. (1999). Sensitivity of tropical convectionto sea surface temperature in the absence of large-scale flow. J. Climate 12,462–476.

Trenberth, K. E., Fasullo, J., and Smith, L. (2005). Trends and variabilityin column-integrated atmospheric water vapor. Climate Dyn. 24, 741–758,doi:10.1007/s00382-005-0017-4.

Trenberth, K. E., and coauthors, (2007). Observations: Surface and atmosphericclimate change. In “Climate Change 2007: The Physical Science Basis” (S.Solomon et al., Eds.), pp. 235–336. Cambridge University Press.

Vecchi, G. A., and Soden, B. J. (2007). Global warming and the weakening ofthe tropical circulation. J. Climate 20, 4316–4340.

Wentz, F. J., and Schabel, M. (2000). Precise climate monitoring usingcomplementary satellite data sets. Nature 403, 414–416.

Wentz, F. J., Ricciardulli, L., Hilburn, K., and Mears, C. (2007). Howmuch more rain will global warming bring? Science 317, 233–235,doi:10.1126/science.1140746.

Williams, K. D., Jones, A., Roberts, D. L., Senior, C. A., and Woodage, M.J. (2001). The response of the climate system to the indirect effects ofanthropogenic sulfate aerosol. Clim. Dyn. 17, 845–856.

Wyant, M. C., Khairoutdinov, M., and Bretherton, C. S. (2006). Climatesensitivity and cloud response of a GCM with a superparameterization.Geophy. Res. Lett. 33, L06714.

Zhang, X., Zwiers, F. W., Hegerl, G. C., Lambert, F. H., Gillett, N. P.,Solomon, S., Stott, P. A., and Nozawa, T. (2007). Detection of humaninfluence on twentieth-century precipitation trends. Nature 448, 461–465,doi:10.1038/nature06025.

Zipser, E. J. (2003). Some views on “Hot Towers” after 50 years of tropical fieldprograms and two years of TRMM data. Meteor. Monogr. 29, 49.

Clouds, Storms, and Global Climate - The RAMS Homepage · Chapter 12 Clouds, Storms, and Global Climate 761 calculated that 1500 to 2500 cumulonimbus “hot towers”, having nearly

Documents