• Albrecht, B., 1989: Aerosols, cloud microphysics, and fractional cloudiness. Science, 245 , 12271230.

  • Beheng, K. D., 1994: A parameterization of warm cloud microphysical conversion processes. Atmos. Res., 33 , 193206.

  • Bolton, D., 1980: The computation of equivalent potential temperature. Mon. Wea. Rev., 108 , 10461053.

  • Bott, A., 1989: A positive definite advection scheme obtained by nonlinear renormalization of the advective fluxes. Mon. Wea. Rev., 117 , 10061015.

    • Search Google Scholar
    • Export Citation
  • Chaumerliac, N., E. Richard, and J. Pinty, 1987: Sulfur scavenging in a mesoscale model with quasi-spectral microphysics: two-dimensional results for continental and maritime clouds. J. Geophys. Res., 92 , 31143126.

    • Search Google Scholar
    • Export Citation
  • Clark, T. L., and R. Farley, 1984: Severe downslope windstorm calculations in two and three spatial dimensions using anelastic interactive grid nesting: A possible mechanism for gustiness. J. Atmos. Sci., 41 , 329350.

    • Search Google Scholar
    • Export Citation
  • Colle, B. A., 2004: Sensitivity of orographic precipitation to changing ambient conditions and terrain geometries: An idealized modeling perspective. J. Atmos. Sci., 61 , 588606.

    • Search Google Scholar
    • Export Citation
  • Cozic, J., B. Verheggen, S. Mertes, P. Connolly, K. Bower, A. Petzold, U. Baltensperger, and E. Weingartner, 2007: Scavenging of black carbon in mixed phase clouds at the high alpine site Jungfraujoch. Atmos. Chem. Phys., 7 , 17971807.

    • Search Google Scholar
    • Export Citation
  • Cozic, J., and Coauthors, 2008: Chemical composition of free tropospheric aerosol for PM1 and coarse mode at the high alpine site Jungfraujoch. Atmos. Chem. Phys., 8 , 407423.

    • Search Google Scholar
    • Export Citation
  • Cziczo, D. J., D. Murphy, P. Hudson, and D. Thomson, 2004: Single-particle measurements of the chemical composition of cirrus ice residue during CRYSTAL-FACE. J. Geophys. Res., 109 .D04201, doi:10.1029/2003JD004032.

    • Search Google Scholar
    • Export Citation
  • Davies, H., 1976: A lateral boundary formulation for multi-level prediction models. Quart. J. Roy. Meteor. Soc., 102 , 405418.

  • Denman, K. L., and Coauthors, 2007: Couplings between changes in the climate system and biogeochemistry. Climate Change 2007: The Physical Science Basis, S. Soloman et al., Eds., Cambridge University Press, 499–588.

    • Search Google Scholar
    • Export Citation
  • Doms, G., and U. Schättler, 2002: A description of the nonhydrostatic regional model LM. Part I: Dynamics and numerics. Deutscher Wetterdienst Tech. Rep., 134 pp.

  • Durran, D., and J. Klemp, 1983: A compressible model for the simulation of moist mountain waves. Mon. Wea. Rev., 111 , 23412361.

  • Ferek, R. J., D. A. Hegg, P. V. Hobbs, P. Durkee, and K. Nielsen, 1998: Measurements of ship-induced tracks in clouds off the Washington coast. J. Geophys. Res., 103 , D18. 2319923206.

    • Search Google Scholar
    • Export Citation
  • Fuhrer, O., and C. Schär, 2005: Embedded cellular convection in moist flow past topography. J. Atmos. Sci., 62 , 28102828.

  • Ghan, S., C. C. Chuang, and J. E. Penner, 1993: A parameterization of cloud droplet nucleation. Part I: Single aerosol type. Atmos. Res., 30 , 197221.

    • Search Google Scholar
    • Export Citation
  • Givati, A., and D. Rosenfeld, 2004: Quantifying precipitation suppression due to air pollution. J. Appl. Meteor., 43 , 10381056.

  • Givati, A., and D. Rosenfeld, 2005: Separation between cloud-seeding and air pollution effects. J. Appl. Meteor., 44 , 12981314.

  • Herzog, H-J., G. Vogel, and U. Schubert, 2002: LLM—A nonhydrostatic model applied to high-resolving simulations of turbulent fluxes over heterogeneous terrain. Theor. Appl. Climatol., 73 , 6786.

    • Search Google Scholar
    • Export Citation
  • Hobbs, P., R. Easter, and A. Fraser, 1973: A theoretical study of the flow of air and fallout of solid precipitation over mountainous terrain: Part II. Microphysics. J. Atmos. Sci., 30 , 813823.

    • Search Google Scholar
    • Export Citation
  • Houze, R., and S. Medina, 2005: Turbulence as a mechanism for orographic precipitation enhancement. J. Atmos. Sci., 62 , 35993623.

  • Hueglin, C., R. Gehrig, U. Baltensperger, M. Gysel, C. Monn, and H. Vonmont, 2005: Chemical characterisation of PM2.5, PM10 and coarse particles at urban, near-city and rural sites at Switzerland. Atmos. Environ., 39 , 637651.

    • Search Google Scholar
    • Export Citation
  • Jaenicke, R., 1993: Tropospheric aerosols. Aerosol–Cloud–Climate Interactions, P. V. Hobbs, Ed., Academic Press, 1–31.

  • Jiang, Q., 2003: Moist dynamics and orographic precipitation. Tellus, 55A , 301316.

  • Jiang, Q., 2006: Precipitation over concave terrain. J. Atmos. Sci., 63 , 22692288.

  • Jiang, Q., 2007: Precipitation over multiscale terrain. Tellus, 59A , 321335.

  • Jiang, Q., and R. Smith, 2003: Cloud timescales and orographic precipitation. J. Atmos. Sci., 60 , 15431559.

  • Jirak, I. L., and W. R. Cotton, 2006: Effect of air pollution on precipitation along the front range of the Rocky Mountains. J. Appl. Meteor., 45 , 236245.

    • Search Google Scholar
    • Export Citation
  • Khairoutdinov, M., and Y. Kogan, 2000: A new cloud physics parameterization in a large-eddy simulation model of marine stratocumulus. Mon. Wea. Rev., 128 , 229243.

    • Search Google Scholar
    • Export Citation
  • Kirshbaum, D. J., and D. R. Durran, 2004: Factors governing cellular convection in orographic precipitation. J. Atmos. Sci., 61 , 682698.

    • Search Google Scholar
    • Export Citation
  • Leaitch, W., C. Banic, G. Isaac, M. Couture, P. Liu, I. Gultepe, and S-M. Li, 1996: Physical and chemical observations in marine stratus during the 1993 North Atlantic Regional Experiment: Factors controlling cloud droplet number concentrations. J. Geophys. Res., 101 , D22. 2912329135.

    • Search Google Scholar
    • Export Citation
  • Lin, H., and R. Leaitch, 1997: Development of an in-cloud aerosol activation parameterization for climate modelling. Proc. Workshop on Measurements of Cloud Properties for Forecasts of Weather and Climate, Mexico City, Mexico, WMO, 328–335.

    • Search Google Scholar
    • Export Citation
  • Lohmann, U., 2002: Possible aerosol effects on ice clouds via contact nucleation. J. Atmos. Sci., 59 , 647656.

  • Lohmann, U., and B. Kärcher, 2002: First interactive simulations of cirrus clouds formed by homogeneous freezing in the ECHAM general circulation model. J. Geophys. Res., 107 .4105, doi:10.1029/ 2001JD000767.

    • Search Google Scholar
    • Export Citation
  • Lohmann, U., and J. Feichter, 2005: Global indirect aerosol effects: A review. Atmos. Chem. Phys., 4 , 75617614.

  • Lohmann, U., P. Stier, C. Hoose, S. Ferrachat, S. Kloster, E. Roeckner, and J. Zhang, 2007: Cloud microphysics and aerosol indirect effects in the global climate model ECHAM5-HAM. Atmos. Chem. Phys., 7 , 34253446.

    • Search Google Scholar
    • Export Citation
  • Miglietta, M. M., and R. Rotunno, 2005: Simulations of moist nearly neutral flow over a ridge. J. Atmos. Sci., 62 , 14101427.

  • Murphy, D. M., and Coauthors, 2006: Single-particle mass spectrometry of tropospheric aerosol particles. J. Geophys. Res., 111 .D23S32, doi:10.1029/2006JD007340.

    • Search Google Scholar
    • Export Citation
  • Peng, Y., U. Lohmann, R. Leaitch, C. Banic, and M. Couture, 2002: The cloud albedo–cloud droplet effective radius relationship for clean and polluted clouds from RACE and FIRE.ACE. J. Geophys. Res., 107 .4106, doi:10.1029/2000JD000281.

    • Search Google Scholar
    • Export Citation
  • Pinsky, M., A. Khain, and M. Shapiro, 2001: Collision efficiency of drops in a wide range of Reynolds numbers: Effect of pressure on spectrum evolution. J. Atmos. Sci., 58 , 742764.

    • Search Google Scholar
    • Export Citation
  • Pruppacher, H., and J. Klett, 1997: Microphysics of Clouds and Precipitation. 2nd ed. Kluwer Academic, 954 pp.

  • Roe, G. H., 2005: Orographic precipitation. Annu. Rev. Earth Planet. Sci., 33 , 645671.

  • Rosenfeld, D., and A. Givati, 2006: Evidence of orographic precipitation suppression by air pollution–induced aerosols in the western united states. J. Appl. Meteor., 45 , 893911.

    • Search Google Scholar
    • Export Citation
  • Rosenfeld, D., J. Dai, X. Yu, Z. Yao, X. Xu, X. Yang, and C. Du, 2007: Inverse relations between amounts of air pollution and orographic precipitation. Science, 315 , 396398.

    • Search Google Scholar
    • Export Citation
  • Rotunno, R., and R. A. Houze, 2007: Lessons on orographic precipitation from the Mesoscale Alpine Programme. Quart. J. Roy. Meteor. Soc., 133 , 811830.

    • Search Google Scholar
    • Export Citation
  • Seifert, A., and K. D. Beheng, 2006: A two-moment cloud microphysics parameterization for mixed-phase clouds. Part I: Model description. Meteor. Atmos. Phys., 92 , 4566.

    • Search Google Scholar
    • Export Citation
  • Seinfeld, J., and S. Pandis, 1998: Atmospheric Chemistry and Physics. John Wiley, 1326 pp.

  • Smith, R., 1979: The influence of mountains on the atmosphere. Advances in Geophysics, Vol. 21, Academic Press, 87–230.

  • Smith, R., 1980: Linear theory of stratified hydrostatic flow past an isolated mountain. Tellus, 32 , 348364.

  • Smith, R., 1989a: Hydrostatic airflow over mountains. Advances in Geophysics, Vol. 31, 1–41.

  • Smith, R., 1989b: Mechanisms of orographic precipitation. Meteor. Mag., 119 , 8588.

  • Smith, R., 2003: A linear upslope time-delay model for orographic precipitation. J. Hydrol., 282 , 29.

  • Smith, R., and I. Barstad, 2004: A linear theory of orographic precipitation. J. Atmos. Sci., 61 , 13771391.

  • Smolarkiewicz, P., R. Rasmussen, and T. Clark, 1988: On the dynamics of hawaiian cloud bands: Island forcing. J. Atmos. Sci., 45 , 18721905.

    • Search Google Scholar
    • Export Citation
  • Solomon, S., and Coauthors, 2007: Climate Change 2007: The Physical Science Basis. Cambridge University Press, 996 pp.

  • Spichtinger, P., 2004: Eisübersättigte Regionen. Ph.D. thesis, Deutsches Zentrum für Luft- und Raumfahrt, 208 pp.

  • Steppeler, J., G. Doms, U. Schättler, H. Bitzer, A. Gassmann, U. Damrath, and G. Gregoric, 2003: Meso-gamma scale forecasts using the nonhydrostatic model LM. Meteor. Atmos. Phys., 82 , 7596.

    • Search Google Scholar
    • Export Citation
  • Stier, P., and Coauthors, 2005: The aerosol–climate model ECHAM5-HAM. Atmos. Chem. Phys., 5 , 11251156.

  • Twomey, S., M. Piepgrass, and T. Wolfe, 1984: An assessment of the impact of pollution on global cloud albedo. Tellus, 36B , 356366.

  • Vehameki, H., M. Kulmala, I. Napari, K. Lehtinen, C. Timmreck, M. Noppel, and A. Laaksonen, 2002: An improved parameterization for sulfuric acid–water nucleation rates for tropospheric and stratospheric conditions. J. Geophys. Res., 107 .4622, doi:10.1029/2002JD002184.

    • Search Google Scholar
    • Export Citation
  • Verheggen, B., and Coauthors, 2007: Aerosol partitioning between the interstitial and the condensed phase in mixed-phase clouds. J. Geophys. Res., D23202, doi:10.1029/2007JD008714.

    • Search Google Scholar
    • Export Citation
  • Vignati, E., J. Wilson, and P. Stier, 2004: M7: An efficient size-resolved aerosol microphysics module for large-scale aerosol transport models. J. Geophys. Res., 109 .D22202, doi:10.1029/2003JD004485.

    • Search Google Scholar
    • Export Citation
  • Weingartner, E., S. Nyeki, and U. Baltensperger, 1999: Seasonal and diurnal variation of aerosol size distributions (10 < d < 750 nm) at a high-alpine site (Jungfraujoch 3580 m asl). J. Geophys. Res., 104 , D21. 2680926820.

    • Search Google Scholar
    • Export Citation
  • Weingartner, E., M. Gysel, and U. Baltensperger, 2002: Hygroscopicity of aerosol particles at low temperatures. 1. New low-temperature h-tdma instrument: Setup and first applications. Environ. Sci. Technol., 36 , 5562.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Vertical profiles of temperature (solid thick line) and dewpoint temperature (dashed thick line). The parameters describing the analytically given profiles are Nd = 0.0115 s−1, TSL = 285 K, pSL = 1000 hPa, and RHSL = 0.8. The lifting condensation level (LCL) is located at approximately zLCL ≈ 425 m.

  • View in gallery

    Initial aerosol size spectra for the clean case (solid lines) and polluted case (dashed lines) at the surface. Size distributions for (a) the number density (cm−3) and (b) the mass distributions (μg m−3).

  • View in gallery

    (a) Initial profiles of the gas-phase sulfuric acid concentration, (b) the number density of sulfate aerosols in nucleation mode, and (c) the mass density of nucleation mode sulfate aerosols (c). Note that the number and mass distributions of the gas-phase sulfuric acid and the nucleation mode aerosols are all in (molecules/particles) per cubic centimeter.

  • View in gallery

    Potential temperature θ (K; contour interval 1 K), equivalent potential temperature θe (K; contour interval 1 K) and horizontal wind speed U (m s−1; contour interval 1 m s−1) for (a), (b) the dry simulation and (c), (d) the moist simulation in experiment A after 10 h. Only part of the computational domain is shown.

  • View in gallery

    Vertical cross section of (a), (b) the cloud water mixing ratio (g kg−1; contour interval 0.05 g kg−1) and (c), (d) for the rain water mixing ratio (g kg−1; contour interval 0.02 g kg−1) for experiment A after 10 h. The clean case simulation is shown in the left column and the polluted case simulation is shown in the right column. Only part of the computational domain is shown.

  • View in gallery

    The precipitation distribution [mm (10 h)−1] for experiment A for (a) the narrow mountain with a = 10 km and (b) the wide mountain with a = 30 km after 10 h, for both the clean case (solid) and the polluted case (dashed). The topographic surface contour for (c) the narrow mountain and (d) the wide mountain. The location of the mountain peak is indicated by the triangle.

  • View in gallery

    Difference in the precipitation distribution [polluted case minus clean case; mm (10 h)−1] along the idealized topography after 10 h for experiment A and different mountain half-widths. The location of the mountain peak is indicated by the triangle.

  • View in gallery

    Difference fields for experiment A after 10 h. The differences (polluted case − clean case) for (a) the water vapor mixing ratio qυ (contour interval 0.04 g kg−1), (b) the potential temperature θ (contour interval 0.1 K), (c) the horizontal wind speed U (contour interval 0.1 m s−1) and (d) the vertical velocity (contour interval 0.02 m s−1). Only part of the domain is plotted.

  • View in gallery

    For experiment B after 10 h, the (a) potential temperature θ (K) and (b) horizontal velocity U (m s−1) for the dry simulation; (c) equivalent potential temperature θe (K) and (d) U (m s−1) for the moist simulation. The contour intervals are (a), (c) 1 K and (b), (d) 2 m s−1. Only part of the computational domain is shown.

  • View in gallery

    Same as Fig. 5, but for experiment B after 10 h and a rain water mixing ratio contour interval of 0.002 g kg−1.

  • View in gallery

    Same as in Fig. 7 but for experiment B.

  • View in gallery

    Equivalent potential temperature field θe (K; contour interval 1 K) and horizontal velocity U (m s−1; contour interval 2 m s−1) in experiment C after 10 h for (a), (b) the clean case and (c), (d) the polluted case simulation. The thick contour indicates zero wind speed. Only part of the computational domain is shown.

  • View in gallery

    Same as in Fig. 5, but for experiment C after 10 h and for a cloud water mixing ratio interval of 0.1 g kg−1 and a rain water mixing ratio of 0.05 g kg−1. Only part of the computational domain is shown.

  • View in gallery

    Same as in Fig. 6 but for experiment C.

  • View in gallery

    Same as in Fig. 7 but for experiment C.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 113 81 10
PDF Downloads 74 47 6

Sensitivity Studies of the Role of Aerosols in Warm-Phase Orographic Precipitation in Different Dynamical Flow Regimes

View More View Less
  • 1 Institute for Atmospheric and Climate Science, ETH Zurich, Zurich, Switzerland
© Get Permissions
Full access

Abstract

Aerosols serve as a source of cloud condensation nuclei (CCN) and influence the microphysical properties of clouds. In the case of orographic clouds, it is suspected that aerosol–cloud interactions reduce the amount of precipitation on the upslope side of the mountain and enhance the precipitation on the downslope side when the number of aerosols is increased. The net effect may lead to a shift of the precipitation distribution toward the leeward side of mountain ranges, which affects the hydrological cycle on the local scale.

In this study aerosol–cloud interactions in warm-phase clouds and the possible impact on the orographic precipitation distribution are investigated. Herein, simulations of moist orographic flow over topography are conducted and the influence of anthropogenic aerosols on the orographic precipitation formation is analyzed. The degree of aerosol pollution is prescribed by different aerosol spectra that are representative for central Switzerland. The simulations are performed with the Consortium for Small-Scale Modeling’s mesoscale nonhydrostatic limited-area weather prediction model (COSMO) with a horizontal grid spacing of 2 km and a fully coupled aerosol–cloud parameterization.

It is found that an increase in the aerosol load leads to a downstream shift of the orographic precipitation distribution and to an increase in the spillover factor. A reduction of warm-phase orographic precipitation is observed at the upslope side of the mountain. The downslope precipitation enhancement depends critically on the width of the mountain and on the flow dynamics. In the case of orographic precipitation induced by stably stratified unblocked flow, the loss in upslope precipitation is not compensated by leeward precipitation enhancement. In contrast, flow blocking may lead to leeward precipitation enhancement and eventually to a compensation of the upslope precipitation loss. The simulations also indicate that latent heat effects induced by aerosol–cloud–precipitation interactions may considerably affect the orographic flow dynamics and consequently feed back on the orographic precipitation development.

Corresponding author address: Andreas Muhlbauer, Institute for Atmospheric and Climate Science, ETH Zurich, Universitaetsstr. 16, 8092 Zurich, Switzerland. Email: andreas.muehlbauer@env.ethz.ch

Abstract

Aerosols serve as a source of cloud condensation nuclei (CCN) and influence the microphysical properties of clouds. In the case of orographic clouds, it is suspected that aerosol–cloud interactions reduce the amount of precipitation on the upslope side of the mountain and enhance the precipitation on the downslope side when the number of aerosols is increased. The net effect may lead to a shift of the precipitation distribution toward the leeward side of mountain ranges, which affects the hydrological cycle on the local scale.

In this study aerosol–cloud interactions in warm-phase clouds and the possible impact on the orographic precipitation distribution are investigated. Herein, simulations of moist orographic flow over topography are conducted and the influence of anthropogenic aerosols on the orographic precipitation formation is analyzed. The degree of aerosol pollution is prescribed by different aerosol spectra that are representative for central Switzerland. The simulations are performed with the Consortium for Small-Scale Modeling’s mesoscale nonhydrostatic limited-area weather prediction model (COSMO) with a horizontal grid spacing of 2 km and a fully coupled aerosol–cloud parameterization.

It is found that an increase in the aerosol load leads to a downstream shift of the orographic precipitation distribution and to an increase in the spillover factor. A reduction of warm-phase orographic precipitation is observed at the upslope side of the mountain. The downslope precipitation enhancement depends critically on the width of the mountain and on the flow dynamics. In the case of orographic precipitation induced by stably stratified unblocked flow, the loss in upslope precipitation is not compensated by leeward precipitation enhancement. In contrast, flow blocking may lead to leeward precipitation enhancement and eventually to a compensation of the upslope precipitation loss. The simulations also indicate that latent heat effects induced by aerosol–cloud–precipitation interactions may considerably affect the orographic flow dynamics and consequently feed back on the orographic precipitation development.

Corresponding author address: Andreas Muhlbauer, Institute for Atmospheric and Climate Science, ETH Zurich, Universitaetsstr. 16, 8092 Zurich, Switzerland. Email: andreas.muehlbauer@env.ethz.ch

1. Introduction

One of the major challenges in today’s efforts in climate prediction within the framework of the Intergovernmental Panel on Climate Change (IPCC) (Solomon et al. 2007) is to estimate the role of aerosol particles, which interact with clouds and radiation in several ways. In contrast to the greenhouse gases, aerosols are more confined to the local scale. Besides radiative effects, aerosols serve as cloud condensation nuclei (CCN) and are considered to alter the cloud droplet size spectrum toward smaller radii (Twomey et al. 1984; Peng et al. 2002), which directly translates into a change of the microphysical properties of clouds. For a fixed liquid water content (LWC), the change in the cloud droplet spectrum implies also an increase in the cloud albedo [the so-called cloud–albedo effect; Twomey et al. (1984)] and a decrease in drizzle production [the so-called cloud lifetime effect; Albrecht (1989)]. In a recent review on aerosol indirect effects, Lohmann and Feichter (2005) summarized that GCM estimates of global annual mean radiative perturbations at the top of atmosphere range from −0.5 to −1.9 W m−2 because of the cloud albedo effect and from −0.3 to −1.4 W m−2 because of the cloud lifetime effect. However, the confidence in these values is very low and no estimate has yet been given for the possible implications of the aerosol–cloud interactions on precipitation formation (Denman et al. 2007).

Case studies suggest a tendency for precipitation to decrease with increasing aerosol load because of smaller collision efficiencies for cloud droplets in warm-phase clouds (Ferek et al. 1998; Peng et al. 2002). Since collision is a prerequisite for producing precipitation efficiently, an inhibition of this cloud microphysical process is assumed to yield a prolongation of the precipitation development, which then potentially leads to precipitation suppression.

On a local scale, orography is a key factor for inducing as well as enhancing precipitation. So far, different mechanisms for the development of orographic precipitation have been identified, such as neutrally stratified frontal air masses impinging on a mountain range, or triggering of convection by either leeward confluence and moisture convergence or slope wind circulations in a valley (Smith 1989b; Roe 2005). The classical picture of orographic precipitation is associated with the smooth upslope ascent of a stably stratified moist air mass along a topographic barrier. Here, air masses are mechanically lifted to the lifting condensation level (LCL) where condensation occurs and latent heat is released. Typical mechanisms that are known to be responsible for enhancing orographic precipitation are either related to the dynamics of the flow such as orographic flow blocking and the turbulence induced by wind shear in a low-level jet (Houze and Medina 2005) or to the potential instability of the upstream sounding, which may trigger cellular embedded convection (Kirshbaum and Durran 2004; Fuhrer and Schär 2005). Furthermore, microphysical processes in mixed-phase clouds, such as the Bergeron–Findeisen process, may lead to considerable precipitation enhancement. A comprehensive overview of the key mechanisms leading to orographic enhanced or initiated precipitation can be found in, for example, Smith (1979, 1989a, b), Smolarkiewicz et al. (1988), Jiang (2003), Roe (2005), and Rotunno and Houze (2007).

The aerosol indirect effect on orographic precipitation based on records of rain gauge measurements was studied by Givati and Rosenfeld (2004), Givati and Rosenfeld (2005), Rosenfeld and Givati (2006), Jirak and Cotton (2006), and Rosenfeld et al. (2007). For example, Givati and Rosenfeld (2004) employed a linear trend analysis to time series of annual precipitation data from several stations in the United States and in Israel. In that paper, they analyzed the ratio of annual rainfall, which is the ratio of the precipitation measured at a mountain station divided by the precipitation measured at a lowland station located far upstream. The trend analysis was applied to the orographic enhancement factor (increased precipitation with increased height), which was calculated based on annually aggregated precipitation data. Givati and Rosenfeld (2004) hypothesized that the orographic precipitation enhancement factor shows a decreasing trend in areas with air pollution, whereas it is constant in areas without air pollution. The hypothesized implications of the aerosol–cloud interactions on orographic precipitation are the tendency toward a loss of precipitation on the upslope side of the mountain and a possible gain of precipitation on the downslope side of the mountain.

So far, other studies have tackled orographic precipitation problems by either focusing on effects induced by the complexity of the terrain (e.g., Jiang 2006, 2007) or on the impacts of the dynamical and thermodynamical situations (e.g., Miglietta and Rotunno 2005) while keeping the microphysics in the model simple. To understand the possible implication of aerosol–cloud interactions within the framework of orographic precipitation, our approach is to increase the degree of complexity of the microphysical processes in the model but to keep the dynamical conditions, as well as the topography, idealized.

Thus, the main goal of this paper is to investigate to what extent aerosols may contribute to a changed spatial orographic precipitation pattern at the surface and how aerosol–cloud interactions in warm-phase clouds translate into the local precipitation budget. Furthermore, we want to assess whether dynamical conditions exist that favor the change in the local precipitation distribution on the surface and if they are sensitive to the aerosol modification.

The paper is structured as follows: In section 2 we introduce the modeling approach focusing on the numerical model and the parameterizations that are employed. In section 3 we discuss the model setup and the experimental design. We apply our new modeling framework to three dynamical situations, which are analyzed in sections 46. Finally, we end with a discussion of our results and the conclusions in section 7.

2. Numerical model

The model simulations are performed with the nonhydrostatic, fully compressible limited-area mesoscale weather prediction model COSMO (formerly known as the LM; Doms and Schättler 2002; Steppeler et al. 2003), which is currently developed within the Consortium for Small-scale Modeling (COSMO, http://www.cosmo-model.org). The elastic equations are solved in a time-splitting approach with a two time-level total variation diminishing (TVD) third-order Runge–Kutta scheme in combination with a fifth-order horizontal advection scheme. All moisture variables and aerosols are advected by a fourth-order positive-definite advection scheme after Bott (1989).

The boundary conditions are open at the lateral boundaries (Davies 1976) and freeslip at the lower boundary. A Rayleigh damping sponge layer is introduced at the upper boundary to damp reflections of vertically propagating gravity waves. The sponge layer starts at 11 km and covers almost one-half of the vertical model domain.

Since the main focus of this study is given to aerosol–cloud–precipitation interactions via the aerosol indirect effect, all radiative effects, such as the change in cloud albedo, are neglected. Thus, no parameterization of radiative processes is considered here. No convection parameterization is employed.

For the vertically turbulent diffusive processes, a Smagorinsky-type scheme with a prognostic turbulent kinetic energy (TKE) equation is used (Herzog et al. 2002).

a. Cloud microphysics

The microphysical processes are treated within the two-moment bulk-microphysics scheme of Seifert and Beheng (2006). Since the main emphasis is on warm-phase clouds, ice- and mixed-phase processes (e.g., ice nucleation, freezing, riming, the Bergeron–Findeisen process) are not considered here. The two-moment scheme predicts two partial moments Mki of the underlying number density size distribution fi(m) for each hydrometeor category i, which are defined such that
i1520-0469-65-8-2522-e1
Here, m denotes the hydrometeor mass, and m1 and m2 are the integration bounds along the number density size distribution fi(m). The first two partial moments k ∈ {0, 1} are the number density Ni (number concentration) and the mass density Li (liquid water content), respectively. The index i ∈ {c, r} denotes the hydrometeor class, which is, in our study, either cloud droplets (i = c) or rain droplets (i = r). For cloud droplets, the integration boundaries are [0, mr,min], with mr,min = 2.6 × 10−10 kg being the minimum mass of rain. This minimum mass separates cloud droplets from raindrops and corresponds to a minimum radius for the raindrops of 40 μm. For raindrops, the integration boundaries are [mr,min, ∞]. For the number density size distribution fi(m), a generalized gamma distribution of the form
i1520-0469-65-8-2522-e2
with two free parameters νi and μi is assumed. Keeping these two free parameters fixed for each hydrometeor class reduces the generalized gamma distribution to a gamma distribution for the cloud droplets (νc = 1, μc = 1) and an exponential distribution for raindrops (νr = 0, μr = ⅓). The coefficients A and λ can then be related to the power moments Mki of the size distribution, as discussed in Seifert and Beheng (2006).
Generally, each partial moment Mki satisfies the budget equation
i1520-0469-65-8-2522-e3
which accounts for advection, turbulent diffusion, and sedimentation. Note that v = (u, υ, w) is the mean wind velocity, Kh is the turbulent diffusivity of heat, and wi,k is the mean vertical sedimentation velocity, which is different for the number and the mass density. Thus, large raindrops fall faster than small raindrops. In the case of warm-phase microphysics, the source term Ski comprises the nucleation of cloud droplets, condensation–evaporation of cloud droplets, autoconversion of cloud droplets to rain, accretion of cloud droplets by rain, self-collection of cloud droplets and rain, evaporation of rain, and the breakup of large raindrops.

1) Nucleation of cloud droplets

For the activation of the soluble/mixed aerosols and cloud droplet nucleation, a modified version of the Ghan et al. (1993) parameterization is used (Leaitch et al. 1996; Lin and Leaitch 1997; Lohmann 2002):
i1520-0469-65-8-2522-e4
i1520-0469-65-8-2522-e5
and
i1520-0469-65-8-2522-e6
Here, α = 0.023 cm4 s−1 is a constant and w is the grid-scale vertical velocity. The minimum cloud droplet mass mc,min is assumed to be mc,min = 4.2 × 10−15 kg, which corresponds to a minimum cloud droplet radius of 1 μm. We use N ta as the number concentration (at time step t) of all soluble/mixed aerosols in the coarse mode (COA) and accumulation mode (ACC) plus the number of mixed Aitken mode (AIT) particles exceeding a critical dry radius for the activation of ractiv = 35 nm:
i1520-0469-65-8-2522-e7
where N tc and N t−1c are the number concentrations of the activated cloud droplets at time step t and t − 1, respectively. The number concentrations NCOA, NACC, and NAIT of the soluble/mixed aerosols in the different modes are provided by the aerosol microphysics scheme, which is described in section 2b.

2) Condensational growth

To remove the saturation adjustment scheme, the condensation and evaporation processes of cloud droplets are calculated explicitly, similar to an approach by Lohmann and Kärcher (2002), but for the warm-phase cloud liquid water content:
i1520-0469-65-8-2522-e8
with
i1520-0469-65-8-2522-e9
Here, T is the temperature, p is the pressure, S is the water saturation ratio, rc is the mean cloud droplet radius, and Nc is the cloud droplet number density. The ventilation factor FRe accounts for turbulent effects in the droplet growth equation as a function of the Reynolds number (Pruppacher and Klett 1997, p. 541). Furthermore, Rυ is the gas constant of water vapor and es(T) is the saturation vapor pressure over water. In addition, Lυ(T) is the latent heat of vaporization (Bolton 1980), ka(T) is the thermal conductivity of air (Seinfeld and Pandis 1998, p. 804), and Dυ(T, p) is the diffusivity of water vapor (Pruppacher and Klett 1997, p. 503). By predicting the two partial moments Nc and Lc of the underlying size distribution, the average cloud droplet radius rc can be diagnosed by assuming spherical droplets:
i1520-0469-65-8-2522-e10
Here, ρw is the density of water.

3) Autoconversion, accretion, and self-collection

Following Seifert and Beheng (2006), the autoconversion process (collisions of cloud droplets leading to rain) is parameterized in the subsequent way:
i1520-0469-65-8-2522-e11
i1520-0469-65-8-2522-e12
Here, kc = 4.44 × 109 m3 kg−2 s−1 is a constant cloud–cloud kernel based on the collision efficiencies after Pinsky et al. (2001). The fraction ρ0/ρ accounts for an increase in the autoconversion rate with decreasing air density ρ (relative to ρ0 = 1.225 kg m3 at the surface) because of larger hydrometeor terminal velocities. Here, ΦAU(τ) is a universal autoconversion function developed from numerical solutions of the stochastic collection equation and depends on an internal dimensionless timescale τ:
i1520-0469-65-8-2522-e13
The denominator of the autoconversion Eq. (11) depends nonlinearly on the cloud droplet number concentration. Hence, an increase in the cloud droplet number concentration leads to a reduction in the autoconversion rate.
The accretion process (collection of cloud droplets by rain) is parameterized in the following form:
i1520-0469-65-8-2522-e14
i1520-0469-65-8-2522-e15
with the cloud–rain kernel kcr = 5.25 m3 kg−1 s−1 and the accretion function ΦAC(τ). Note that the accretion in Eq. (14) depends linearly on the cloud and rainwater content.
The self-collection (collection of cloud–rain that does not lead to a change in hydrometeor class) is parameterized in the following way for cloud droplets,
i1520-0469-65-8-2522-e16
and for rain
i1520-0469-65-8-2522-e17
with the rain–rain kernel krr = 7.12 m3 kg−1 s−1 and krr = 60.7 kg−1/3. Here, it is assumed that the raindrop number distribution follows an exponential size distribution with slope parameter λr.

b. Aerosol microphysics

For the treatment of aerosol size distributions for different aerosol species and in order to simulate the aerosol indirect effect on precipitation consistently, the aerosol microphysics module M7 (Vignati et al. 2004; Stier et al. 2005) has been coupled to the COSMO model. The M7 module distinguishes between different aerosol species, which can be sulfate (SU), carbonaceous aerosols (organic carbon, OC, and black carbon, BC), sea salt (SS), and mineral dust (DU). This modeling framework allows for the prediction of aerosol spectra by a superposition of up to seven lognormal distributions of the form
i1520-0469-65-8-2522-e18
by keeping the geometric standard deviation σi fixed for each mode. The seven aerosol modes are composed of either soluble/mixed modes (e.g., sulfate, sea salt, mixed organic, or black carbon) or insoluble modes (e.g., freshly formed black carbon or mineral dust). Here, the prognostic variables are the number density Ni and the mass density mi,j for each mode i and aerosol species k (e.g., mAIT,BC insol, denoting the mass density of insoluble black carbon aerosols in the Aitken mode). The count median radius i can then be diagnosed from the number densities Ni and the mass densities mi,j. Generally, the number densities Ni and mass densities mi,j are also governed by a budget equation such as Eq. (3) including advection and vertical turbulent diffusion. However, aerosol sedimentation processes are neglected throughout this study. One may argue that the (size dependent) sedimentation time scales of the aerosols that are considered here are much larger than the simulation time. For the aerosol size ranges that are considered here, the most dominant aerosol removal process is in-cloud nucleation scavenging rather than dry deposition. The aerosol dynamical processes in M7 that contribute to the source term [such as Ski in Eq. (3)] include the nucleation of gas-phase sulfuric acid (Vehameki et al. 2002), the condensation of sulfuric acid on preexisting aerosol particles, coating of insoluble aerosols by sulfuric acid, inter- and intramodal coagulation, and the uptake of water vapor. For a more detailed description of the parameterized processes and the underlying equations, we refer to Vignati et al. (2004).

3. Model setup and experimental design

In this study, idealized two-dimensional simulations of moist orographic flows over a bell-shaped topography are conducted. The influence of aerosol particles on the orographic precipitation distribution and the orographic wave response are analyzed by comparing a polluted case against a clean reference case. The dimensionless height (inverse Froude number)
i1520-0469-65-8-2522-e19
may be used to characterize the behavior of the orographic flow and the orographic flow regime in terms of the dry Brunt–Väisälä frequency Nd, the mountain height h, and the horizontal wind speed U. In a dry atmosphere with constant Nd and U and ĥ < 1, the linear theory (Smith 1979, 1980) predicts a linear hydrostatic mountain wave. For a critical value of ĥ ≈ 1, a stagnation point forms at the leeward side of the mountain, which leads to a steepening of the isentropes and gravity wave breaking above the mountain crest. In this case, linear theory is formally no longer valid and breaks down. However, the critical value of the dimensionless mountain height depends also on other factors such as the geometry of the obstacle, the moisture content in the atmosphere, and the surface friction. For high values of ĥ (low Froude number), flow stagnation also occurs on the windward side, which leads to flow splitting or even flow reversal (Smolarkiewicz and Rotunno 1990). Durran and Klemp (1983) found that the addition of moisture in the atmosphere damps the amplitude of the mountain waves and considerably reduces the horizontal wind speeds, while the vertical wavelength of the mountain wave is increased. If clouds and precipitation are considered, the latent heat release may delay the onset of gravity wave breaking and orographic flow splitting (Jiang 2003). This phenomenon can be understood within the concept of the moist Brunt–Väisälä frequency Nm, which accounts for the water vapor in the atmosphere. Since the addition of moisture reduces the effective stability, the moist Brunt–Väisälä frequency Nm is always lower than the corresponding dry Brunt–Väisälä frequency. Jiang (2003) noted that under saturated conditions the dimensionless parameter ĥ may be adapted by the moist Brunt–Väisälä frequency as long as Nm does not change severely within one vertical wavelength of the mountain wave. Since both conditions (saturated environment and vertically constant Nm within one vertical wavelength) are not satisfied in our simulations, we only give values for the dry dimensionless mountain height in order to characterize the flow regime.

a. Computational domain

The 2D computational domain is composed of 200 grid points in the horizontal with a grid spacing of 2 km, which yields a domain of −200 km ≤ x ≤ 200 km along the x axis. The time step of the model is 10 s. A hybrid stretched height-based Gal-Chen vertical coordinate is used with 50 layers and a vertical grid spacing varying between 10 m in the lowermost and 1000 m in the uppermost layer. The model top is located at 21.5 km.

b. Topography

The idealized bell-shaped topography has the form
i1520-0469-65-8-2522-e20
with the peak mountain height h0 located at x0 and a being the mountain half-width (Kirshbaum and Durran 2004). The mountain range is centered in the computational domain such that x0 = 100 in gridpoint space. Throughout this study we use this idealized topography with different values of h0 (1000, 3000 m) and a (10–30 km) to simulate different relevant topographic barriers.

c. Dynamical initialization

The model is initialized with a horizontally homogeneous basic state, which is given by a dry atmosphere at rest with surface pressure pSL = 1000 hPa and surface temperature TSL = 288.15 K. The basic state is hydrostatically balanced and the temperature increases with the logarithm of pressure such that ∂T/(∂ lnp) = 42 K. The initial horizontally homogeneous profile of the pressure p(z) and the temperature T(z) is calculated analytically as a function of the surface pressure pSL, the surface temperature TSL, and the dry Brunt–Väisälä frequency Nd (Clark and Farley 1984). In our simulations, the surface pressure and surface temperature are prescribed by pSL = 1000 hPa and TSL = 285 K, respectively. The dry Brunt–Väisälä frequency is chosen to be constant with height with Nd = 0.0115 s−1. The vertical profile of the relative humidity is prescribed by a modified Fermi function (Spichtinger 2004) of the type
i1520-0469-65-8-2522-e21
with the parameters a = 0.8, b = 0.03, c = 0.0015 m−1, z0 = 5000 m, and 0 ≤ RH ≤ 1. The modified Fermi function gives a vertical profile of the relative humidity, which starts with the value a = 0.8 at the surface and decays smoothly with height toward the value b = 0.03.

The vertical decay is controlled with the parameters c and z0. The horizontal wind velocity U is prescribed unidirectionally and is vertically constant with values in the range of U = 14–20 m s−1 for the different experiments. The vertical profiles of temperature and dewpoint temperature are shown in the skew T–log p diagram in Fig. 1. The sounding depicts an almost neutrally stratified atmosphere in the lower layers with marginal potential instability and increasing stability with height. Although the lowermost layers are potentially unstable, the sounding is still statically stable since the moist squared Brunt–Väisälä frequency N2m is positive throughout the whole atmosphere. Thus, we expect this profile to yield an orographic cloud that is induced by smooth upslope ascent without contributions of thermally induced cellular convective processes. The lifting condensation level (LCL) is located at approximately zLCL ≈ 425 m.

Since the Rossby number Ro = U/fL is usually greater than unity for the characteristic scales that are considered here (e.g., U = 20 m s−1, f ≈ 10−4 s−1 at midlatitudes, and L = 10 km), the effect of the Coriolis force is neglected in this study. Moreover, model simulations by Colle (2004) showed that the inclusion of rotation has little impact on the orographic precipitation sensitivity.

d. Aerosol initialization

To simulate characteristic clean and polluted aerosol conditions, initial aerosol distributions are prescribed. The degree of aerosol pollution is simulated by prescribing initial aerosol spectra that characterize a clean and a polluted case, as shown in Fig. 2. The aerosol number concentrations were measured with a scanning mobility particle sizer (SMPS) at the Jungfraujoch in the Bernese Alps during a field campaign in 1999 (Weingartner et al. 1999) and may be seen as representative values for central Switzerland. The aerosol mass distribution is calculated analytically from the number distribution by assuming spherical particles and an average aerosol density of 1.9 g cm−3.

Since the Jungfraujoch is a free-tropospheric site during wintertime, the aerosol size distributions are typical for remote continental conditions throughout the winter months. However, during summertime, the relatively clean air at the Jungfraujoch is often entrained with air from the planetary boundary layer and the aerosol number concentrations are usually higher. This mixing of boundary layer air is induced by either convective processes or locally confined slope circulations, which transport anthropogenic pollutants into the free troposphere (Weingartner et al. 1999). In contrast to the wintertime profile, the summertime profile exhibits a well-pronounced bimodal structure. Throughout this study we use the averaged wintertime aerosol size distributions as the clean reference case and the average summertime size distributions as the polluted case. All aerosols are assumed to be internally mixed, which also corresponds with the observations made at the Jungfraujoch (Weingartner et al. 2002; Cozic et al. 2007; Verheggen et al. 2007). The clean and polluted cases are superimposed on a vertically constant background aerosol composition that consists of an internal mixture of organic carbon and sulfate particles in the accumulation mode. From aerosol mass spectrometry it was found that organics and sulfate contribute dominantly to the chemical aerosol (PM1) composition at the Jungfraujoch (Cozic et al. 2008) and in the upper troposphere (Cziczo et al. 2004; Murphy et al. 2006). Sulfate nucleation mode particles are assumed to form by homogeneous nucleation (gas to particle conversion) from the supersaturated gas-phase sulfuric acid concentration at low temperatures and, therefore, preferably in the upper troposphere. These nucleation mode aerosols are smaller than 10 nm in diameter and are not available from the SMPS measurements. To get an estimate of the vertical distribution of the background gas-phase sulfuric acid concentration and the sulfate aerosols in nucleation mode, pseudosoundings are taken from the ECHAM5-Hamburg Aerosol Model (HAM; Stier et al. 2005) for the nearest Jungfraujoch grid point. These vertical profiles are averaged over the winter (October–March) and summer months (April–September), respectively, and are shown in Fig. 3. The sulfuric acid concentrations and concentrations for the sulfate nucleation mode aerosols are taken from simulations that use the year 2000 emission inventories (Lohmann et al. 2007).

The aerosol number and mass concentrations for the lognormally distributed aerosol species are shown in Table 1. In the vertical, the initial aerosol number concentrations Ni and the corresponding mass concentrations Mi,j of aerosol species j in mode i decay exponentially with height such that
i1520-0469-65-8-2522-e22
i1520-0469-65-8-2522-e23
assuming a typical scale height for remote continental aerosols of HN = HM = 730 m (Jaenicke 1993) for the number and mass densities of all aerosol species. The initial number and mass densities are superimposed on the vertically constant background aerosol. Note that, for the sake of simplicity and because of the lack of adequate measurements, the scale heights of the number and mass densities are considered to be identical for all aerosol species. Throughout this study we restrict our focus to black carbon, organic carbon, and sulfate aerosols since these particles are, besides nitrates, usually most abundant in central Switzerland (Hueglin et al. 2005) except during intense southwesterly flows, which occasionally lead to Saharan dust events (Cozic et al. 2008).

4. Experiment A: Mountain wave

In the first experiment, the wind speed is set to U = 20 m s−1 and the height of the idealized topography is set to h = 1000 m with a half-width of a = 10 km. Considering a vertically constant Brunt–Väisälä frequency of Nd = 0.0115 s−1 gives a dimensionless mountain height of ĥ = 0.58. After 10 h of simulation, the flow develops a hydrostatic mountain wave with upstream phase tilt in the lower-atmospheric layers as shown in Fig. 4. In the dry simulation, the horizontal wind speeds reach maximum values of up to 35 m s−1. The corresponding moist simulation shows appreciably lower horizontal wind speeds with maximum values of approximately 30 m s−1. Although the orographic flow is in the same flow regime for both simulations, the addition of moisture weakens the mountain wave and reduces the horizontal wind velocity but increases the vertical wavelength. These results are consistent with the findings of Durran and Klemp (1983).

a. Aerosol–cloud interactions

The smooth ascent in the updraft region of the hydrostatic mountain wave results in the formation of an orographic cloud on the upslope side of the mountain as shown in Fig. 5. As a consequence of the phase tilt of the mountain wave, the updraft region shifts slightly upstream with height. The cloud base of the orographic cloud is located at the LCL at an altitude of approximately 425 m and the cloud top is located at roughly 4500-m height.

The orographic cloud shows a well-pronounced maximum in the cloud water mixing ratio of approximately 1 g kg−1, located on the upstream slope close to the mountaintop. The rainwater mixing ratio shows a maximum of almost 0.2 g kg−1, which is located at the downslope side of the mountain in the region of the mountain crest. In contrast to the conventional upslope model of orographic precipitation (e.g., Smith and Barstad 2004), the condensate does not fall out immediately but is delayed because of the microphysical conversion processes (e.g., cloud droplet conversion into rain). The delay in the microphysical conversion affects the downstream advection and sedimentation of the hydrometeors and explains the downstream shift of the rain maximum (Hobbs et al. 1973; Smith 2003; Smith and Barstad 2004).

Since the aerosol number density is higher in the polluted case than in the clean case, more cloud droplets are activated that are smaller than in the clean case. The maximum cloud droplet number concentrations vary between approximately 60 cm−3 in the clean case and 380 cm−3 in the polluted case. The time-averaged mean cloud droplet radius is 17 μm in the clean case and 12 μm in the polluted case. This size effect makes diffusional growth more effective but reduces the efficacy of the rain production through the kinetic growth mechanisms such as autoconversion and accretion in the polluted case. Hence, the liquid water content is slightly higher in the polluted case than in the clean case since more condensate stays in the cloud while the opposite is true for the rainwater content. A maximum of approximately 1 g kg−1 in the cloud water mixing ratio is attained in the polluted case whereas the maximum rainwater content reaches values on the order of 0.02 g kg−1. Thus, the rainwater content is almost an order of magnitude lower than in the clean case simulation and the rainwater maximum shifts from the region of the mountain crest farther toward the leeward side. The truncation of the orographic cloud in the lee indicates the importance of evaporation processes, which predominantly occur in the downdraft region of the mountain wave on the leeward side. The role of evaporating rain is also reflected in the time-averaged mean raindrop radius, which decreases in the polluted case (rr = 20 μm) compared to the clean case (rr = 47 μm).

b. Precipitation budgets

The surface precipitation distribution after 10 h is shown in Fig. 6 for the idealized bell-shaped orography with a height of 1000 m and half-widths of 10 and 30 km, respectively. For the narrow mountain and in the clean case, the precipitation mainly falls at the mountain crest and on the leeward side of the mountain with a pronounced maximum on the downslope side. This result is consistent with the expected orographic precipitation distribution for narrow mountain ranges (Smith 1979). The spillover factor, which is the fraction of leeward precipitation to the total precipitation (Jiang 2003), is 0.96, meaning that 96% of the precipitation falls on the leeward side of the mountain.

In contrast to the clean case, the polluted case shows an inhibition of the upslope precipitation together with a shift of the precipitation maximum of roughly 6 km farther toward the leeward side. As a consequence, the spillover factor increases nearly to 1 in the polluted case (see also Table 2). Hence, almost all precipitation that is generated by the mountain wave falls on the leeward side. This precipitation shift suggests that the conversion of cloud droplets to rain is retarded because of the indirect aerosol effect on precipitation and, therefore, the cloud droplets and raindrops are more likely to be advected downstream. Since in the clean case simulation the precipitation maximum is already on the leeward side, the strong decrease in precipitation in the polluted case may be explained by hydrometeors entering the downdraft region of the mountain wave and are thus more likely subject to leeward evaporation. However, the spillover factor increases as a result of the increased aerosol load by 0.04. This increase in spillover reflects how the microphysical changes in the cloud properties can translate into changes in the spatial distribution of the orographic precipitation pattern via the aerosol–cloud–precipitation interactions and by retarding the hydrometeor conversion processes. Throughout the simulation, the time-averaged mean mass conversion rates of rain are roughly an order of magnitude lower in the polluted case than in the clean case. It is interesting to note that the decrease in the total mass conversion rates is caused by a reduction of the autoconversion rate as well as the accretion rate to a comparable extent. Hence, the reduced rainfall in the polluted case is a combined effect of decreasing rain formation due to decreasing collision and collection of cloud droplets and the evaporation of raindrops on the leeward side of the mountain.

If the half-width of the mountain is increased to a = 30 km, the spatial precipitation distribution maximizes on the upslope side of the mountain and the precipitation maximum is considerably weaker than in the case of the narrow mountain (with a = 10 km), although the amount of total domain precipitation along the topography is higher. In the case of a broad mountain, the steepness of the slope is decreased along the topography compared to the narrow mountain, which leads to weaker vertical velocities in the mountain wave and thus to a lower precipitation maximum. An upstream shift of the precipitation maximum with increased width of the topographic barrier is also consistent with observations and other studies (e.g., Smith 1979; Jiang 2003) because the updraft region of the mountain wave shifts upstream with increased half-widths. This upstream shift of the precipitation pattern is also evident in the polluted case but the maximum precipitation remains located on the downstream side of the mountain and the amount of upslope precipitation is reduced. Again, the spillover factor increases with increased aerosol load from 0.58 in the clean case to 0.89 in the polluted case (see also Table 2) but is generally lower than for the narrow mountain, which is in agreement with the findings of Jiang (2003). Here, the effects of adding aerosols to the mountain flow can be seen as a decrease of the effective mountain half-width since, generally, more precipitation is advected downstream and the spillover factor is increased. This decrease in the spillover factor with increased mountain half-width is consistently found for all simulations in experiment A and the details are summarized in Table 2.

Figure 7 shows the spatial precipitation differences for different mountain widths. It turns out that the aerosol–cloud–precipitation interaction leads to an overall loss of precipitation regardless of the mountain half-width since the gain in leeward precipitation is virtually zero after 10 h. In terms of absolute values, the maximum precipitation loss occurs on the leeward side in the case of a narrow mountain (e.g., a = 10 km) and on the upslope side in the case of a wider mountain (e.g., a = 30 km). It is noted that leeward precipitation enhancement may critically depend on the temperature and moisture stratification on the leeward side but also on the time scale of the microphysical processes. For example, the simulations with mountain half-widths of a = 20 km and a = 30 km give a slight leeward precipitation enhancement but only after 24 h of simulation, which may be a rather long time for a naturally occurring orographic precipitation event. However, the upslope precipitation loss is not compensated for by the leeward precipitation gain in all simulations of experiment A.

c. Aerosol–cloud dynamics interactions

The modification of the orographic precipitation pattern caused by the addition of aerosols to the background flow leads also to a dynamical feedback especially in terms of the horizontal wind speed. Simulations by Chaumerliac et al. (1987) suggested that increasing the CCN concentration leads to higher values of the horizontal wind speed in a moist mountain wave flow. Their maritime simulation having less CCN depicted the lowest maximum wind speeds when compared to the corresponding continental case simulation with more CCN. Their dry simulation showed the highest wind speeds. These maxima in the horizontal wind speed occurred in the region of the mountain wave directly over the mountain crest, suggesting that the addition of aerosols leads to a more vigorous orographic wave response. In our case, the continental simulation corresponds to the polluted case since more aerosols are activated, which leads to more but smaller cloud droplets. In our mountain wave experiment, the differences in maximum horizontal wind speed between the clean and the polluted case simulation are fairly small compared to the differences obtained in the simulations of Chaumerliac et al. (1987). A closer look at the difference fields for the horizontal wind speed, shown in Fig. 8, reveals that the aerosol–cloud–precipitation interactions also change the leeward flow dynamics.

The increased transport of hydrometeors toward the leeward side of the mountain, which is also reflected by the increased spillover factor in the polluted case, leads to strong evaporative cooling. This leeward evaporation enriches the water vapor content and leads to a cooling below the ridge height on the downstream side of the mountain. Both effects increase the relative humidity locally. The addition of moisture then decreases the effective stability on the leeward side, which in turn translates into a damping of the amplitude of the orographic mountain wave on the leeward side. This weakening of the mountain wave response causes lower wind speeds in the polluted case below the ridge height with the effect being largest close to the surface.

5. Experiment B: Hydraulic jump

To investigate the role of aerosols in the formation of orographic precipitation under different dynamical conditions, further experiments are conducted. In the second experiment the orographic flow response is severely changed by modifying the flow regime from a linear hydrostatic mountain wave to a nonlinear flow regime with features of wave breaking and the development of a hydraulic jump. The change in the flow regime is introduced by reducing the unidirectional horizontal wind speed from U = 20 m s−1 to U = 14 m s−1 while keeping the mountain height and the dry Brunt–Väisälä frequency constant, as described in section 4. Modifying these parameters gives a nondimensional mountain height of ĥ = 0.82. Figure 9 shows the dry as well as the moist simulations for the clean case after 10 h.

Jiang (2003) showed that the addition of moisture offsets the formation of the stagnation point above the mountain ridge in nonlinear orographic flows, which then leads to a delay in the onset of the gravity wave breaking and overturning of the isentropes. In other words, the increase in moisture pushes the flow regime toward smaller values of the dimensionless mountain height and makes the orographic wave response more linear than the corresponding dry simulation. In our case, the moist simulation depicts lower horizontal wind speeds than the dry simulation but the strong nonlinear features, such as the leeward wave overturning and the hydraulic jump, are retained in both simulations. Thus, just the addition of moisture does not change the flow regime in our simulations.

Similar to the first experiment, an orographic cloud develops on the upslope side of the mountain, which is heavily truncated by the strong downdraft region of the mountain flow in the lee (Fig. 10).

These strong downdrafts transport potentially warm air masses from the upper-tropospheric layers down to the near-surface layers. After 10 h, the resulting downslope winds reach values of up to u = 49 m s−1 in the dry and u = 47 m s−1 in the moist clean simulations, respectively. In both cases, the wind speeds maximize close to the surface. Here, the development and the vertical extent of the orographic cloud are also governed by the phase tilt of the upstream flow and the spatial onset of the downslope winds on the upstream side of the mountain. In the polluted case, the phase tilt is slightly stronger developed, which truncates the orographic cloud also in the vertical. Nevertheless, the vertical cloud extent is still comparable in both cases. However, the reduction in precipitation is even more severe in this experiment since the evaporation processes are enforced because of the stronger and warmer downdrafts which pose a barrier for the orographic precipitation. Investigating the orographic precipitation budgets in Fig. 11 reveals that the strong loss in upslope precipitation is not compensated by a leeward gain. The aerosol modification leads to an overall reduction of orographic precipitation along the topography.

6. Experiment C: Flow reversal

In this experiment the mountain height of the bell-shaped topography is increased to h0 = 3000 m and the half-width is a = 10 km. These choices yield an orographic barrier similar to that of the Jungfraujoch in central Switzerland. Keeping the dry Brunt–Väisälä frequency and the horizontal wind speed fixed gives a nondimensional height of ĥ = 2.46, which leads to windward flow blocking. The flow fields for the clean and the polluted case simulations are shown in Fig. 12.

Similarly to the experiment in the previous section, the leeward dynamics are governed by severe downslope winds with the maximum wind speeds being highest in the polluted case. However, the more interesting features occur on the windward side of the mountain in this experiment. Because of the strong orographic blocking, a zone of flow reversal develops in the lower atmospheric levels. At the position where the reversed flow meets the incoming flow on the windward side of the mountain, a perturbation in the vertical velocity field is generated. This perturbation propagates slowly against the incoming flow direction as time proceeds. The reverse circulation, together with the upstream propagating velocity perturbation, are responsible for the formation of a secondary rainband, which is shown in Fig. 13.

The development of a secondary rainband in association with blocked orographic flows is an important factor in modulating the orographic precipitation distribution far upstream of mountain ranges as discussed by Jiang (2003). The polluted case simulation shows considerably higher values of the cloud water mixing ratio close to the mountain peak whereas the rainwater mixing ratio is reduced. In contrast to the clean case, the formation of the secondary rainband is retarded in the polluted case, which leads to a loss of precipitation far upstream of the mountain. However, the precipitation along the topography is comparable in both simulations with slightly more precipitation in the polluted case than in the clean case, which is shown in Fig. 14.

The simulation with the wide mountain and a = 30 km gives similar results such as reduced precipitation in the secondary rain belt but, this time, the total precipitation along the topography is lower in the polluted case than in the clean case. Figure 15 shows the precipitation differences along the topography for different mountain half-widths ranging from a = 10 km to a = 30 km. For the narrow mountain range, a loss in the precipitation in the secondary rainband is accompanied by a gain in upslope precipitation and a narrow region of precipitation enhancement on the leeward side. However, the precipitation distribution and the upslope precipitation loss is much more complex owing to the nonsteadiness of the reverse circulation. Similar features can be seen for the wider mountain with a = 20 km despite their spatial displacements due to the change in mountain half-width. In contrast, the simulation with the widest mountain (a = 30 km) gives no precipitation enhancement on the leeward side and the leeward precipitation gain is most pronounced for the narrow mountain.

Regarding the orographic precipitation budgets, the outcome of experiment C is fundamentally different from experiment A in terms of the microphysical processes contributing to the mean mass conversion. Similarly to experiment A, the increased aerosol load leads to more but smaller cloud droplets. The time-averaged mean cloud droplet size decreases from 20 μm in the clean case to 16 μm in the polluted case and the raindrop size decreases from 54 to 50 μm, respectively. Although the time-averaged mean autoconversion rates are again lower in the polluted case than in the clean case, the opposite is true for the accretion rates. In this experiment, the formation of orographic precipitation is mainly dominated by the accretion process, which compensates for the reduction in autoconversion.

7. Discussion and conclusions

Idealized 2D simulations of moist orographic flow are conducted with the numerical weather prediction model COSMO at a horizontal grid spacing of 2 km. To simulate the indirect aerosol effects on orographic precipitation formation, the aerosol microphysics and cloud microphysics parameterizations are fully coupled following a two-moment approach.

Our model simulations suggest that in warm-phase clouds more aerosols lead to more cloud droplets, which are then smaller in size. Throughout the simulations, the reduction in the average cloud droplet size leads also to a reduction in the average autoconversion rate. Furthermore, in the simulations with increased aerosol, a reduction of the mean raindrop size can be observed. These microphysical changes in the orographic cloud modify the orographic precipitation distribution along the topography. A downstream shift in the orographic precipitation pattern is observed, which leads to a more pronounced advection of precipitation toward the downslope side of mountain ranges. It is also shown that increasing the aerosol load increases also the orographic spillover factor and hence the fraction of precipitation falling on the leeward side. However, the modification of the orographic precipitation pattern, and in particular the question of leeward precipitation enhancement, depend also on factors that are directly connected to the topography, such as the half-width and the height of the mountain. These geometric factors govern the orographic flow response and modulate the relative contribution of microphysical processes such as autoconversion, accretion, and evaporation. The orographic precipitation budgets are controlled by the complex interaction of dynamical and microphysical processes.

In the case of orographic precipitation from a linear hydrostatic mountain wave, the downstream advection of hydrometeors favors evaporation processes in the warm and dry downdraft region of the orographic flow on the leeward side of the mountain. These evaporation processes may contribute dramatically to the orographic precipitation loss in addition to the loss induced by the retarded hydrometeor conversion. Hence, in this case, the altered orographic precipitation distribution leads to a loss of upslope precipitation, which is not compensated for by a precipitation gain on the leeward side. Sensitivity studies indicate that the effect of aerosols suppressing upslope precipitation development in linear mountain waves is most pronounced for narrow mountain ranges. For narrow mountains, the updraft region is smaller and the advection time of air parcels in cloud is shorter than for wide mountains. Both factors limit the time available for hydrometeor growth.

In situations of blocked orographic flow, the contribution of the accretion process to the orographic precipitation becomes more dominant than in the unblocked flow, which may lead to a net compensation of the retarded autoconversion process. This microphysical compensation effect may result in comparable or even more total precipitation in the polluted situation, although upslope precipitation is still reduced. Precipitation enhancement on the leeward side of the mountain is only observed in the simulations where the topographic barrier is rather large and narrow and where the orographic flow is blocked. In these situations, the orographic blocking is accompanied by a zone of flow reversal on the upstream side of the mountain and the formation of a secondary rainband. Here, precipitation is reduced in the secondary precipitation belt far upstream of the mountain under polluted conditions. Again, the orographic precipitation budgets depend also on the width of the underlying topography.

It is emphasized that aerosol–cloud–precipitation interactions also affect the dynamics of the flow via latent heat release (e.g., condensation–evaporation processes). Below the ridge height, the downstream advected hydrometeors evaporate in the downdraft region of the mountain wave, which leads to evaporative cooling and moistening of the lower-atmospheric layers. The increased moisture then reduces the effective stability in these layers, damps the amplitude of mountain waves, and reduces the horizontal wind speed. Thus, aerosols interact indirectly with the ambient flow dynamics via the latent heating induced by clouds and precipitation.

A quantification of the aerosol–cloud–precipitation interactions in terms of the orographic enhancement factor as proposed by Givati and Rosenfeld (2004) critically depends on the spatial precipitation pattern and, therefore, on the location of the reference points used for the calculation along the orography. Therefore, we quantify the aerosol–cloud–precipitation interactions in terms of the orographic spillover factors (leeward precipitation fraction), following the definition of Jiang and Smith (2003). Our simulations indicate a clear increase in the spillover factor with increased aerosol load especially in the situation of an unblocked flow. However, the absolute differences in the spillover factor depend also on the mountain half-width.

Our model simulations emphasize the importance of considering aerosols and the indirect aerosol effect on local-scale precipitation as a further piece in understanding the orographic precipitation distribution in mountainous terrain. Our results suggest that aerosols may have the potential to affect the precipitation development and the hydrological cycle in mountainous regions, such as the Alps, although the net effect in terms of the water budget is still unclear. Under changed climatic conditions, the reduction of upslope orographic precipitation due to aerosol–cloud–precipitation interactions may have considerable impacts on the local water resources. Addressing this question motivates the consideration of the indirect aerosol effect on precipitation also for regional climate simulations.

To establish confidence in our key findings, given the uncertainty that is introduced by the microphysical representation, we also performed sensitivity experiments with other two-moment cloud microphysics parameterizations such as the Beheng (1994) scheme or the Khairoutdinov and Kogan (2000) scheme. We also tested the sensitivities with respect to different cloud collision kernels in the Seifert and Beheng (2006) parameterization. These sensitivity studies indicate that quantitative differences with respect to aerosol–cloud–precipitation interaction exist among the schemes that are attributable to the microphysical representation and the intrinsic time scales of the schemes to develop precipitation. However, despite these quantitative differences, qualitatively similar results are obtained. Thus, we argue that our major conclusions and inferences are robust within the range of state-of-the-art microphysical parameterizations. Finally, it is emphasized that our results are constrained to single warm-phase orographic clouds and idealized dynamical situations. Investigating the effects of aerosol–cloud–precipitation interactions in single and/or multiple mixed-phase orographic clouds is the subject of forthcoming publications.

Acknowledgments

We thank the two anonymous reviewers and the editor Bjorn Stevens for their helpful comments and suggestions. We are grateful to Axel Seifert and Peter Spichtinger for numerous fruitful comments and discussions and to Daniel Cziczo and Sarah Kew for carefully proofreading the manuscript. We also thank the European Centre for Medium-Range Weather Forecasts (ECMWF) for providing computing time within the special project Cloud Aerosol Interactions (SPCHCLAI) and gratefully appreciate the user support for the help with the supercomputing facility.

REFERENCES

  • Albrecht, B., 1989: Aerosols, cloud microphysics, and fractional cloudiness. Science, 245 , 12271230.

  • Beheng, K. D., 1994: A parameterization of warm cloud microphysical conversion processes. Atmos. Res., 33 , 193206.

  • Bolton, D., 1980: The computation of equivalent potential temperature. Mon. Wea. Rev., 108 , 10461053.

  • Bott, A., 1989: A positive definite advection scheme obtained by nonlinear renormalization of the advective fluxes. Mon. Wea. Rev., 117 , 10061015.

    • Search Google Scholar
    • Export Citation
  • Chaumerliac, N., E. Richard, and J. Pinty, 1987: Sulfur scavenging in a mesoscale model with quasi-spectral microphysics: two-dimensional results for continental and maritime clouds. J. Geophys. Res., 92 , 31143126.

    • Search Google Scholar
    • Export Citation
  • Clark, T. L., and R. Farley, 1984: Severe downslope windstorm calculations in two and three spatial dimensions using anelastic interactive grid nesting: A possible mechanism for gustiness. J. Atmos. Sci., 41 , 329350.

    • Search Google Scholar
    • Export Citation
  • Colle, B. A., 2004: Sensitivity of orographic precipitation to changing ambient conditions and terrain geometries: An idealized modeling perspective. J. Atmos. Sci., 61 , 588606.

    • Search Google Scholar
    • Export Citation
  • Cozic, J., B. Verheggen, S. Mertes, P. Connolly, K. Bower, A. Petzold, U. Baltensperger, and E. Weingartner, 2007: Scavenging of black carbon in mixed phase clouds at the high alpine site Jungfraujoch. Atmos. Chem. Phys., 7 , 17971807.

    • Search Google Scholar
    • Export Citation
  • Cozic, J., and Coauthors, 2008: Chemical composition of free tropospheric aerosol for PM1 and coarse mode at the high alpine site Jungfraujoch. Atmos. Chem. Phys., 8 , 407423.

    • Search Google Scholar
    • Export Citation
  • Cziczo, D. J., D. Murphy, P. Hudson, and D. Thomson, 2004: Single-particle measurements of the chemical composition of cirrus ice residue during CRYSTAL-FACE. J. Geophys. Res., 109 .D04201, doi:10.1029/2003JD004032.

    • Search Google Scholar
    • Export Citation
  • Davies, H., 1976: A lateral boundary formulation for multi-level prediction models. Quart. J. Roy. Meteor. Soc., 102 , 405418.

  • Denman, K. L., and Coauthors, 2007: Couplings between changes in the climate system and biogeochemistry. Climate Change 2007: The Physical Science Basis, S. Soloman et al., Eds., Cambridge University Press, 499–588.

    • Search Google Scholar
    • Export Citation
  • Doms, G., and U. Schättler, 2002: A description of the nonhydrostatic regional model LM. Part I: Dynamics and numerics. Deutscher Wetterdienst Tech. Rep., 134 pp.

  • Durran, D., and J. Klemp, 1983: A compressible model for the simulation of moist mountain waves. Mon. Wea. Rev., 111 , 23412361.

  • Ferek, R. J., D. A. Hegg, P. V. Hobbs, P. Durkee, and K. Nielsen, 1998: Measurements of ship-induced tracks in clouds off the Washington coast. J. Geophys. Res., 103 , D18. 2319923206.

    • Search Google Scholar
    • Export Citation
  • Fuhrer, O., and C. Schär, 2005: Embedded cellular convection in moist flow past topography. J. Atmos. Sci., 62 , 28102828.

  • Ghan, S., C. C. Chuang, and J. E. Penner, 1993: A parameterization of cloud droplet nucleation. Part I: Single aerosol type. Atmos. Res., 30 , 197221.

    • Search Google Scholar
    • Export Citation
  • Givati, A., and D. Rosenfeld, 2004: Quantifying precipitation suppression due to air pollution. J. Appl. Meteor., 43 , 10381056.

  • Givati, A., and D. Rosenfeld, 2005: Separation between cloud-seeding and air pollution effects. J. Appl. Meteor., 44 , 12981314.

  • Herzog, H-J., G. Vogel, and U. Schubert, 2002: LLM—A nonhydrostatic model applied to high-resolving simulations of turbulent fluxes over heterogeneous terrain. Theor. Appl. Climatol., 73 , 6786.

    • Search Google Scholar
    • Export Citation
  • Hobbs, P., R. Easter, and A. Fraser, 1973: A theoretical study of the flow of air and fallout of solid precipitation over mountainous terrain: Part II. Microphysics. J. Atmos. Sci., 30 , 813823.

    • Search Google Scholar
    • Export Citation
  • Houze, R., and S. Medina, 2005: Turbulence as a mechanism for orographic precipitation enhancement. J. Atmos. Sci., 62 , 35993623.

  • Hueglin, C., R. Gehrig, U. Baltensperger, M. Gysel, C. Monn, and H. Vonmont, 2005: Chemical characterisation of PM2.5, PM10 and coarse particles at urban, near-city and rural sites at Switzerland. Atmos. Environ., 39 , 637651.

    • Search Google Scholar
    • Export Citation
  • Jaenicke, R., 1993: Tropospheric aerosols. Aerosol–Cloud–Climate Interactions, P. V. Hobbs, Ed., Academic Press, 1–31.

  • Jiang, Q., 2003: Moist dynamics and orographic precipitation. Tellus, 55A , 301316.

  • Jiang, Q., 2006: Precipitation over concave terrain. J. Atmos. Sci., 63 , 22692288.

  • Jiang, Q., 2007: Precipitation over multiscale terrain. Tellus, 59A , 321335.

  • Jiang, Q., and R. Smith, 2003: Cloud timescales and orographic precipitation. J. Atmos. Sci., 60 , 15431559.

  • Jirak, I. L., and W. R. Cotton, 2006: Effect of air pollution on precipitation along the front range of the Rocky Mountains. J. Appl. Meteor., 45 , 236245.

    • Search Google Scholar
    • Export Citation
  • Khairoutdinov, M., and Y. Kogan, 2000: A new cloud physics parameterization in a large-eddy simulation model of marine stratocumulus. Mon. Wea. Rev., 128 , 229243.

    • Search Google Scholar
    • Export Citation
  • Kirshbaum, D. J., and D. R. Durran, 2004: Factors governing cellular convection in orographic precipitation. J. Atmos. Sci., 61 , 682698.

    • Search Google Scholar
    • Export Citation
  • Leaitch, W., C. Banic, G. Isaac, M. Couture, P. Liu, I. Gultepe, and S-M. Li, 1996: Physical and chemical observations in marine stratus during the 1993 North Atlantic Regional Experiment: Factors controlling cloud droplet number concentrations. J. Geophys. Res., 101 , D22. 2912329135.

    • Search Google Scholar
    • Export Citation
  • Lin, H., and R. Leaitch, 1997: Development of an in-cloud aerosol activation parameterization for climate modelling. Proc. Workshop on Measurements of Cloud Properties for Forecasts of Weather and Climate, Mexico City, Mexico, WMO, 328–335.

    • Search Google Scholar
    • Export Citation
  • Lohmann, U., 2002: Possible aerosol effects on ice clouds via contact nucleation. J. Atmos. Sci., 59 , 647656.

  • Lohmann, U., and B. Kärcher, 2002: First interactive simulations of cirrus clouds formed by homogeneous freezing in the ECHAM general circulation model. J. Geophys. Res., 107 .4105, doi:10.1029/ 2001JD000767.

    • Search Google Scholar
    • Export Citation
  • Lohmann, U., and J. Feichter, 2005: Global indirect aerosol effects: A review. Atmos. Chem. Phys., 4 , 75617614.

  • Lohmann, U., P. Stier, C. Hoose, S. Ferrachat, S. Kloster, E. Roeckner, and J. Zhang, 2007: Cloud microphysics and aerosol indirect effects in the global climate model ECHAM5-HAM. Atmos. Chem. Phys., 7 , 34253446.

    • Search Google Scholar
    • Export Citation
  • Miglietta, M. M., and R. Rotunno, 2005: Simulations of moist nearly neutral flow over a ridge. J. Atmos. Sci., 62 , 14101427.

  • Murphy, D. M., and Coauthors, 2006: Single-particle mass spectrometry of tropospheric aerosol particles. J. Geophys. Res., 111 .D23S32, doi:10.1029/2006JD007340.

    • Search Google Scholar
    • Export Citation
  • Peng, Y., U. Lohmann, R. Leaitch, C. Banic, and M. Couture, 2002: The cloud albedo–cloud droplet effective radius relationship for clean and polluted clouds from RACE and FIRE.ACE. J. Geophys. Res., 107 .4106, doi:10.1029/2000JD000281.

    • Search Google Scholar
    • Export Citation
  • Pinsky, M., A. Khain, and M. Shapiro, 2001: Collision efficiency of drops in a wide range of Reynolds numbers: Effect of pressure on spectrum evolution. J. Atmos. Sci., 58 , 742764.

    • Search Google Scholar
    • Export Citation
  • Pruppacher, H., and J. Klett, 1997: Microphysics of Clouds and Precipitation. 2nd ed. Kluwer Academic, 954 pp.

  • Roe, G. H., 2005: Orographic precipitation. Annu. Rev. Earth Planet. Sci., 33 , 645671.

  • Rosenfeld, D., and A. Givati, 2006: Evidence of orographic precipitation suppression by air pollution–induced aerosols in the western united states. J. Appl. Meteor., 45 , 893911.

    • Search Google Scholar
    • Export Citation
  • Rosenfeld, D., J. Dai, X. Yu, Z. Yao, X. Xu, X. Yang, and C. Du, 2007: Inverse relations between amounts of air pollution and orographic precipitation. Science, 315 , 396398.

    • Search Google Scholar
    • Export Citation
  • Rotunno, R., and R. A. Houze, 2007: Lessons on orographic precipitation from the Mesoscale Alpine Programme. Quart. J. Roy. Meteor. Soc., 133 , 811830.

    • Search Google Scholar
    • Export Citation
  • Seifert, A., and K. D. Beheng, 2006: A two-moment cloud microphysics parameterization for mixed-phase clouds. Part I: Model description. Meteor. Atmos. Phys., 92 , 4566.

    • Search Google Scholar
    • Export Citation
  • Seinfeld, J., and S. Pandis, 1998: Atmospheric Chemistry and Physics. John Wiley, 1326 pp.

  • Smith, R., 1979: The influence of mountains on the atmosphere. Advances in Geophysics, Vol. 21, Academic Press, 87–230.

  • Smith, R., 1980: Linear theory of stratified hydrostatic flow past an isolated mountain. Tellus, 32 , 348364.

  • Smith, R., 1989a: Hydrostatic airflow over mountains. Advances in Geophysics, Vol. 31, 1–41.

  • Smith, R., 1989b: Mechanisms of orographic precipitation. Meteor. Mag., 119 , 8588.

  • Smith, R., 2003: A linear upslope time-delay model for orographic precipitation. J. Hydrol., 282 , 29.

  • Smith, R., and I. Barstad, 2004: A linear theory of orographic precipitation. J. Atmos. Sci., 61 , 13771391.

  • Smolarkiewicz, P., R. Rasmussen, and T. Clark, 1988: On the dynamics of hawaiian cloud bands: Island forcing. J. Atmos. Sci., 45 , 18721905.

    • Search Google Scholar
    • Export Citation
  • Solomon, S., and Coauthors, 2007: Climate Change 2007: The Physical Science Basis. Cambridge University Press, 996 pp.

  • Spichtinger, P., 2004: Eisübersättigte Regionen. Ph.D. thesis, Deutsches Zentrum für Luft- und Raumfahrt, 208 pp.

  • Steppeler, J., G. Doms, U. Schättler, H. Bitzer, A. Gassmann, U. Damrath, and G. Gregoric, 2003: Meso-gamma scale forecasts using the nonhydrostatic model LM. Meteor. Atmos. Phys., 82 , 7596.

    • Search Google Scholar
    • Export Citation
  • Stier, P., and Coauthors, 2005: The aerosol–climate model ECHAM5-HAM. Atmos. Chem. Phys., 5 , 11251156.

  • Twomey, S., M. Piepgrass, and T. Wolfe, 1984: An assessment of the impact of pollution on global cloud albedo. Tellus, 36B , 356366.

  • Vehameki, H., M. Kulmala, I. Napari, K. Lehtinen, C. Timmreck, M. Noppel, and A. Laaksonen, 2002: An improved parameterization for sulfuric acid–water nucleation rates for tropospheric and stratospheric conditions. J. Geophys. Res., 107 .4622, doi:10.1029/2002JD002184.

    • Search Google Scholar
    • Export Citation
  • Verheggen, B., and Coauthors, 2007: Aerosol partitioning between the interstitial and the condensed phase in mixed-phase clouds. J. Geophys. Res., D23202, doi:10.1029/2007JD008714.

    • Search Google Scholar
    • Export Citation
  • Vignati, E., J. Wilson, and P. Stier, 2004: M7: An efficient size-resolved aerosol microphysics module for large-scale aerosol transport models. J. Geophys. Res., 109 .D22202, doi:10.1029/2003JD004485.

    • Search Google Scholar
    • Export Citation
  • Weingartner, E., S. Nyeki, and U. Baltensperger, 1999: Seasonal and diurnal variation of aerosol size distributions (10 < d < 750 nm) at a high-alpine site (Jungfraujoch 3580 m asl). J. Geophys. Res., 104 , D21. 2680926820.

    • Search Google Scholar
    • Export Citation
  • Weingartner, E., M. Gysel, and U. Baltensperger, 2002: Hygroscopicity of aerosol particles at low temperatures. 1. New low-temperature h-tdma instrument: Setup and first applications. Environ. Sci. Technol., 36 , 5562.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

Vertical profiles of temperature (solid thick line) and dewpoint temperature (dashed thick line). The parameters describing the analytically given profiles are Nd = 0.0115 s−1, TSL = 285 K, pSL = 1000 hPa, and RHSL = 0.8. The lifting condensation level (LCL) is located at approximately zLCL ≈ 425 m.

Citation: Journal of the Atmospheric Sciences 65, 8; 10.1175/2007JAS2492.1

Fig. 2.
Fig. 2.

Initial aerosol size spectra for the clean case (solid lines) and polluted case (dashed lines) at the surface. Size distributions for (a) the number density (cm−3) and (b) the mass distributions (μg m−3).

Citation: Journal of the Atmospheric Sciences 65, 8; 10.1175/2007JAS2492.1

Fig. 3.
Fig. 3.

(a) Initial profiles of the gas-phase sulfuric acid concentration, (b) the number density of sulfate aerosols in nucleation mode, and (c) the mass density of nucleation mode sulfate aerosols (c). Note that the number and mass distributions of the gas-phase sulfuric acid and the nucleation mode aerosols are all in (molecules/particles) per cubic centimeter.

Citation: Journal of the Atmospheric Sciences 65, 8; 10.1175/2007JAS2492.1

Fig. 4.
Fig. 4.

Potential temperature θ (K; contour interval 1 K), equivalent potential temperature θe (K; contour interval 1 K) and horizontal wind speed U (m s−1; contour interval 1 m s−1) for (a), (b) the dry simulation and (c), (d) the moist simulation in experiment A after 10 h. Only part of the computational domain is shown.

Citation: Journal of the Atmospheric Sciences 65, 8; 10.1175/2007JAS2492.1

Fig. 5.
Fig. 5.

Vertical cross section of (a), (b) the cloud water mixing ratio (g kg−1; contour interval 0.05 g kg−1) and (c), (d) for the rain water mixing ratio (g kg−1; contour interval 0.02 g kg−1) for experiment A after 10 h. The clean case simulation is shown in the left column and the polluted case simulation is shown in the right column. Only part of the computational domain is shown.

Citation: Journal of the Atmospheric Sciences 65, 8; 10.1175/2007JAS2492.1

Fig. 6.
Fig. 6.

The precipitation distribution [mm (10 h)−1] for experiment A for (a) the narrow mountain with a = 10 km and (b) the wide mountain with a = 30 km after 10 h, for both the clean case (solid) and the polluted case (dashed). The topographic surface contour for (c) the narrow mountain and (d) the wide mountain. The location of the mountain peak is indicated by the triangle.

Citation: Journal of the Atmospheric Sciences 65, 8; 10.1175/2007JAS2492.1

Fig. 7.
Fig. 7.

Difference in the precipitation distribution [polluted case minus clean case; mm (10 h)−1] along the idealized topography after 10 h for experiment A and different mountain half-widths. The location of the mountain peak is indicated by the triangle.

Citation: Journal of the Atmospheric Sciences 65, 8; 10.1175/2007JAS2492.1

Fig. 8.
Fig. 8.

Difference fields for experiment A after 10 h. The differences (polluted case − clean case) for (a) the water vapor mixing ratio qυ (contour interval 0.04 g kg−1), (b) the potential temperature θ (contour interval 0.1 K), (c) the horizontal wind speed U (contour interval 0.1 m s−1) and (d) the vertical velocity (contour interval 0.02 m s−1). Only part of the domain is plotted.

Citation: Journal of the Atmospheric Sciences 65, 8; 10.1175/2007JAS2492.1

Fig. 9.
Fig. 9.

For experiment B after 10 h, the (a) potential temperature θ (K) and (b) horizontal velocity U (m s−1) for the dry simulation; (c) equivalent potential temperature θe (K) and (d) U (m s−1) for the moist simulation. The contour intervals are (a), (c) 1 K and (b), (d) 2 m s−1. Only part of the computational domain is shown.

Citation: Journal of the Atmospheric Sciences 65, 8; 10.1175/2007JAS2492.1

Fig. 10.
Fig. 10.

Same as Fig. 5, but for experiment B after 10 h and a rain water mixing ratio contour interval of 0.002 g kg−1.

Citation: Journal of the Atmospheric Sciences 65, 8; 10.1175/2007JAS2492.1

Fig. 11.
Fig. 11.

Same as in Fig. 7 but for experiment B.

Citation: Journal of the Atmospheric Sciences 65, 8; 10.1175/2007JAS2492.1

Fig. 12.
Fig. 12.

Equivalent potential temperature field θe (K; contour interval 1 K) and horizontal velocity U (m s−1; contour interval 2 m s−1) in experiment C after 10 h for (a), (b) the clean case and (c), (d) the polluted case simulation. The thick contour indicates zero wind speed. Only part of the computational domain is shown.

Citation: Journal of the Atmospheric Sciences 65, 8; 10.1175/2007JAS2492.1

Fig. 13.
Fig. 13.

Same as in Fig. 5, but for experiment C after 10 h and for a cloud water mixing ratio interval of 0.1 g kg−1 and a rain water mixing ratio of 0.05 g kg−1. Only part of the computational domain is shown.

Citation: Journal of the Atmospheric Sciences 65, 8; 10.1175/2007JAS2492.1

Fig. 14.
Fig. 14.

Same as in Fig. 6 but for experiment C.

Citation: Journal of the Atmospheric Sciences 65, 8; 10.1175/2007JAS2492.1

Fig. 15.
Fig. 15.

Same as in Fig. 7 but for experiment C.

Citation: Journal of the Atmospheric Sciences 65, 8; 10.1175/2007JAS2492.1

Table 1.

The underlying number densities Ni and mass densities Mi,j for the initial aerosol size distributions shown in Fig. 2 together with the standard deviation of each lognormal mode. The abbreviations denote the background state (bg), the clean case (cc), and the polluted case (pc). Note that the aerosol concentrations in the clean case as well as in the polluted case are added on top of the aerosol concentrations of the background state.

Table 1.
Table 2.

Local budgets for upslope, crest, downslope, and total precipitation as well as the spillover factor after 10 h. Units are mm (10 h)−1 for the precipitation values. All precipitation values are rounded to the first decimal place.

Table 2.
Save