• Berry, E. X., and R. L. Reinhardt, 1974: An analysis of cloud drop growth by collection. Part IV: A new parameterization. J. Atmos. Sci, 31 , 21272135.

    • Search Google Scholar
    • Export Citation
  • Bosart, L. F., 2003: Whither the weather analysis and forecasting process? Wea. Forecasting, 18 , 520529.

  • Braun, S. A., and W-K. Tao, 2000: Sensitivity of high-resolution simulations of Hurricane Bob (1991) to planetary boundary layer parameterizations. Mon. Wea. Rev, 128 , 39413961.

    • Search Google Scholar
    • Export Citation
  • Brown, P. R. A., and H. A. Swann, 1997: Evaluation of key microphysical parameters in three-dimensional cloud-model simulation using aircraft and multiparameter radar data. Quart. J. Roy. Meteor. Soc, 123 , 22452275.

    • Search Google Scholar
    • Export Citation
  • Bruintjes, R. T., T. L. Clark, and W. D. Hall, 1994: Interactions between topographic airflow and cloud/precipitation development during the passage of a winter storm in Arizona. J. Atmos. Sci, 51 , 4867.

    • Search Google Scholar
    • Export Citation
  • Burrows, D. A., 1992: Evaluation of a two-dimensional kinematic cloud model using data from a central Sierra Nevada orographic cloud system. J. Appl. Meteor, 31 , 5163.

    • Search Google Scholar
    • Export Citation
  • Charba, J. P., D. W. Reynolds, B. E. McDonald, and G. M. Carter, 2003: Comparative verification of recent quantitative precipitation forecasts in the National Weather Service: A simple approach for scoring forecast accuracy. Wea. Forecasting, 18 , 161183.

    • 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
  • Colle, B. A., and C. F. Mass, 2000: The 5–9 February 1996 flooding event over the Pacific Northwest: Sensitivity studies and evaluation of the MM5 precipitation forecasts. Mon. Wea. Rev, 128 , 593617.

    • Search Google Scholar
    • Export Citation
  • Colle, B. A., and Y. Zeng, 2004: Bulk microphysical sensitivities within the MM5 for orographic precipitation. Part II: Impact of barrier width and freezing level. Mon. Wea. Rev.,132, 2802–2815.

    • Search Google Scholar
    • Export Citation
  • Colle, B. A., K. J. Westrick, and C. F. Mass, 1999: Evaluation of the MM5 and Eta-10 precipitation forecasts over the Pacific Northwest during the cool season. Wea. Forecasting, 14 , 137154.

    • Search Google Scholar
    • Export Citation
  • Colle, B. A., C. F. Mass, and K. J. Westrick, 2000: MM5 precipitation verification over the Pacific Northwest during the 1997–1999 cool seasons. Wea. Forecasting, 15 , 730744.

    • Search Google Scholar
    • Export Citation
  • Cooper, W. A., 1986: Ice initiation in natural clouds. Precipitation Enhancement—A Scientific Challenge, Meteor. Monogr., No. 21, Amer. Meteor. Soc., 29–32.

    • Search Google Scholar
    • Export Citation
  • Cox, G. P., 1988: Modeling precipitation in frontal rainbands. Quart. J. Roy. Meteor. Soc, 114 , 115127.

  • Doyle, J. D., and Coauthors, 2000: An intercomparison of model-predicted wave breaking for the 11 January 1972 Boulder windstorm. Mon. Wea. Rev, 128 , 901914.

    • Search Google Scholar
    • Export Citation
  • Dudhia, J., 1989: Numerical study of convection observed during the winter monsoon experiment using a mesoscale two-dimensional model. J. Atmos. Sci, 46 , 30773107.

    • Search Google Scholar
    • Export Citation
  • Ferrier, B. S., 1994: A double-moment, multiple-phase, four-class bulk ice scheme. Part I: Description. J. Atmos. Sci, 51 , 249280.

  • Ferrier, B. S., W-K. Tao, and J. Simpson, 1995: A double-moment, multiple-phase, four-class bulk ice scheme. Part II: Simulations of convective storms in different large-scale environments and comparisons with other bulk parameterizations. J. Atmos. Sci, 52 , 10011033.

    • Search Google Scholar
    • Export Citation
  • Flatau, P. J., G. J. Tripoli, J. Verlinde, and W. R. Cotton, 1989: The CSU–RAMS cloud microphysical module: General theory and code documentation. Atmospheric Science Paper 451, 88 pp. [Available from Colorado State University, Dept. of Atmospheric Science, Fort Collins, CO 80523.].

    • Search Google Scholar
    • Export Citation
  • Fletcher, N. H., 1962: Physics of Rain Clouds. Cambridge University Press, 386 pp.

  • Fritsch, J. M., and Coauthors, 1998: Quantitative precipitation forecasting: Report of the eighth prospectus development team, U.S. Weather Research Program. Bull. Amer. Meteor. Soc, 79 , 285299.

    • Search Google Scholar
    • Export Citation
  • Gaudet, B., and W. R. Cotton, 1998: Statistical characteristics of real-time precipitation forecasting model. Wea. Forecasting, 13 , 966982.

    • Search Google Scholar
    • Export Citation
  • Grell, G. A., J. Dudhia, and D. R. Stauffer, 1994: A description of the fifth-generation Penn State/NCAR Mesoscale Model (MM5). NCAR Tech. Note NCAR/TN-398+STR, 138 pp. [Available from National Center for Atmospheric Research, P.O. Box 3000, Boulder, CO 80307.].

    • Search Google Scholar
    • Export Citation
  • Hallet, M. F., and S. C. Mossop, 1974: Production of secondary ice particles during the riming process. Nature, 249 , 2628.

  • Hobbs, P. V., 1975: The nature of winter clouds and precipitation in the Cascade Mountains and their modification by artificial seeding. Part I: Natural conditions. J. Appl. Meteor, 14 , 783804.

    • Search Google Scholar
    • Export Citation
  • Hobbs, P. V., and P. O. G. Persson, 1982: The mesoscale and microscale structure and organization of clouds and precipitation in midlatitude cyclones. Part V: The substructure of narrow cold-frontal rainbands. J. Atmos. Sci, 39 , 280295.

    • Search Google Scholar
    • Export Citation
  • Hobbs, P. V., T. J. Matejka, P. H. Herzegh, J. D. Locatelli, and R. A. Houze Jr., 1980: The mesoscale and microscale structure and organization of clouds and precipitation in midlatitude cyclones. I: A case study of a cold front. J. Atmos. Sci, 37 , 568596.

    • Search Google Scholar
    • Export Citation
  • Hong, S-Y., and H-L. Pan, 1996: Nonlocal boundary layer vertical diffusion in a medium-range forecast model. Mon. Wea. Rev, 124 , 23222339.

    • Search Google Scholar
    • Export Citation
  • Houze, R. A., P. V. Hobbs, and P. H. Herzegh, 1979: Size distributions of precipitation particles in frontal clouds. J. Atmos. Sci, 36 , 156162.

    • Search Google Scholar
    • Export Citation
  • Hsie, E. Y., R. A. Anthes, and D. Keyser, 1984: Numerical simulation of frontogenesis in a moist atmosphere. J. Atmos. Sci, 41 , 25812594.

    • Search Google Scholar
    • Export Citation
  • Jiang, Q., 2003: Moist dynamics and orographic precipitation. Tellus, 55A , 301316.

  • Jiang, Q., and R. B. Smith, 2003: Cloud timescales and orographic precipitation. J. Atmos. Sci, 60 , 11591172.

  • Kessler, E., 1969: On the Distribution and Continuity of Water Substance in Atmospheric Circulations. Meteor. Monogr., No. 32, Amer. Meteor. Soc., 84 pp.

    • Search Google Scholar
    • Export Citation
  • Klemp, J. B., and D. R. Durran, 1983: An upper boundary condition permitting internal gravity wave radiation in numerical mesoscale models. Mon. Wea. Rev, 111 , 430444.

    • Search Google Scholar
    • Export Citation
  • Lin, Y. L., R. Farley, and H. D. Orville, 1983: Bulk parameterization of the snow field in a cloud model. J. Climate Appl. Meteor, 22 , 10651092.

    • Search Google Scholar
    • Export Citation
  • Locatelli, J. D., and P. V. Hobbs, 1974: Fall speeds and masses of solid precipitation particles. J. Geophys. Res, 79 , 21852197.

  • Meyers, M. P., and W. R. Cotton, 1992: Evaluation of the potential for wintertime quantitative precipitation forecasting over mountainous terrain with an explicit cloud model. Part I: Two-dimensional sensitivity experiments. J. Appl. Meteor, 31 , 2650.

    • Search Google Scholar
    • Export Citation
  • Meyers, M. P., P. J. Demott, and W. R. Cotton, 1992: New primary ice-nucleation parameterization in an explicit cloud model. J. Appl. Meteor, 31 , 708721.

    • Search Google Scholar
    • Export Citation
  • Rasmussen, R. M., and Coauthors, 1992: Winter Icing and Storm Projects (WISP). Bull. Amer. Meteor. Soc, 73 , 951974.

  • Rasmussen, R. M., I. Geresdi, G. Thompson, K. Manning, and E. Karplus, 2002: Freezing drizzle formation in stably stratified layer clouds: The role of radiative cooling of cloud droplets, cloud condensation nuclei, and ice initiation. J. Appl. Meteor, 59 , 837860.

    • Search Google Scholar
    • Export Citation
  • Rauber, R. M., 1992: Microphysical structure and evolution of a Sierra Nevada shallow orographic cloud system. J. Appl. Meteor, 31 , 324.

    • Search Google Scholar
    • Export Citation
  • Reisner, J., R. M. Rasmussen, and R. T. Bruintjes, 1998: Explicit forecasting of supercooled liquid water in winter storm using the MM5 mesoscale model. Quart. J. Roy. Meteor. Soc, 124 , 10711107.

    • Search Google Scholar
    • Export Citation
  • Reynolds, D. W., and A. S. Dennis, 1986: A review of the Sierra Cooperative Pilot Project. Bull. Amer. Meteor. Soc, 67 , 513523.

  • Rutledge, S. A., and P. V. Hobbs, 1983: The mesoscale and microscale structure and organization of clouds and precipitation in mid-latitude cyclones: A model for the “seeder-feeder” process in warm-frontal rainbands. J. Atmos. Sci, 40 , 11851206.

    • Search Google Scholar
    • Export Citation
  • Rutledge, S. A., and P. V. Hobbs, 1984: The mesoscale and microscale structure and organization of clouds and precipitation in mid-latitude cyclones: A diagnostic modeling study of precipitation development in narrow cold-frontal rainbands. J. Atmos. Sci, 41 , 29492972.

    • Search Google Scholar
    • Export Citation
  • Schultz, P., 1995: An explicit cloud physics parameterization for operational numerical weather prediction. Mon. Wea. Rev, 123 , 33313343.

    • Search Google Scholar
    • Export Citation
  • Sekhon, R. S., and R. C. Srivastava, 1970: Snow size spectra and radar reflectivity. J. Atmos. Sci, 27 , 299307.

  • Stoelinga, M., and Coauthors, 2003: Improvement of microphysical parameterizations through observational verification experiments (IMPROVE). Bull. Amer. Meteor. Soc, 84 , 18071826.

    • Search Google Scholar
    • Export Citation
  • Tao, W-K., and J. Simpson, 1993: The Goddard Cumulus Ensemble Model. Part I: Model description. Terr. Atmos. Oceanic Sci, 4 , 3572.

  • Thompson, G., R. M. Rasmussen, and K. Manning, 2004: Explicit forecasts of winter precipitation using an improved bulk microphysics scheme. Part I: Description and sensitivity analysis. Mon. Wea. Rev, 132 , 519542.

    • Search Google Scholar
    • Export Citation
  • Twomey, S., and T. A. Wojciechowski, 1969: Observations of the geographic variation of cloud nuclei. J. Atmos. Sci, 26 , 684688.

  • View in gallery

    Conceptual model of the microphysical processes over the Sierra Nevada on 12 Feb 1986 (from Fig. 18 of Rauber 1992). Arrows indicate cloud droplet (C), needle (N), and dendritic particle (D) trajectories in the plane of the cross section. The few key isotherms (dashed) are also shown

  • View in gallery

    The terrain profile across the coastal range and Sierra Nevada used in this study. The two bold arrows outline the position of the box used for the microphysical budget, and SH is the location of the Sheridan, CA, sounding used in this study

  • View in gallery

    The 1500 UTC 12 Feb 1986 Sheridan sounding used in this study (see SH on Fig. 2) showing temperatures (black) and dewpoint (gray). The geostrophic winds (one full barb = 10 m s−1) were obtained from the 1200 UTC 12 Feb NCEP synoptic charts

  • View in gallery

    Microphysical flowchart for the Reisner2 scheme. The circles represent the various water species (water vapor, cloud water, cloud ice, rain, snow, and graupel), and the arrows are the processes that link the species (see appendix for the list of processes)

  • View in gallery

    Cross section at hour 6 showing (a) rain (thin solid every 0.04 g kg−1), snow (gray), and graupel (black solid) every 0.08 g kg−1, and wind vectors at 4-km grid spacing (run 3 in Table 1). (b) Same as (a), except for 2-km grid spacing (run 1). (c) Same as (a), except for 1.3-km grid spacing (run 2)

  • View in gallery

    Accumulated 6–12-h surface precipitation (mm) for the 4‐, 2-, and 1.33-km horizontal grid spacings. The terrain is also shown for reference

  • View in gallery

    Cross section for the control run (run 1) at hour 6 showing (a) u (normal) wind component (every 2.0 m s−1) and (b) υ (parallel) wind component (every 2.0 m s−1), and (c) cloud water (solid black every 0.08 g kg−1), cloud ice (solid gray every 0.02 g kg−1), and circulation vectors

  • View in gallery

    Cross section showing the distribution of (a) rain (thin solid), snow (gray), and graupel (thick black) for the Reisner2 scheme. (b) Same as (a), except for cloud water (black) and cloud ice (gray). The contour interval is 0.04 g kg−1 for rain, 0.08 g kg−1 for snow, 0.08 g kg−1 for graupel, 0.08 g kg−1 for cloud water, and 0.02 g kg−1 for cloud ice. (c), (d) Same as (a), (b), except for the warm-rain scheme; (e), (f) same as (a), (b), except for the simple ice scheme; (g), (h) same as (a), (b), except for the Reisner1 scheme; (i), (j) same as (g), (h), except for the Goddard scheme; (k), (l) same as (g), (h), except for the Schultz scheme

  • View in gallery

    (Continued )

  • View in gallery

    Surface-accumulated (6–12 h) precipitation (mm) vs distance for the different microphysical schemes: Reisner2 (bold solid), warm rain (dotted), simple ice (dashed), Reisner1 (solid), Goddard (dash–dot–dot–dot), and Schultz (long dash). The observed precipitation profile over a portion of the Sierra Nevada (derived from Rauber 1992) is shown by the thick gray line

  • View in gallery

    Flowchart of the microphysical processes between 6–12 h of the control run (run 1) for the box in Fig. 2. The values shown are the ratio of each microphysical process rate to the total WVL rate (cond + sdep + gdep + idsn + idep) within the box. The processes are listed in the appendix. The sum of all the microphysical process tendencies for each species is given by (wv:, cw:, r:, ci:, g: and s:). This sum does not include horizontal advection and diffusion/ divergence, which are labeled as hadv and other, respectively. The fallout tendency of rain (rprc), snow (sprc), graupel (gprc), and cloud ice (iprc) are also shown. Microphysical processes greater than 10% of the WVL rate are in bold

  • View in gallery

    Cross section of the spatial distribution of the major microphysical processes contributing to snow production, showing the ratio of (a) sdep (gray), icns (black), and ssacr (dashed gray), (b) saci (gray), ssacw (dashed gray), and siacr (black) to the total snow production rate from these processes at each point contoured every 20%. (c), (d) Same as (a), except for graupel production showing (c) gdep (gray), gsacw (dashed gray), and gacw (black), and (d) scng (black) and gacr (gray). (e), (f) Same as (a), except rain production showing (e) racw (gray) and racs (black), and (f) gmlt (black), ccnr (dashed gray), and racs (gray)

  • View in gallery

    (a) Cross section difference between run 9 (Fletcher ice initiation) and run 1 (Cooper ice initiation) (run 9 − run 1) showing cloud ice (black) every 0.01 g kg−1 and snow (gray) every 0.01 g kg−1, with negative values dashed. (b) Same as (a), except the difference between run 11 (no ice production above 7 km) and run 1 (run 11 − run 1) showing cloud ice (black) every 0.02 g kg−1 and snow (gray) every 0.02 g kg−1

  • View in gallery

    (a) Cross-section difference between run 12 (larger CNP) and run 1 (run 12 − run 1) showing snow (gray) and graupel (bold) every 0.04 g kg−1, and rain (thin solid) every 0.02 g kg−1, with negative values dashed. (b) Same as (a), except for run 13 (smaller CNP) minus run 1

  • View in gallery

    Accumulated 6–12-h surface precipitation (mm) for the different CNP experiments

  • View in gallery

    (a) Cross-section difference between run 14 (fixed Nos) and run 1 (Nosqs) (run 14 − run 1) showing snow (gray) every 0.08 g kg−1 and graupel (bold) every 0.04 g kg−1, and rain (thin solid) every 0.02 g kg−1, with negative values dashed. (b) Same as (a), except for run 15 (NosT) minus run 1

  • View in gallery

    (a) Accumulated 6–12-h surface precipitation (mm) for different Nos experiments. (b) Same as (a), except for the snow fall speed (Vs) experiments

  • View in gallery

    (a) Same as Fig. 10, except showing the difference between (a) run 14 (fixed Nos) and run 1 (Nosqs) (run 14 − run 1). Each process was normalized by the average water vapor loss rate for the two runs before taking the difference. (b) Same as (a), except for run 15 (NosT) minus run 1. Those differences greater than 1% of the water vapor loss rate are in bold.

  • View in gallery

    Snow fall velocity (m s−1) vs diameter of snow (mm) for different Vs schemes. A dotted line is for the control (run 1), dashed line for Ferrier (run 16), dash–dot for Brown and Swann (run 18), and dash–dot–dot–dot for Cox (run 17)

  • View in gallery

    Cross section difference between run 16 (Ferrier Vs) and run 1 (run 16 − run 1) showing snow (gray) every 0.04 g kg−1 and graupel (bold) and rain (thin solid) every 0.02 g kg−1, with negative values dashed

  • View in gallery

    (a) Accumulated 6–12-h surface precipitation (mm) for the different graupel density experiments. (b) Same as (a), except for the graupel fall speed (Vg) experiments

  • View in gallery

    (a) Accumulated 6–12-h surface precipitation (mm) for the different ice crystal autoconversion experiments. (b) Same as (a), except for the cloud water autoconversion experiments

  • View in gallery

    Same as Fig. 10, except for the simple ice BMP. Those processes greater than 5% of the WVL rate are bold. Table 2 lists the process abbreviations

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 150 101 1
PDF Downloads 77 46 0

Bulk Microphysical Sensitivities within the MM5 for Orographic Precipitation. Part I: The Sierra 1986 Event

View More View Less
  • 1 Institute for Terrestrial and Planetary Atmospheres, State University of New York at Stony Brook, Stony Brook, New York
Full access

Abstract

This paper investigates the microphysical pathways and sensitivities within the Reisner2 bulk microphysical parameterization (BMP) of the fifth-generation Pennsylvania State University–National Center for Atmospheric Research (PSU–NCAR) Mesoscale Model (MM5) for a precipitation event over the central Sierra Nevada on 12 February 1986. Using a single sounding initialization, the MM5 was run two-dimensionally at 2-km horizontal grid spacing, which was needed to realistically simulate the embedded convective cells within the orographic cloud. Unlike previous modeling studies of this event, a microphysical budget over the windward slope was calculated for each experiment, in which the importance of each microphysical process was quantified relative to the water vapor loss (WVL) rate. For the control MM5, the largest microphysical processes that contribute to surface precipitation over the Sierra windward slope are condensation (63% of WVL), snow deposition (33%), riming to form graupel (35%), and melting of graupel (28%). The amount of supercooled water aloft is larger than observed and in previous modeling studies of this event using the Regional Atmospheric Modeling System (RAMS). The surface precipitation and microphysical processes over the Sierra Nevada are most sensitive to those parameters associated with the snow distribution, cloud condensation nuclei (CCN) concentrations, and snow/graupel fall speeds, while there is less sensitivity to ice initiation and autoconversions; however, all experiments overpredict the surface precipitation over the windward slope. If ice production is turned off in the cloud-ice region (above 7 km or <250 K), deposition acting on the small amount of cloud ice nucleated at warmer temperatures can still generate a similar snow cloud below 4 km and surface precipitation. The precipitation differences between the BMPs in the MM5 are greater than any single process experiment within Reisner2. The process experiments do help reveal some of the fundamental differences between BMP schemes.

Corresponding author address: Dr. B. A. Colle, Marine Sciences Research Center, State University of New York at Stony Brook, Stony Brook, NY 11794-5000. Email: bcolle@notes.cc.sunysb.edu

Abstract

This paper investigates the microphysical pathways and sensitivities within the Reisner2 bulk microphysical parameterization (BMP) of the fifth-generation Pennsylvania State University–National Center for Atmospheric Research (PSU–NCAR) Mesoscale Model (MM5) for a precipitation event over the central Sierra Nevada on 12 February 1986. Using a single sounding initialization, the MM5 was run two-dimensionally at 2-km horizontal grid spacing, which was needed to realistically simulate the embedded convective cells within the orographic cloud. Unlike previous modeling studies of this event, a microphysical budget over the windward slope was calculated for each experiment, in which the importance of each microphysical process was quantified relative to the water vapor loss (WVL) rate. For the control MM5, the largest microphysical processes that contribute to surface precipitation over the Sierra windward slope are condensation (63% of WVL), snow deposition (33%), riming to form graupel (35%), and melting of graupel (28%). The amount of supercooled water aloft is larger than observed and in previous modeling studies of this event using the Regional Atmospheric Modeling System (RAMS). The surface precipitation and microphysical processes over the Sierra Nevada are most sensitive to those parameters associated with the snow distribution, cloud condensation nuclei (CCN) concentrations, and snow/graupel fall speeds, while there is less sensitivity to ice initiation and autoconversions; however, all experiments overpredict the surface precipitation over the windward slope. If ice production is turned off in the cloud-ice region (above 7 km or <250 K), deposition acting on the small amount of cloud ice nucleated at warmer temperatures can still generate a similar snow cloud below 4 km and surface precipitation. The precipitation differences between the BMPs in the MM5 are greater than any single process experiment within Reisner2. The process experiments do help reveal some of the fundamental differences between BMP schemes.

Corresponding author address: Dr. B. A. Colle, Marine Sciences Research Center, State University of New York at Stony Brook, Stony Brook, NY 11794-5000. Email: bcolle@notes.cc.sunysb.edu

1. Introduction

a. Background

Compared to other forecast parameters such as geopotential height and temperature, quantitative precipitation forecasts (QPFs) have improved more slowly during the past few decades (Bosart 2003). This slow improvement in operational model QPF has been from the lack of horizontal resolution, spatial and timing errors, and uncertainties in moist physical parameterizations. As a result, improving QPF is one of the priorities of the U.S. Weather Research Program (Fritsch et al. 1998).

Precipitation forecasting over mountainous areas is often especially difficult since orographic precipitation is controlled by a number of dynamical and microphysical processes. For example, it is widely known that moist flow ascending a mountain enhances the precipitation along the windward side, but the amount and distribution of orographic precipitation is also affected by thermodynamic stratification, moisture availability, wind profile above the barrier, and hydrometeor advection and generation rates (Colle 2004; Jiang and Smith 2003).

Several high-resolution model simulations (Bruintjes et al. 1994; Gaudet and Cotton 1998; Colle and Mass 2000; Colle et al. 1999, 2000) have shown that orographic precipitation structures become more realistic with increasing model horizontal resolution. As a result of the terrain forcing, the day 3 precipitation forecasts over the western United States are more skillful than day 1 over the eastern United States (Charba et al. 2003). However, Colle et al. (2000) found that mesoscale models when run at high resolution often produce too much precipitation along the windward slope, which they attributed to deficiencies in model microphysical parameterizations. Therefore, more investigation of the microphysics in numerical models is needed. In particular, more research is needed to determine the important processes within a bulk microphysical parameterization (BMP) that model developers and field studies can focus on.

The explicit prediction of clouds and precipitation in numerical models depends on sophisticated BMPs (e.g., Lin et al. 1983; Rutledge and Hobbs 1983, 1984; Reisner et al. 1998). For example, for the fifth-generation Pennsylvania State University–National Center for Atmospheric Research (PSU–NCAR) Mesoscale Model (MM5), there are several BMPs that have become more sophisticated during the past two decades. A warm-rain scheme includes cloud water and rainwater processes, but no ice processes (Hsie et al. 1984). A “simple” ice scheme that includes some snow and ice phase processes has been available for several years (Dudhia 1989), but there is no supercooled water, and snow immediately melts above 0°C. The Reisner1 (or mixed-phase Reisner) scheme in the MM5 allows for supercooled water below 0°C and a melting equation for snow (Grell et al. 1994), but there are no graupel and riming processes. The Reisner2 scheme (Reisner et al. 1998; Thompson et al. 2004) builds on the Reisner1 scheme, but Reisner2 includes graupel and prognostic equations for cloud-ice number concentration. The Goddard scheme is a modification of the Lin et al. (1983) scheme (Tao and Simpson 1993; Braun and Tao 2000), which includes prediction equations for graupel or hail. A more simplified graupel-based scheme has also been developed for real-time operational weather forecasting (Schultz 1995), which runs 20%–30% faster than Reisner2.

The early development of these BMPs utilized datasets from field experiments, such as the Cascade Project (1969–74) over the Cascade Mountains (Hobbs 1975) and Cyclonic Extratropical Storms (CYCLES) from 1980 to 1983 (Hobbs et al. 1980; Hobbs and Persson 1982; among others). The Sierra Cooperative Pilot Project (SCPP), over the central Sierra Nevada of California from 1978 to 1987 (Reynolds and Dennis 1986), provided additional insights into the microphysics and dynamics of orographic precipitation, and the results have been used to evaluate the BMPs within the Regional Atmospheric Modeling System (RAMS) (Burrows 1992; Meyers and Cotton 1992; Meyers et al. 1992). The objective of this paper is to investigate the microphysical sensitivities of an SCPP event from an MM5 BMP perspective.

b. Previous microphysical results from the Sierra project

The SCPP case revisited in this paper occurred on 12 February 1986 (hereafter referred to as the CA86 case), which included a landfalling Pacific front, a terrain-parallel barrier jet to 20 m s−1, and a shallow orographic cloud over the barrier (Rauber 1992, hereafter R92). Using aircraft microphysical data and ground-based radar, R92 described two intersection zones of hydrometeor species over the windward slope during this event (Fig. 1). One region over the middle windward slope involved large cloud droplets rising from cloud base and interacting with the dendrites descending from cloud top, which led to secondary ice production from rime splintering and the formation of graupel. The other region near the barrier crest had enhanced dentritic aggregate production as the crystals produced by secondary ice production interacted with the falling snow crystals. During the few hours of intense observations the orographic cloud became more established (>25 dBZ) over the lower windward slope (cf. Fig. 10 of R92).

Meyers and Cotton (1992, hereafter referred to as MC92) investigated some of the microphysical sensitivities within RAMS for the CA86 case. They found that an increase in graupel density (from 450 to 900 kg m−3) decreased the residence time of graupel aloft, which in turn reduced the graupel mass and total precipitation over the barrier. MC92 found that a reduction in cloud condensation nuclei (CCN) led to slightly more warm-rain precipitation over the barrier, while more CCN resulted in more suspended cloud water, which promoted riming of snow and graupel. MC92 found little sensitivity to ice multiplication from the Hallett– Mossop (H–M) process, which involves ice splinter production during the riming growth of snow and graupel (Hallett and Mossop 1974).

Meyers et al. (1992) examined the primary ice nucleation parameterization for the CA86 case in the RAMS model. They noted that the Fletcher (1962) formulation for deposition and condensation-freezing nucleation overpredicts pristine ice concentration at cold temperatures (<−30°C) and underpredicts ice at warm temperatures (−5° to −20°C). Overall, they showed that the pristine ice fields were very sensitive to the changes in the ice-nucleation formulation, but they did not evaluate the impact on surface precipitation or other microphysical processes.

c. Other microphysical sensitivity studies

More recent field studies have addressed other aspects of microphysical parameterizations. Using data from the Winter Ice Storm Project (WISP 1990) over Colorado (Rasmussen et al. 1992), Reisner et al. (1998) found that the fixed slope intercept for the snow number concentration (Nos) resulted in too much riming and deposition of snow, and underprediction in cloud water. They found that using a diagnostic formulation by Sekhon and Srivastava (1970), which relates Nos to the snowfall rate, slows snow deposition and maintains cloud water mass. Since Nos decreases by two orders of magnitude from −20° to 0°C (Houze et al. 1979), Thompson et al. (2004) implemented a temperature-dependent form of Nos in Reisner2, which mimics the reduction in number concentration by aggregation.

Brown and Swann (1997) examined the effect of the H–M process on a convective cloud simulated over southern England. For this event the H–M process contributed to a majority of ice crystal concentrations in areas of riming between −3° and −8°C. Meanwhile, they found that a reduction in the graupel slope intercept (Nog) reduced the total surface precipitation and decreased the accretion of snow by graupel, while a reduction of graupel fall speed reduced the accumulated surface precipitation.

Colle and Mass (2000) tested the snow fall speed sensitivity for a flooding event over the Pacific Northwest on 5–9 February 1996. The standard MM5 uses the Rutledge and Hobbs (1983) expression for the snow fall speed, which is based on the fall speed for unrimed radiating assemblages of plates, side planes, bullets, and columns (Locatelli and Hobbs 1974). They found that using Cox (1988) fall speed (20% smaller and closer to unrimed assemblages of dendrites) shifted more precipitation from the windward to the lee side of the barrier.

More recently, Rasmussen et al. (2002) found that freezing drizzle formation was very sensitive to ice initiation. In their simulations, the ice initiation scheme of Meyers et al. (1992) produced more ice than the Cooper (1986) scheme. The surface precipitation using Meyers came primarily from rimed crystals, while the Cooper formulation allowed for more freezing drizzle.

d. Motivation and goals

There are many uncertainties in BMPs; therefore, in the late 1990s and early 2000s, a number of field studies were designed to better understand orographic precipitation processes and microphysics. For example, the Improvement of Microphysical Parametrization through Observational Verification Experiment (IMPROVE) along the Washington coast in January 2001 and Oregon Cascades in December 2001 was solely designed to verify and improve BMPs (Stoelinga et al. 2003). In addition to verifying these BMPs using field data, it is also important to understand which parameters are the most important within the BMPs. Even though there are hundreds of users of the MM5, there has been little detailed systematic evaluation of its BMPs or comparison with other models. Many modeling studies, such as Colle and Mass (2000) and MC92, simply compare mixing ratio outputs between the various BMP experiments; therefore, a more comprehensive study is needed to quantify the microphysical budget for a sophisticated BMP in order to determine how a change to a BMP impacts the other microphysical processes.

This paper revisits the CA86 case from an MM5 BMP perspective, in which several parameters within the Reisner2 are systematically evaluated using a microphysical water/ice budget, and the results are compared with other BMPs. A series of MM5 simulations are completed to show the utility of the budget approach for a well-documented event (CA86) as well as address several important questions:

  • What are the most important parameters in the BMPs that result in the greatest precipitation sensitivity for orographic precipitation?

  • What are the important microphysical pathways in the production of precipitation?

  • What are the major differences between MM5 and the well-documented RAMS model BMP for the CA86 case?

  • How do the Reisner2 results compare with other BMPs available in MM5?

Section 2 provides a detailed description of the model, experimental design, and analysis methods. Section 3 shows sensitivity results to horizontal resolution and the various BMPs in the MM5, and the microphysical budget is presented for the control. The microphysical process studies are shown in section 4. A discussion and summary is given in the final section.

2. Experimental design and analysis methods

As in MC92, in order to simplify the evaluation of the BMP sensitivities, such that the ambient winds and temperatures are nearly constant, a two-dimensional version of the MM5 (version 3.6) was applied. Previously, the 2D MM5 has been used to investigate microphysics (Rasmussen et al. 2002; Thompson et al. 2004) and mountain flow dynamics (Doyle et al. 2000; Colle 2004). The MM5 was run nonhydrostatically with 39 sigma levels.1 Figure 2 shows a cross section of the Sierra topography used in this study, which was constructed by interpolating a 30-s topography dataset to an east–west cross section at 4-km grid spacing along 39°N from a sea level point over the Pacific to 50 km east of the Sierra crest. East of this point the terrain height was constant (1700 m). A few different horizontal grid spacings (4, 2, or 1.33 km) were tested using the 4-km terrain, and the model domain was exceptionally long (4000 km) to prevent spurious reflections from the fixed lateral boundary conditions.

The Reisner2 scheme was used for the control BMP with no convective parameterization. The version of Reisner2 is the same version as in Thompson et al. (2004), except that the slope intercept for the snow number concentration for our control run was a function of precipitation rate (Reisner et al. 1998), and the Kessler (1969) autoconversion was used instead of the newly implemented Berry and Reinhardt (1974) scheme. The planetary boundary layer (PBL) was parameterized using the Medium-Range Forecast (MRF) scheme (Hong and Pan 1996), but no heat and moisture surface fluxes were applied. The simulations also did not include radiative effects. An upper-radiative boundary condition (Klemp and Durran 1983) and a sponge layer were applied at the model top (50 mb) to prevent gravity waves from being reflected, while the lateral boundaries were fixed. As in MC92, the simulations included the Coriolis force and were run using the approach outlined in Colle (2004); therefore, a terrain-parallel barrier jet could develop during the simulation.

As in MC92, a sounding from Sheridan, California, on 1500 UTC 12 February 1986 was used to initialize the temperature and moisture profile (Fig. 3). The sounding was saturated up to 750 mb, with a potentially unstable layer between 750 and 625 mb. The initial wind profile came from MC92, which was derived by calculating the geostrophic zonal wind component from the National Centers for Environmental Prediction (NCEP) synoptic charts at 1200 UTC 12 February, and the lateral boundary conditions were fixed using the initial profile. Table 1 lists the experiments used to evaluate the impact of horizontal resolution, parameters within the Reisner2 scheme, and other MM5 BMPs.

One of the important tools for this study is the calculation of a full microphysical budget for a box upstream of the crest (Fig. 2). By quantifying the relationship between water species, one can determine which microphysical process contributes most to the production and destruction of a specific hydrometeor species. Figure 4 and Eqs. (1)–(6) show how the various Reisner2 processes described in the appendix are related through the water vapor (qυ), cloud water (qc), cloud ice (qi), rain (qr), snow (qs), and graupel (qg) mixing ratios. The prognostic equations for each hydrometeor species described in Reisner et al. (1998) are
i1520-0493-132-12-2780-e1
where p* is the pressure difference between the surface and model top.

From Eqs. (1)–(6) there are three contributions to the mixing ratio tendency. The first set of Pxxxx terms on the right-hand side are the contributions from the microphysical processes, while the terms in all capital letters represent advection (horizontal/vertical), divergence, and diffusion. Since the horizontal advection term (ADV) is usually larger than the other terms in italics, it will be noted separately below, while the other italic terms will be lumped together as “other_x.” The double-underlined term (Pxxxx) represents the fallout of a precipitable species.

The conversion rate for each term in Eqs. (1)–(6) was output every 5 min between hours 6 and 12 and integrated for a box upstream of the crest (Fig. 2). In order to determine the relative importance of each process in moving water/ice mass, each process was normalized by the integrated water vapor loss (WVL) within that same box using
i1520-0493-132-12-2780-e7
where Pqqqq(i, k) is the conversion rate of a specific microphysical process averaged for the two adjacent sigma levels, WVL is the sum of Cond + Pidsn + Pidep + Psdep + Pgdep, and Δσ is the sigma-level difference.

3. Grid resolution and MM5 schemes

a. Horizontal resolution

Before comparing the microphysical schemes and processes within the MM5, the sensitivity to horizontal grid spacing was tested using the 4-km terrain and physics (runs 1–3 in Table 1). Unlike the 2- (run 1) and 1.3-km (run 2) grid spacings at hour 6 (Figs. 5b,c), the 4-km simulation (run 3) over the Sierra Nevada could not resolve the convective cells of graupel associated with layer lifting of a potentially unstable layer around 700 mb (Fig. 5a). Meanwhile, there was little difference with resolution over the coastal range. The 1.3-km run was able to resolve more convective plumes than the 2-km run, but the 2-km grid was used for the control run since the microphysical results are similar (see section 3b). Overall, these resolution experiments show that even though 4-km grid spacing can resolve the orographic flows induced by the Sierra Nevada, substantial differences can occur within the cloud at higher resolutions when some embedded convection develops.

Since the convective cores shift randomly during the simulation (not shown), their time-integrated effect on the 6–12-h surface precipitation is less dramatic (Fig. 6). At 4-km grid spacing, there is slightly less surface precipitation at the base of the barrier and 8% more over the upper windward slope than at 2-km spacing. At 1.3-km grid spacing, there is slightly more precipitation immediately upstream of the Sierra Nevada, since it has more elevated convection (Fig. 5c). Meanwhile, the 1.3-km run has less precipitation than the 2-km grid along the windward slope, since the 1.3-km has slightly weaker (0.40 g kg−1) graupel plumes than the 2-km (0.48 g kg−1).

b. Comparison with MC92 and observations

The simulated wind and precipitation structures over the barrier at hour 6 were qualitatively compared with the observations in R92 and the MC92 results using RAMS2 at 1.5-km grid spacing. Because of flow blocking, the u component (cross barrier) is less than 2 m s−1 near the surface at the base of the Sierra barrier (Fig. 7a), which is similar to the MC92 simulation (their Fig. 4a) and the observations (Fig. 4a of R92). The υ component (terrain parallel) has an elongated barrier jet near the base of the barrier (Fig. 7b), with its location close to MC92 (their Fig. 4b) and the observations (Fig. 4b of R92). The magnitude of the barrier jet is 15 m s−1, which is less than observed (20 m s−1), but greater than MC92 (9 m s−1). During the 6–12-h forecast period, the MM5 barrier jet increased slightly to 17 m s−1.

The simulated rain mixing ratios at hour 6 in the MM5 extend from the surface up to 2 km above sea level (ASL) over the lower windward slope (Fig. 5b), with a peak of 0.34 g kg−1. MC92's rain (their Fig. 5b) is similar, with a maximum on the windward side of 0.45 g kg−1. The simulated maximum in snow mixing ratio intersect near the Sierra crest, which agrees well with MC92 (their Fig. 5d) and the observations (Fig. 1). The MM5 snow field extends from 2 to 7.7 km ASL, with a maximum of 0.40 g kg−1 over the windward slope, while the snow field in MC92 has a maximum value of 1.0 g kg−1. The less snow in MM5 is more consistent with the weaker reflectivities (10–15 dBZ) observed aloft between 4 and 5 km, although the MM5 is still overpredicted by 5–10 dBZ (not shown). The graupel region extending from 1 to 4.7 km ASL has a cellular pattern (Fig. 5b), with a peak of 0.56 g kg−1 near 1.5 km ASL along the windward slope, which is close to the observed fallout location (Fig. 1). MC92's graupel (their Fig. 5e) also has this cellular structure, with a maximum value of 0.50 g kg−1.

The cloud water in Reisner2 at hour 6 extends from the surface to 7 km ASL over the windward slope (Fig. 7c), with values around 0.25 g kg−1 below 3 km ASL and localized maxima to 0.5 g kg−1 around 4–5 km ASL. In MC92 (their Fig. 5a), cloud water exists only between 1 to 2 km ASL, with a maximum of 0.4 g kg−1. The observations in R92 suggest values around 0.4 g kg−1 concentrated mainly below 3 km ASL (Fig. 1); therefore, the simulated cloud water in MC92 is too shallow while the MM5 is too deep. The cloud ice in the MM5 is situated between 8 and 9 km ASL, with a maximum of 0.1 g kg−1 (Fig. 7c), while the cloud ice in MC92 is located between 3 and 7 km ASL, with a maximum of 1.0 g kg−1 (their Fig. 5c). One likely reason for this difference is that the MM5 autoconverts cloud ice to snow at a smaller size threshold, which will be investigated below. The relatively large amounts of cloud ice above 7 km in MM5 and MC92 are also not consistent with the observations (Rauber 1992); therefore, the ice in this region will be removed in MM5 to test the impact on the lower-level cloud (section 4a).

The MM5 precipitation for the 6–12-h period has a maximum about 20 km upstream of the crest (see Fig. 9 later), which is similar to MC92 (their Fig. 6) and the observations (not shown). However, the simulated maximum precipitation rate 20 km upwind of the crest (5.3 mm h−1) is 66% greater than observed (about 3.2 mm h−1), while MC92 (5.9 mm h−1) is 84% greater than observed (not shown).

c. Comparison of MM5 schemes

The six microphysical schemes (warm rain, simple ice, Reisner1, Goddard, Reisner2, and Schultz) in the MM5 were compared at 2-km grid spacing using 5-min averaged output between hours 6 and 12 (runs 1 and 4–8 in Table 1). Figure 8 illustrates the average 6–12-h mixing ratios for the various schemes, while Fig. 9 shows the surface precipitation distributions. The convective plumes aloft are not evident using the time-averaged output (Figs. 8a,b), thereby allowing one to better quantify the net effect of each scheme during the period.

The warm-rain scheme (run 4) has 30%–50% more precipitation than Reisner2 along the lower windward slope of the Sierra Nevada (Fig. 9), and 30%–70% less precipitation than Reisner2 near the crest and lee. Without ice microphysics, all condensate goes to cloud water (Fig. 8d), which rapidly autoconverts to rain and precipitates over the lower windward slope (Fig. 8c). The warm rain also produces 35% more precipitation over the narrow peaks of the coastal range than Reisner2. The warm-rain precipitation is closer to the observed over the upper windward slope than the other BMPs.

The simple ice scheme (run 5) does not include graupel or supercooled water processes (Figs. 8e,f), which results in about 10%–20% less precipitation along lower and middle windward slope than Reisner2 (Fig. 8). Simple ice also produces twice as much snow aloft as Reisner2, which advects downwind, resulting in about 20% more precipitation over the crest. The simple ice scheme produces more rain over the narrow coastal range than Reisner2.

Since Reisner1 (run 6) has many of the same snow processes as Reisner2, both schemes have similar snow and supercooled water distributions aloft as well as surface precipitation (Figs. 8g, 8h, and 9); however, Reisner1 produces slightly more precipitation over the lower windward slope and less over the upper windward side and crest than Reisner2 (less than 10%). These differences may be from Reisner1 having no graupel processes, a fixed slope intercept for rain in Reisner1, and different autoconversion thresholds.

Even though the Goddard scheme (run 7) includes graupel, its precipitation distribution over the windward slope is more similar to simple ice than Reisner2 (Fig. 9), with 10%–15% less precipitation along the lower windward slope. There is very little graupel and supercooled water in the Goddard (Figs. 8i,j), which results in less fallout over the lower windward slope. The Goddard also has 3 times as much cloud ice aloft than Reisner2 and similar cloud-ice amounts shown in MC92.

The precipitation distribution for the Schultz scheme (run 8) has similar characteristics as the Reisner2 and the warm-rain scheme. As in the warm-rain scheme, the Schultz produces more precipitation along the lower windward slope and less in the lee than Reisner2. Meanwhile, the Schultz precipitation profile is similar to Reisner2 over the upper windward slope. Schultz has more cloud ice and less snow over the windward slope than Reisner2 (Figs. 8k,l).

Although the surface precipitation varies between 20% and 30% between MM5 BMPs, all schemes overpredict the observed precipitation upwind of the crest as derived from R92 (Fig. 9). This may be from deficiencies in the BMPs (Colle and Mass 2000), as well as two-dimensional modeling and steady-state assumptions noted in MC92. Both the warm rain and Schultz have too little precipitation spillover into the lee. The uncertainties in the BMPs do warrant further investigation of the specific processes within these schemes.

d. Microphysical budget for Reisner2

In order to better understand the microphysical pathways within the Reisner2 control (CTL) run (run 1), Fig. 10 shows the average microphysical budget over the windward slope between hours 6 and 12, with process values scaled greater than 10% of the WVL rate highlighted in bold. The two most important pathways of water vapor depletion are condensation (cond = 63.24% of WVL) and snow deposition (sdep = 33.48%). For temperatures greater than 0°C, cloud water converts to rain via coalescence processes (racw = 12.40%), while at colder temperatures graupel increases rapidly through cloud water riming onto snow (gsacw = 31.83%) and graupel (ggacw = 4.85%). It is evident that melting of graupel (gmlt = 28.36%), accretion of cloud water by rain (racw = 12.40%), and collection of snow by rain (racs = 10.40%) are important sources of rain. There is little water movement through the cloud-ice category, since any cloud-ice growth that does occur (idep = 2.67%) gets rapidly autoconverted to snow (icns = 2.45%). However, this small amount of cloud-ice generation aloft is important, since without it there would be no ice to snow autoconversion and snow deposition. Overall, most of the surface rain generation comes from melting of graupel rather than coalescence of cloud droplets (racw) or melting of snow (smlt).

By summing the precipitation fallout terms in Fig. 10 (rprc, sprc, gprc), one can calculate the windward precipitation efficiency (PE; Table 2) within the budget box, which is defined as the ratio of the surface precipitation fallout upstream of the crest to the water vapor loss over that same region (Colle 2004; Jiang 2003). The windward PE for the CTL (run 1) is 80%, which is relatively high since this is a relatively wide mountain with large amounts of graupel falling out over the windward slope.

The microphysical budget was also calculated for the horizontal resolution experiments (runs 1–3 in Table 1). With less embedded convection over the Sierra Nevada in the 4-km run (Fig. 5), there is less accretion of cloud water by rain and graupel melting than in the 2-km run (Table 2). The 4-km run has 20% more snow than the higher-resolution runs, which results in more snow advection into the lee. Thus, the 4-km PE is 7%–10% less than higher resolutions. There is little microphysical budget difference from 2- to 1.33-km grid spacing, thus the reason why 2-km grid spacing was used for all experiments.

It is also important to quantify the spatial distribution of the microphysical processes over the windward slope. Figure 11a shows the percentage contribution that each process has on the snow production rate at each point. For the CTL (run 1), snow growth occurs primarily via snow deposition (sdep) over the windward slope (3–8 km). It is maximized near 5.5 km (−15°C), which is a favored region for ice growth since the supersaturation with respect to ice is well below that of water. Conversion of cloud ice to snow (icns) and accretion of cloud ice by snow (saci) dominates in higher levels (7– 9 km). Collection of cloud water by snow (ssacw) occurs between 4 and 6 km ASL, while collection of rainwater by snow (ssacr) is located between 2 and 3 km ASL.

Figure 11b illustrates the spatial distribution of graupel production. Since the supercooled water in this region is limited (Fig. 7c), graupel only grows slowly aloft (5–7 km) in an area of deposition (gdep). Collection of cloud water by snow to graupel (gsacw) dominates lower layers (2–5 km), while the conversion of snow to graupel (scng) also contributes between 2 and 4 km ASL. Collection of rainwater by graupel (gacr) dominates over the crest and immediate lee.

Between 1 and 2 km ASL, most of the rain production is from melting of graupel (gmlt), melting of snow (smlt) (not shown), and collection of snow by rain (racs). Meanwhile, accretion of cloud water by rainwater (racw) primarily occurs in the layer from the surface to 1 km ASL. The racw and cloud water autoconversion (ccnr) terms dominate at higher layers (2–6 km), but there is very little rain production aloft (cf. Fig. 5).

4. Process sensitivity experiments for CA86

a. Ice initiation

There are three options in the MM5 for ice initiation (Fletcher 1962; Cooper 1986; Meyers 1992), which are given as
i1520-0493-132-12-2780-e8
where Si is the supersaturation (percent) with respect to ice, Ni is the number concentration (L−1), T0 = 273.12 K, and T is the temperature in Kelvin. When the temperature is warmer than 252 K, the number of ice particles initiated from greatest to least is Meyers, Cooper, and Fletcher (not shown). In contrast, when the temperature is less than 243 K, the Fletcher parameterization has a far greater number concentration.

The variation of number of cloud ice (Ni) with temperature is consistent with the ice mixing ratios, in which the Fletcher scheme (run 9) has twice (0.1 g kg−1) as much cloud ice as the Cooper scheme (CTL) above 7 km (Fig. 12a), while Meyers (run 10) has about half (0.05 g kg−1) as much as the Cooper (not shown). Interestingly, changing ice initiation schemes has little effect on the other microphysical processes (Table 2) or the snow field (Fig. 12a), and there is less than 1 mm (2%) difference in surface precipitation (not shown).

Since differences in ice initiation do not impact the surface precipitation, how important is ice initiation in precipitation generation? When ice initiation (idsn) is turned off in addition to heterogeneous (ifzc) and homogeneous (ihfzc) freezing of cloud droplets, there is no cloud ice or snow over the barrier and a 20% reduction in surface precipitation near the crest (not shown), so some ice initiation is necessary for snow production. An experiment was also completed by turning off ice production above 7 km or <250 K (run 11), since the ice may have been overpredicted in this region. Without upper-level cloud ice (Fig. 12b), the difference in snow as compared to the CTL becomes much smaller (<10% of CTL) as snow grows and falls to lower levels, resulting in only a slight (<5%) decrease in surface precipitation over the windward slope (not shown). Condensation and accretion of cloud water by rain increase slightly, while there is slightly less riming of snow (Table 2), but overall the differences are smaller than the other microphysical experiments to be shown later.

The ice initiation results suggest that Reisner2 is capable of generating a similar snow field with only a small amount of cloud ice. For example, even with only heterogeneous (ifzc) and homogeneous (ihfzc) freezing allowed to produce cloud ice (no ice initiation), which are combined less than 0.005% of WVL, the surface precipitation and snow aloft differences with the CTL are still less than 10% (not shown). Thus, a very small amount of ice generation aloft can grow rapidly via snow deposition and riming and result in similar snow, graupel, and surface precipitation amounts. This is consistent with the microphysical budget for the CTL (Fig. 11), which showed that the small amount of snow that is autoconverted from cloud ice (icns = 2.45% of WVL) can grow rapidly via deposition (sdep = 33.48% of WVL).

b. Cloud condensation nuclei

The CCN parameter (CNP) was changed from 100 cm−3 in the CTL (typical for a maritime air mass; Twomey and Wojciechowski 1969) to either 600 cm−3 (more typical for a continental airmass) or 10 cm−3 (runs 12 and 13 in Table 1). MC92 found that the RAMS model was only slightly sensitive to CCN, since warm-rain processes accounted for only a small part of the precipitation; however, Reisner2 has more cloud water than MC92 (Figs. 5 and 7c), and therefore may be more sensitive.

Figure 13 shows cross-section differences from the CTL for those experiments using different CCN concentrations; Table 2 lists the flowchart differences from the CTL; and Fig. 14 shows the surface precipitation distribution. Increasing the CCN to 600 cm−3 results in 20%–30% more cloud water (not shown), which results in greater cloud water overpredictions. The greater cloud water leads to more accretion by rain (racw), less riming of graupel given the smaller cloud drop diameters and decreased collision efficiency, and a decrease of graupel by 0.20 g kg−1 along windward slope (Fig. 13a). Less snow is converted to graupel, so snow increases by around 0.20 g kg−1. Rain is decreased along the lower windward slope by about 0.05 g kg−1 due to decreased melting of graupel, but it is increased by around 0.05 g kg−1 over the upper windward slope and crest. As a result, there is about 3 mm (10%) less precipitation along the lower windward slope and 2 mm (15%) more over the crest (Fig. 14). With the shift of precipitation toward the lee, the windward PE decreases by 6% relative to the CTL (Table 2).

Reducing the CCN to 10 cm−3 increases graupel while snow is decreased (Fig. 13b), and rain is increased over the upper windward slope. There is about 2 mm (10%) more precipitation at the base of the Sierra Nevada and 1.5 mm (5%) less over the crest (Fig. 14). The windward PE is 4% larger than the CTL experiment.

c. Slope intercept for snow number concentration

The slope intercept in the Marshall–Palmer distribution of snow, Nos, is not only important to the snow fall speed, but it also influences riming, deposition, and melting of snow. To examine the effect of Nos on microphysical processes and surface precipitation, the snow intercept parameter Nos in the CTL (Nosqs) was changed to either a fixed value (run 14) or a temperature-dependent Nos (run 15), described in Thompson et al. (2004) as
T6T0T
where T0 = 273.15 K.

From Eq. (11), as the temperature decreases, NosT increases, which favors more small snow particles and fewer large particles. When the temperature is colder than −19°C, NosT is larger than a fixed Nos. For a representative snow mixing ratio of 4.5 × 10−4 kg kg−1, temperature of −12°C, and air density of 0.83 kg m−3, the number concentration using NosT is between the smaller Nosqs and the larger fixed Nos (not shown). In general, Nosqs has more large snow particles and fewer small particles than the other Nos approaches.

Using a fixed Nos (Fig. 15a), graupel and cloud water are decreased compared to the CTL, while snow is increased along the windward slope (maximum change of +0.88 g kg−1 for snow and −0.32 g kg−1 for graupel). There is 2 mm (7%) less surface precipitation over the lower windward slope due to a decrease of rain (Fig. 16a). Over the middle windward slope, rain increases greatly but graupel decreases, so there is only 1 mm (4%) more total precipitation. Over the upper windward slope and crest, snow increases while graupel decreases, and there is about 4 mm (15%–25%) more precipitation. Most of the increase in snow aloft still falls out before the lee, so there is little windward PE difference with the CTL (Table 2).

A fixed Nos decreases the condensation contribution to WVL from 63% to 35% (Fig. 17a), while snow deposition increases from 33% to 62% of WVL. Less available cloud water results in 53% less riming of cloud water by snow and 82% less collection of cloud water by graupel as compared to the CTL run, which in turn results in less graupel. Overall, a fixed Nos favors a microphysical pathway involving more snow deposition (wv → sdep → snow) rather than graupel production via riming (wv → cond → ggacw, gsacw → graupel). These changes are consistent with the greater number of smaller snow particles using fixed Nos, which can more efficiently deplete the available supersaturated water vapor. As a result, the fixed Nos did help with the overpredictions of cloud water aloft noted in the CTL as compared to R92.

The temperature-dependent Nos scheme has a similar effect as a fixed Nos when compared with Nosqs, but the changes are about half as large (Fig. 15b). The increase of rain leads to 2.5 mm (10%) more precipitation over the lower windward slope, while the increase of snow and decrease of graupel leads to 3.5 mm (26%) more over the crest of the barrier (Fig. 16b). The budget shows an increase of snow deposition from 33% to 43% of WVL using NosT (Fig. 17b), while there is a corresponding decrease in condensation and riming processes.

d. Snow fall speed (Vs)

Since the terminal velocity of snow is relatively small (<1.5 m s−1), a small variation in its fall speed can substantially change where it ultimately reaches the surface. The CTL uses the Rutledge and Hobbs (1983) expression for fall speed (Fig. 18), which is based on the Locatelli and Hobbs (1974) fall speed for unrimed radiating assemblages of plates, side planes, bullets, and columns [Eq. (12)]. Colle and Mass (2000) compared this expression with Cox [1988; Eq. (13)], which is 20% smaller than Rutledge and Hobbs and close to the Ferrier [1994; Eq. (14)] expression for unrimed radiating assemblages of dendrites. Lin et al. (1983), Brown and Swann (1997), and MC92 used the snow fall speed for graupel-like snow of hexagonal type [Eq. (15)], which is similar in magnitude to the Rutledge and Hobbs (1983) expression (Fig. 18). The various fall speed expressions can be written as
i1520-0493-132-12-2780-e12
where D is the particle diameter in meters and V is the fall speed (m s−1).

Since the Ferrier (1994) changes (run 16) are similar to Cox (1988; run 17), only the cross-section differences from Ferrier are shown (Fig. 19). With the smaller Ferrier fall speed, snow particles are suspended for a longer time, and 20%–30% more snow mass is created as compared to the CTL. This snow increase results from more collection of rain by snow (ssacr), collection of cloud water by snow (ssacw) and deposition of snow (sdep), less conversion of snow to graupel (scng), and melting of snow (smlt) (Table 2). Accretion of cloud water by rain (racw) is increased, so more water vapor goes through wv → sdep → snow and wv → cond → racw → rain instead of cond → ggacw, gsacw → graupel. As a result, the graupel mass is decreased, which results in less melting and rain. Therefore, precipitation is decreased by 10%–20% over the lower part of the barrier from the decrease of rain, while it is increased by 10%– 60% over the upper windward slope, crest, and lee side from the snow increase (Fig. 16b). Overall, a reduction in snow fall speed (Ferrier 1994; Cox 1988) shifts more precipitation from the windward side to the lee, and the windward PE decreases from 80% in the CTL to around 70%. The Brown and Swann (1997) fall speed (run 18) is similar to the CTL (Fig. 19); therefore, the microphysical and PE results are similar (Table 2).

e. Graupel density

MC92 found that the surface precipitation in the RAMS was quite sensitive to graupel density. To determine how this density change impacts Reisner2, the graupel density was changed from 400 to 900 kg m−3 (run 19) or 200 kg m−3 (run 20), which ranges from hail-like graupel to rimed snow, respectively.

For a larger graupel density, rain is decreased at the base of windward slope (Fig. 20a), while over the upper slope graupel is decreased by 0.02 g kg−1 and snow is increased by 0.10 g kg−1 (not shown). There is 1.5 mm (6%) less rain precipitation over the lower windward slope (Fig. 20a) and 1.2 mm (8%) more precipitation over the crest. Graupel decreases since a larger graupel density favors less autoconversion from snow to graupel (scng in Table 2). Meanwhile, the decrease of conversion of snow to graupel (scng) and increase of collection of snow by rain (racs) leads to a snow increase.

For smaller graupel density, there is 2.3 mm (10%) more precipitation along the lower slope, while a decrease of snow near the crest results in 2.7 mm (20%) less surface precipitation. A decrease in the collection of snow by rain (racs) and increase in the conversion of snow to graupel leads to the decrease of snow precipitation near the crest.

f. Graupel fall speed

The graupel fall speed (Vg = agDbg, where D is the particle diameter (m) and Vg is the fall speed (m s−1) was changed from Ferrier et al. (1995) (ag = 19.3, bg = 0.37) in the CTL (run 1) to a larger Vg in Lin et al. (1983) (ag = 94.5, bg = 0.5) (run 21). A larger graupel fall speed results in less residence time for graupel aloft; therefore, graupel is decreased by 0.20 g kg−1 while rain is increased by 0.05 g kg−1 (not shown). This results in 6 mm (28%) more precipitation at the base of the Sierra Nevada (Fig. 20b), and the decrease of graupel aloft results in 4 mm (20%) less precipitation near the crest. The amount of graupel decreases since most source terms decrease, such as collection of cloud water by snow to graupel (gsacw), collection of rain by graupel (gacr), and conversion of snow to graupel (scng), while the primary sink of graupel (gmlt) is increased (Table 2). Changing the graupel fall speed has little effect on the windward PE.

g. Cloud ice to snow autoconversion

The amount of cloud ice to snow autoconversion (Picns) is dependent on the radius of the smallest snow particle (Rso) allowed. The Rso parameter in Reisner2 is set to a relatively small value of 75 μm, as compared to 100 μm in the RAMS model (Flatau et al. 1989). When the Rso threshold is increased by a factor of 4 (run 22), the cloud ice deposition (idep) is 16% of WVL as compared to 3% in the CTL (run 1), which leads to the increase of cloud ice by 0.20 g kg−1 (not shown). Meanwhile, condensation decreases, such that riming and accretion processes decrease (Table 2), thus resulting in 0.20 g kg−1 less graupel. Even though deposition of snow (sdep) decreases, an ice to snow autoconversion increase of 11% (icns) results in 0.20 g kg−1 more snow. This leads to slightly (5%) less precipitation over the windward slope and more near the crest than the CTL, and the windward PE decreases slightly from 80% to 77%.

With a reduced Rso (run 23; rso × 0.25), there is little precipitation difference with the CTL and little change in the microphysical budget (Fig. 21a; Table 2), since the conversion of cloud ice to snow is already relatively small in the CTL (Fig. 10).

h. Cloud water autoconversion

The collision and coalescence of cloud droplets to rainwater is parameterized by the formulation of Kessler (1969) as
aHqcqcth
where a = 1 × 10−3 s−1, H is a Heaviside function, and qcth = 0.35 × 10−3 is a threshold value for the CTL run.

When doubling the autoconversion threshold from cloud water to rainwater (run 24), the amount of autoconversion (ccnr) is decreased slightly (1% of WVL). Both rain and snow are decreased slightly (<0.03 g kg−1) over the windward slope (not shown), while graupel and cloud water are increased by an equivalent amount. The reduction in rain results in less accretion of cloud water by rain (racw) (Table 2), while the increase in graupel results in more riming of snow (gsacw). With a reduced ccnr threshold (qcth × 0.5) (run 25), there is 0.08 g kg−1 less cloud water, while 0.05 g kg−1 of supercooled rain exists up to 5 km ASL (not shown). A reduction in cloud water results in less riming (Table 2); therefore, graupel decreases by 0.08 g kg−1, while snow is increased.

There is little impact on the surface precipitation over the Sierra Nevada when doubling or halving the autoconversion threshold (Fig. 21b). The changes in graupel and rainwater nearly offset each other, and since both species fall relatively fast, there is little change in the surface precipitation. In contrast, there is more sensitivity to ccnr over the coastal range, since the localized areas of strong upward motion and the relatively high freezing level results in more low-level cloud water and little graupel aloft (Figs. 8a,b).

5. Summary and discussion

This study has investigated the sensitivities of surface precipitation, bulk microphysical processes, and water/ ice mixing ratios to selected parameters within the Reisner2 bulk microphysical parameterization (BMP) of MM5. The MM5 was run two-dimensionally using a horizontal grid spacing of 2 km, which was needed in order to realistically simulate the embedded convective plumes over the Sierra Nevada during the 12–13 February 1986 (CA86) event. Compared to a similar 2D study of this event by MC92 using the RAMS model, the MM5 Reisner2 scheme produces more cloud water and less cloud ice and snow.

Unlike previous microphysical sensitivity studies (MC92; Thompson et al. 2004), a sophisticated BMP was fully diagnosed by completing a microphysical budget. Each microphysical process from Reisner2 was output for a box over the windward slope of the Sierra Nevada and normalized by the water vapor loss rate. This unique tool allows one to quantify how the water mass moves between each hydrometeor species, thereby allowing one to diagnose how a change made to a BMP can alter the other processes. Several different parameters were tested within the Reisner2 BMP. For the relatively wide Sierra barrier, the Reisner2 was most sensitive to those parameters associated with snow and graupel distributions, cloud condensation nuclei (CCN) concentrations, and fall speeds, and less sensitive to ice initiation and autoconversions.

The amount of cloud ice aloft is much larger when the Fletcher ice initiation is used rather than the Meyers and Cooper approaches; however, these changes result in only small precipitation differences (<2%) and microphysical sensitivities. The sensitivity is small since cloud ice gets autoconverted to snow at small (75 μm) sizes. Therefore, even a small amount of cloud ice autoconverted to snow can grow rapidly via deposition and riming. Increasing the ice autoconversion threshold by a factor of 4 results in twice as much cloud ice aloft, but the windward surface precipitation still only changes by around 5%.

Changing CCN from 100 cm−3 (typical for a maritime air mass) to 600 cm−3 (typical for a continental air mass) results in more cloud water and snow but less graupel. Less (10%) precipitation is created at the base of the barrier, but more (15%) over the crest. Decreasing CCN to 10 cm−3 leads to more graupel and less snow, and there is more (10%) precipitation at the base of the barrier but less (5%) over the crest.

Surface precipitation varied from 10%–30% when using different snow and graupel fall speeds. Reducing the snow fall speed (Ferrier Vs and Cox Vs) allows for a longer time for snow growth and advection over the barrier. Similarly, an increased graupel fall speed decreases the residence time aloft; therefore, the amount of graupel decreases while rain is increased. The surface precipitation is shifted upstream of the mountain with a larger graupel fall speed.

For a larger graupel density, graupel is decreased while snow is increased since increasing graupel density effectively decreases its time for growth aloft. In contrast, for a smaller graupel density, graupel is increased and snow is decreased. The surface precipitation changes by 10%–20%, with more advecting from the windward to the lee side with a smaller graupel density.

There was little precipitation sensitivity (<3%) over the Sierra Nevada to the magnitude of the cloud water to rainwater autoconversion (ccnr). With an increased ccnr threshold, rain is decreased slightly while cloud water is increased. As a result, more cloud water is converted to graupel through collection processes, but less is converted to rain through accretion of cloud water, and then to snow via collection of rain by snow.

The microphysical budgets and surface precipitation change significantly with the method used to define the snow intercept parameter (Nos). A fixed Nos scheme favors more small snow particles and fewer large ones at warmer temperatures, so deposition of snow aloft doubles in magnitude. In contrast, using a variable Nos as a function of mixing ratio (Nosqs) has fewer snow particles, which favors less depositional growth aloft, more supercooled water, and riming. A temperature-dependent NosT scheme is between fixed Nos and variable Nosqs results, and NosT is currently used in the Reisner2 BMP (Thompson et al. 2004).

Several other BMPs in the MM5 were tested for the CA86 event. The differences in precipitation between the various BMP schemes is generally greater than changing any single Reisner2 process in this study. The larger differences in the BMP experiments is likely from a combination of different microphysical processes and/ or hydrometeor species allowed. Overall, it appears that to improve these BMPs will require more than one microphysical process change (i.e., there is no silver bullet).

The sensitivity simulations in this study help explain some of the differences between those MM5 BMPs that use a fixed slope intercept for snow (i.e, simple ice and Goddard) and a Nosqs (Reisner1 and Reisner2). For example, those schemes that use a fixed slope intercept (simple ice, Goddard, and Schultz) have more snow aloft and less cloud water than those using a Nosqs. A microphysical budget was completed for the simple ice scheme for the CA86 event to see how it compares with the more sophisticated Reisner2 (Fig. 22). As with the fixed Nos experiment for Reisner2 (Fig. 17a), 50% of the WVL in simple ice goes to snow deposition. However, unlike Reisner2, simple ice has 21% of WVL going to cloud-ice deposition. Simple ice does not include supercooled water, so only ice deposition occurs aloft. The cloud ice is rapidly autoconverted to snow (icns = 37.65%); thus there is less suspended cloud ice aloft as compared to snow (Figs. 8e,f).

Overall, this study has improved our understanding of some of these complex BMPs. More three-dimensional simulations using recent field data such as IMPROVE are needed. It appears that some of the microphysical sensitivities for this CA86 event are different between the coastal range and Sierra Nevada. This suggests that other factors, such as barrier width, control the time scale for microphysical growth and may impact some of the results presented. Therefore, in Part II (Colle and Zeng 2004) of this study, a set of idealized studies are completed to test the importance of barrier width as well as freezing level for some of the important BMPs processes in Reisner2.

Acknowledgments

This research was supported by the National Science Foundation (ATM-0094524). This work benefited from several useful discussions with Greg Thompson on recent changes made to the Reisner2 MM5 code. Thanks to the two anonymous reviewers who helped improve this manuscript. Use of the MM5 was made possible by the Microscale and Mesoscale Meteorological (MMM) Division of the National Center for Atmospheric Research (NCAR), which is supported by the National Science Foundation. The 2D version of the MM5 was developed by Prof. Dave Dempsey and the MMM Division of NCAR.

REFERENCES

  • Berry, E. X., and R. L. Reinhardt, 1974: An analysis of cloud drop growth by collection. Part IV: A new parameterization. J. Atmos. Sci, 31 , 21272135.

    • Search Google Scholar
    • Export Citation
  • Bosart, L. F., 2003: Whither the weather analysis and forecasting process? Wea. Forecasting, 18 , 520529.

  • Braun, S. A., and W-K. Tao, 2000: Sensitivity of high-resolution simulations of Hurricane Bob (1991) to planetary boundary layer parameterizations. Mon. Wea. Rev, 128 , 39413961.

    • Search Google Scholar
    • Export Citation
  • Brown, P. R. A., and H. A. Swann, 1997: Evaluation of key microphysical parameters in three-dimensional cloud-model simulation using aircraft and multiparameter radar data. Quart. J. Roy. Meteor. Soc, 123 , 22452275.

    • Search Google Scholar
    • Export Citation
  • Bruintjes, R. T., T. L. Clark, and W. D. Hall, 1994: Interactions between topographic airflow and cloud/precipitation development during the passage of a winter storm in Arizona. J. Atmos. Sci, 51 , 4867.

    • Search Google Scholar
    • Export Citation
  • Burrows, D. A., 1992: Evaluation of a two-dimensional kinematic cloud model using data from a central Sierra Nevada orographic cloud system. J. Appl. Meteor, 31 , 5163.

    • Search Google Scholar
    • Export Citation
  • Charba, J. P., D. W. Reynolds, B. E. McDonald, and G. M. Carter, 2003: Comparative verification of recent quantitative precipitation forecasts in the National Weather Service: A simple approach for scoring forecast accuracy. Wea. Forecasting, 18 , 161183.

    • 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
  • Colle, B. A., and C. F. Mass, 2000: The 5–9 February 1996 flooding event over the Pacific Northwest: Sensitivity studies and evaluation of the MM5 precipitation forecasts. Mon. Wea. Rev, 128 , 593617.

    • Search Google Scholar
    • Export Citation
  • Colle, B. A., and Y. Zeng, 2004: Bulk microphysical sensitivities within the MM5 for orographic precipitation. Part II: Impact of barrier width and freezing level. Mon. Wea. Rev.,132, 2802–2815.

    • Search Google Scholar
    • Export Citation
  • Colle, B. A., K. J. Westrick, and C. F. Mass, 1999: Evaluation of the MM5 and Eta-10 precipitation forecasts over the Pacific Northwest during the cool season. Wea. Forecasting, 14 , 137154.

    • Search Google Scholar
    • Export Citation
  • Colle, B. A., C. F. Mass, and K. J. Westrick, 2000: MM5 precipitation verification over the Pacific Northwest during the 1997–1999 cool seasons. Wea. Forecasting, 15 , 730744.

    • Search Google Scholar
    • Export Citation
  • Cooper, W. A., 1986: Ice initiation in natural clouds. Precipitation Enhancement—A Scientific Challenge, Meteor. Monogr., No. 21, Amer. Meteor. Soc., 29–32.

    • Search Google Scholar
    • Export Citation
  • Cox, G. P., 1988: Modeling precipitation in frontal rainbands. Quart. J. Roy. Meteor. Soc, 114 , 115127.

  • Doyle, J. D., and Coauthors, 2000: An intercomparison of model-predicted wave breaking for the 11 January 1972 Boulder windstorm. Mon. Wea. Rev, 128 , 901914.

    • Search Google Scholar
    • Export Citation
  • Dudhia, J., 1989: Numerical study of convection observed during the winter monsoon experiment using a mesoscale two-dimensional model. J. Atmos. Sci, 46 , 30773107.

    • Search Google Scholar
    • Export Citation
  • Ferrier, B. S., 1994: A double-moment, multiple-phase, four-class bulk ice scheme. Part I: Description. J. Atmos. Sci, 51 , 249280.

  • Ferrier, B. S., W-K. Tao, and J. Simpson, 1995: A double-moment, multiple-phase, four-class bulk ice scheme. Part II: Simulations of convective storms in different large-scale environments and comparisons with other bulk parameterizations. J. Atmos. Sci, 52 , 10011033.

    • Search Google Scholar
    • Export Citation
  • Flatau, P. J., G. J. Tripoli, J. Verlinde, and W. R. Cotton, 1989: The CSU–RAMS cloud microphysical module: General theory and code documentation. Atmospheric Science Paper 451, 88 pp. [Available from Colorado State University, Dept. of Atmospheric Science, Fort Collins, CO 80523.].

    • Search Google Scholar
    • Export Citation
  • Fletcher, N. H., 1962: Physics of Rain Clouds. Cambridge University Press, 386 pp.

  • Fritsch, J. M., and Coauthors, 1998: Quantitative precipitation forecasting: Report of the eighth prospectus development team, U.S. Weather Research Program. Bull. Amer. Meteor. Soc, 79 , 285299.

    • Search Google Scholar
    • Export Citation
  • Gaudet, B., and W. R. Cotton, 1998: Statistical characteristics of real-time precipitation forecasting model. Wea. Forecasting, 13 , 966982.

    • Search Google Scholar
    • Export Citation
  • Grell, G. A., J. Dudhia, and D. R. Stauffer, 1994: A description of the fifth-generation Penn State/NCAR Mesoscale Model (MM5). NCAR Tech. Note NCAR/TN-398+STR, 138 pp. [Available from National Center for Atmospheric Research, P.O. Box 3000, Boulder, CO 80307.].

    • Search Google Scholar
    • Export Citation
  • Hallet, M. F., and S. C. Mossop, 1974: Production of secondary ice particles during the riming process. Nature, 249 , 2628.

  • Hobbs, P. V., 1975: The nature of winter clouds and precipitation in the Cascade Mountains and their modification by artificial seeding. Part I: Natural conditions. J. Appl. Meteor, 14 , 783804.

    • Search Google Scholar
    • Export Citation
  • Hobbs, P. V., and P. O. G. Persson, 1982: The mesoscale and microscale structure and organization of clouds and precipitation in midlatitude cyclones. Part V: The substructure of narrow cold-frontal rainbands. J. Atmos. Sci, 39 , 280295.

    • Search Google Scholar
    • Export Citation
  • Hobbs, P. V., T. J. Matejka, P. H. Herzegh, J. D. Locatelli, and R. A. Houze Jr., 1980: The mesoscale and microscale structure and organization of clouds and precipitation in midlatitude cyclones. I: A case study of a cold front. J. Atmos. Sci, 37 , 568596.

    • Search Google Scholar
    • Export Citation
  • Hong, S-Y., and H-L. Pan, 1996: Nonlocal boundary layer vertical diffusion in a medium-range forecast model. Mon. Wea. Rev, 124 , 23222339.

    • Search Google Scholar
    • Export Citation
  • Houze, R. A., P. V. Hobbs, and P. H. Herzegh, 1979: Size distributions of precipitation particles in frontal clouds. J. Atmos. Sci, 36 , 156162.

    • Search Google Scholar
    • Export Citation
  • Hsie, E. Y., R. A. Anthes, and D. Keyser, 1984: Numerical simulation of frontogenesis in a moist atmosphere. J. Atmos. Sci, 41 , 25812594.

    • Search Google Scholar
    • Export Citation
  • Jiang, Q., 2003: Moist dynamics and orographic precipitation. Tellus, 55A , 301316.

  • Jiang, Q., and R. B. Smith, 2003: Cloud timescales and orographic precipitation. J. Atmos. Sci, 60 , 11591172.

  • Kessler, E., 1969: On the Distribution and Continuity of Water Substance in Atmospheric Circulations. Meteor. Monogr., No. 32, Amer. Meteor. Soc., 84 pp.

    • Search Google Scholar
    • Export Citation
  • Klemp, J. B., and D. R. Durran, 1983: An upper boundary condition permitting internal gravity wave radiation in numerical mesoscale models. Mon. Wea. Rev, 111 , 430444.

    • Search Google Scholar
    • Export Citation
  • Lin, Y. L., R. Farley, and H. D. Orville, 1983: Bulk parameterization of the snow field in a cloud model. J. Climate Appl. Meteor, 22 , 10651092.

    • Search Google Scholar
    • Export Citation
  • Locatelli, J. D., and P. V. Hobbs, 1974: Fall speeds and masses of solid precipitation particles. J. Geophys. Res, 79 , 21852197.

  • Meyers, M. P., and W. R. Cotton, 1992: Evaluation of the potential for wintertime quantitative precipitation forecasting over mountainous terrain with an explicit cloud model. Part I: Two-dimensional sensitivity experiments. J. Appl. Meteor, 31 , 2650.

    • Search Google Scholar
    • Export Citation
  • Meyers, M. P., P. J. Demott, and W. R. Cotton, 1992: New primary ice-nucleation parameterization in an explicit cloud model. J. Appl. Meteor, 31 , 708721.

    • Search Google Scholar
    • Export Citation
  • Rasmussen, R. M., and Coauthors, 1992: Winter Icing and Storm Projects (WISP). Bull. Amer. Meteor. Soc, 73 , 951974.

  • Rasmussen, R. M., I. Geresdi, G. Thompson, K. Manning, and E. Karplus, 2002: Freezing drizzle formation in stably stratified layer clouds: The role of radiative cooling of cloud droplets, cloud condensation nuclei, and ice initiation. J. Appl. Meteor, 59 , 837860.

    • Search Google Scholar
    • Export Citation
  • Rauber, R. M., 1992: Microphysical structure and evolution of a Sierra Nevada shallow orographic cloud system. J. Appl. Meteor, 31 , 324.

    • Search Google Scholar
    • Export Citation
  • Reisner, J., R. M. Rasmussen, and R. T. Bruintjes, 1998: Explicit forecasting of supercooled liquid water in winter storm using the MM5 mesoscale model. Quart. J. Roy. Meteor. Soc, 124 , 10711107.

    • Search Google Scholar
    • Export Citation
  • Reynolds, D. W., and A. S. Dennis, 1986: A review of the Sierra Cooperative Pilot Project. Bull. Amer. Meteor. Soc, 67 , 513523.

  • Rutledge, S. A., and P. V. Hobbs, 1983: The mesoscale and microscale structure and organization of clouds and precipitation in mid-latitude cyclones: A model for the “seeder-feeder” process in warm-frontal rainbands. J. Atmos. Sci, 40 , 11851206.

    • Search Google Scholar
    • Export Citation
  • Rutledge, S. A., and P. V. Hobbs, 1984: The mesoscale and microscale structure and organization of clouds and precipitation in mid-latitude cyclones: A diagnostic modeling study of precipitation development in narrow cold-frontal rainbands. J. Atmos. Sci, 41 , 29492972.

    • Search Google Scholar
    • Export Citation
  • Schultz, P., 1995: An explicit cloud physics parameterization for operational numerical weather prediction. Mon. Wea. Rev, 123 , 33313343.

    • Search Google Scholar
    • Export Citation
  • Sekhon, R. S., and R. C. Srivastava, 1970: Snow size spectra and radar reflectivity. J. Atmos. Sci, 27 , 299307.

  • Stoelinga, M., and Coauthors, 2003: Improvement of microphysical parameterizations through observational verification experiments (IMPROVE). Bull. Amer. Meteor. Soc, 84 , 18071826.

    • Search Google Scholar
    • Export Citation
  • Tao, W-K., and J. Simpson, 1993: The Goddard Cumulus Ensemble Model. Part I: Model description. Terr. Atmos. Oceanic Sci, 4 , 3572.

  • Thompson, G., R. M. Rasmussen, and K. Manning, 2004: Explicit forecasts of winter precipitation using an improved bulk microphysics scheme. Part I: Description and sensitivity analysis. Mon. Wea. Rev, 132 , 519542.

    • Search Google Scholar
    • Export Citation
  • Twomey, S., and T. A. Wojciechowski, 1969: Observations of the geographic variation of cloud nuclei. J. Atmos. Sci, 26 , 684688.

Fig. 1.
Fig. 1.

Conceptual model of the microphysical processes over the Sierra Nevada on 12 Feb 1986 (from Fig. 18 of Rauber 1992). Arrows indicate cloud droplet (C), needle (N), and dendritic particle (D) trajectories in the plane of the cross section. The few key isotherms (dashed) are also shown

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 2.
Fig. 2.

The terrain profile across the coastal range and Sierra Nevada used in this study. The two bold arrows outline the position of the box used for the microphysical budget, and SH is the location of the Sheridan, CA, sounding used in this study

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 3.
Fig. 3.

The 1500 UTC 12 Feb 1986 Sheridan sounding used in this study (see SH on Fig. 2) showing temperatures (black) and dewpoint (gray). The geostrophic winds (one full barb = 10 m s−1) were obtained from the 1200 UTC 12 Feb NCEP synoptic charts

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 4.
Fig. 4.

Microphysical flowchart for the Reisner2 scheme. The circles represent the various water species (water vapor, cloud water, cloud ice, rain, snow, and graupel), and the arrows are the processes that link the species (see appendix for the list of processes)

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 5.
Fig. 5.

Cross section at hour 6 showing (a) rain (thin solid every 0.04 g kg−1), snow (gray), and graupel (black solid) every 0.08 g kg−1, and wind vectors at 4-km grid spacing (run 3 in Table 1). (b) Same as (a), except for 2-km grid spacing (run 1). (c) Same as (a), except for 1.3-km grid spacing (run 2)

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 6.
Fig. 6.

Accumulated 6–12-h surface precipitation (mm) for the 4‐, 2-, and 1.33-km horizontal grid spacings. The terrain is also shown for reference

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 7.
Fig. 7.

Cross section for the control run (run 1) at hour 6 showing (a) u (normal) wind component (every 2.0 m s−1) and (b) υ (parallel) wind component (every 2.0 m s−1), and (c) cloud water (solid black every 0.08 g kg−1), cloud ice (solid gray every 0.02 g kg−1), and circulation vectors

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 8.
Fig. 8.

Cross section showing the distribution of (a) rain (thin solid), snow (gray), and graupel (thick black) for the Reisner2 scheme. (b) Same as (a), except for cloud water (black) and cloud ice (gray). The contour interval is 0.04 g kg−1 for rain, 0.08 g kg−1 for snow, 0.08 g kg−1 for graupel, 0.08 g kg−1 for cloud water, and 0.02 g kg−1 for cloud ice. (c), (d) Same as (a), (b), except for the warm-rain scheme; (e), (f) same as (a), (b), except for the simple ice scheme; (g), (h) same as (a), (b), except for the Reisner1 scheme; (i), (j) same as (g), (h), except for the Goddard scheme; (k), (l) same as (g), (h), except for the Schultz scheme

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 8.
Fig. 8.

(Continued )

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 9.
Fig. 9.

Surface-accumulated (6–12 h) precipitation (mm) vs distance for the different microphysical schemes: Reisner2 (bold solid), warm rain (dotted), simple ice (dashed), Reisner1 (solid), Goddard (dash–dot–dot–dot), and Schultz (long dash). The observed precipitation profile over a portion of the Sierra Nevada (derived from Rauber 1992) is shown by the thick gray line

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 10.
Fig. 10.

Flowchart of the microphysical processes between 6–12 h of the control run (run 1) for the box in Fig. 2. The values shown are the ratio of each microphysical process rate to the total WVL rate (cond + sdep + gdep + idsn + idep) within the box. The processes are listed in the appendix. The sum of all the microphysical process tendencies for each species is given by (wv:, cw:, r:, ci:, g: and s:). This sum does not include horizontal advection and diffusion/ divergence, which are labeled as hadv and other, respectively. The fallout tendency of rain (rprc), snow (sprc), graupel (gprc), and cloud ice (iprc) are also shown. Microphysical processes greater than 10% of the WVL rate are in bold

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 11.
Fig. 11.

Cross section of the spatial distribution of the major microphysical processes contributing to snow production, showing the ratio of (a) sdep (gray), icns (black), and ssacr (dashed gray), (b) saci (gray), ssacw (dashed gray), and siacr (black) to the total snow production rate from these processes at each point contoured every 20%. (c), (d) Same as (a), except for graupel production showing (c) gdep (gray), gsacw (dashed gray), and gacw (black), and (d) scng (black) and gacr (gray). (e), (f) Same as (a), except rain production showing (e) racw (gray) and racs (black), and (f) gmlt (black), ccnr (dashed gray), and racs (gray)

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 12.
Fig. 12.

(a) Cross section difference between run 9 (Fletcher ice initiation) and run 1 (Cooper ice initiation) (run 9 − run 1) showing cloud ice (black) every 0.01 g kg−1 and snow (gray) every 0.01 g kg−1, with negative values dashed. (b) Same as (a), except the difference between run 11 (no ice production above 7 km) and run 1 (run 11 − run 1) showing cloud ice (black) every 0.02 g kg−1 and snow (gray) every 0.02 g kg−1

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 13.
Fig. 13.

(a) Cross-section difference between run 12 (larger CNP) and run 1 (run 12 − run 1) showing snow (gray) and graupel (bold) every 0.04 g kg−1, and rain (thin solid) every 0.02 g kg−1, with negative values dashed. (b) Same as (a), except for run 13 (smaller CNP) minus run 1

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 14.
Fig. 14.

Accumulated 6–12-h surface precipitation (mm) for the different CNP experiments

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 15.
Fig. 15.

(a) Cross-section difference between run 14 (fixed Nos) and run 1 (Nosqs) (run 14 − run 1) showing snow (gray) every 0.08 g kg−1 and graupel (bold) every 0.04 g kg−1, and rain (thin solid) every 0.02 g kg−1, with negative values dashed. (b) Same as (a), except for run 15 (NosT) minus run 1

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 16.
Fig. 16.

(a) Accumulated 6–12-h surface precipitation (mm) for different Nos experiments. (b) Same as (a), except for the snow fall speed (Vs) experiments

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 17.
Fig. 17.

(a) Same as Fig. 10, except showing the difference between (a) run 14 (fixed Nos) and run 1 (Nosqs) (run 14 − run 1). Each process was normalized by the average water vapor loss rate for the two runs before taking the difference. (b) Same as (a), except for run 15 (NosT) minus run 1. Those differences greater than 1% of the water vapor loss rate are in bold.

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 18.
Fig. 18.

Snow fall velocity (m s−1) vs diameter of snow (mm) for different Vs schemes. A dotted line is for the control (run 1), dashed line for Ferrier (run 16), dash–dot for Brown and Swann (run 18), and dash–dot–dot–dot for Cox (run 17)

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 19.
Fig. 19.

Cross section difference between run 16 (Ferrier Vs) and run 1 (run 16 − run 1) showing snow (gray) every 0.04 g kg−1 and graupel (bold) and rain (thin solid) every 0.02 g kg−1, with negative values dashed

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 20.
Fig. 20.

(a) Accumulated 6–12-h surface precipitation (mm) for the different graupel density experiments. (b) Same as (a), except for the graupel fall speed (Vg) experiments

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 21.
Fig. 21.

(a) Accumulated 6–12-h surface precipitation (mm) for the different ice crystal autoconversion experiments. (b) Same as (a), except for the cloud water autoconversion experiments

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Fig. 22.
Fig. 22.

Same as Fig. 10, except for the simple ice BMP. Those processes greater than 5% of the WVL rate are bold. Table 2 lists the process abbreviations

Citation: Monthly Weather Review 132, 12; 10.1175/MWR2821.1

Table 1.

List of microphysical simulations completed for the CA86

Table 1.
Table 2.

The top row shows the windward PE and the microphysical budget for select processes by scaling each process as a percentage of the total water vapor loss rate for the CTL (run 1). The difference between each sensitivity run and run 1 (run X − run 1) is shown for runs 2–25. Each process was normalized by the average water vapor loss rate for the two runs before taking the difference. The difference values are shown in bold when the percentage change between the CTL and experiment is greater than 10%

Table 2.

APPENDIX Abbreviation and Description of Each Microphysical Process in the Reisner2 Scheme

i1520-0493-132-12-2780-ta01

1

The 38 half-sigma levels were σ = 0.997, 0.991, 0.985, 0.978, 0.971, 0.963, 0.954, 0.944, 0.933, 0.922, 0.910, 0.896, 0.881, 0.865, 0.848, 0.829, 0.808, 0.786, 0.763, 0.737, 0.710, 0.681, 0.650, 0.617, 0.583, 0.546, 0.507, 0.468, 0.427, 0.384, 0.341, 0.297, 0.253, 0.209, 0.167, 0.126, 0.086, 0.049.

2

Only hour 6 is shown since this is the only time MC92 presented kinematic and mixing ratio results.

Save