• Baker, R. D., B. H. Lynn, A. Boone, W-K. Tao, and J. Simpson, 2001: The influence of soil moisture, coastline curvature, and land-breeze circulations on sea-breeze-initiated precipitation. J. Hydrometeor, 2 , 193211.

    • Search Google Scholar
    • Export Citation
  • Berry, E. X., and R. L. Reinhardt, 1974: An analysis of cloud drop growth by collection: Part I. Double distributions. J. Atmos. Sci, 31 , 18141824.

    • Search Google Scholar
    • Export Citation
  • Blanchard, D. O., and R. E. Lopez, 1985: Spatial patterns of convection in south Florida. Mon. Wea. Rev, 113 , 12821299.

  • Colle, B. A., C. F. Mass, and K. J. Westrick, 2000: MM5 precipitation verification over the Pacific Northwest during the 1997–99 cool seasons. Wea. Forecasting, 15 , 730744.

    • Search Google Scholar
    • Export Citation
  • Field, P. R., 2000: Bimodal ice spectra in frontal clouds. Quart. J. Roy. Meteor. Soc, 126 , 379392.

  • Grell, G., 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, 121 pp.

    • Search Google Scholar
    • Export Citation
  • Halverson, J., M. Garstang, J. Scala, and W-K. Tao, 1996: Water and energy budgets of a Florida mesoscale convective system: A combined observational and modeling study. Mon. Wea. Rev, 124 , 11611180.

    • Search Google Scholar
    • Export Citation
  • Hunter, S. M., S. J. Underwood, R. L. Holle, and T. L. Mote, 2001: Winter lightning and heavy frozen precipitation in the southeast United States. Wea. Forecasting, 16 , 478489.

    • Search Google Scholar
    • Export Citation
  • Kessler, E., 1969: On the Distribution and Continuity of Water Substance in Atmospheric Circulations. Meteor. Monogr., No. 10, Amer. Meteor. Soc., 84 pp.

    • Search Google Scholar
    • Export Citation
  • Khain, A. P., and I. Sednev, 1996: Simulation of precipitation formation in the eastern Mediterranean coastal zone using a spectral microphysics cloud ensemble model. Atmos. Res, 43 , 77110.

    • Search Google Scholar
    • Export Citation
  • Khain, A. P., M. Ovtchinnikov, M. Pinsky, A. Pokrovsky, and H. Krugliak, 2000: Notes on the state-of-the art numerical modeling of cloud microphysics. Atmos. Res, 55 , 159224.

    • Search Google Scholar
    • Export Citation
  • Khain, A. P., M. Pinsky, M. Shapiro, and A. Pokrovsky, 2001a: Collision rate of small graupel and water drops. J. Atmos. Sci, 58 , 25712595.

    • Search Google Scholar
    • Export Citation
  • Khain, A. P., D. Rosenfeld, and A. Pokrovsky, 2001b: Simulating deep convective clouds with sustained supercooled liquid water down to −37.5°C using a spectral microphysics model. Geophys. Res. Lett, 28 , 38873890.

    • Search Google Scholar
    • Export Citation
  • Khain, A. P., A. Pokrovsky, M. Pinsky, A. Seifert, and V. Philips, 2004a: Simulation of effects of atmospheric aerosols on deep turbulent convective clouds by using a spectral microphysics mixed-phase cumulus cloud model. Part I: Model description and possible applications. J. Atmos. Sci, 61 , 29632982.

    • Search Google Scholar
    • Export Citation
  • Khain, A. P., D. Rosenfeld, and A. Pokrovsky, 2004b: Aerosol effects on dynamics and microphysics of cumulus clouds. Proc. 14th Int. Conf. on Cloud and Precipitation, Vol. 1, Bologna, Italy, ICCP and WMO, 119–122.

    • Search Google Scholar
    • Export Citation
  • Lackmann, G. M., K. Keeter, L. G. Lee, and M. B. Ek, 2002: Model representation of freezing and melting precipitation: Implications for winter weather forecasting. Wea. Forecasting, 17 , 10161033.

    • Search Google Scholar
    • Export Citation
  • Lin, Y-L., R. D. 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
  • Lynn, B., W-K. Tao, and P. J. Wetzel, 1998: A study of landscape-generated deep moist convection. Mon. Wea. Rev, 126 , 928942.

  • Lynn, B., and Coauthors, 2001: Improved simulation of Florida summer convection using the PLACE land model and a 1.5-order turbulence parameterization coupled to the Penn State–NCAR mesoscale model. Mon. Wea. Rev, 129 , 14411461.

    • Search Google Scholar
    • Export Citation
  • Lynn, B., A. Khain, J. Dudhia, D. Rosenfeld, A. Pokrovsky, and A. Seifert, 2005: A spectral (bin) microphysics coupled with a mesoscale model (MM5). Part I: Model description and first results. Mon. Wea. Rev, 133 , 4458.

    • Search Google Scholar
    • Export Citation
  • Mass, C. F., D. Ovens, K. Westrick, and B. A. Colle, 2002: Does increasing horizontal resolution produce better forecasts? The results of two years of real-time numerical weather prediction over the Pacific Northwest. Bull. Amer. Meteor. Soc, 83 , 407430.

    • Search Google Scholar
    • Export Citation
  • McCumber, M., W-K. Tao, J. Simpson, R. Penc, and S-T. Soong, 1991: Comparison of ice-phase microphysical parameterization schemes using numerical simulations of tropical convection. J. Appl. Meteor, 30 , 9851004.

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

    • Search Google Scholar
    • Export Citation
  • Roebber, P. J., and J. Eise, 2001: The 21 June 1997 flood: Storm-scale simulations and implications for operational forecasting. Wea. Forecasting, 16 , 197218.

    • 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
  • Seifert, A., and K. Beheng, 2004: A two-moment cloud microphysics parameterization for mixed-phase clouds. Part I: Model description. Meteor. Atmos. Phys., in press.

    • Search Google Scholar
    • Export Citation
  • Seifert, A., A. Khain, U. Blahak, and K. Beheng, 2005: Possible effects of collisional breakup on mixed-phase deep convection simulated by a spectral (bin) cloud model. J. Atmos. Sci., in press.

    • Search Google Scholar
    • Export Citation
  • Tao, W-K., and J. Simpson, 1984: Cloud interactions and merging: Numerical simulations. J. Atmos. Sci, 41 , 29012917.

  • Tao, W-K., and Coauthors, 2003: Microphysics, radiation and surface processes in a non-hydrostatic model. Meteor. Atmos. Phys, 82 , 97137.

    • Search Google Scholar
    • Export Citation
  • 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
  • Weisman, M. L., W. C. Skamarock, and J. B. Klemp, 1997: The resolution dependence of explicitly modeled convective systems. Mon. Wea. Rev, 125 , 527548.

    • Search Google Scholar
    • Export Citation
  • Yuter, S. E., and R. A. Houze Jr., 1995: Three-dimensional kinematic and microphysical evolution of Florida cumulonimbus. Part I: Spatial distribution of updrafts, downdrafts, and precipitation. Mon. Wea. Rev, 123 , 19211940.

    • Search Google Scholar
    • Export Citation
  • View in gallery
    Fig. 1.

    Model geometry used in simulations. (center) The 3-km domain was used for the nested domain, with the spectral (bin) microphysics (right) nested inside at the same model grid resolution

  • View in gallery
    Fig. 2.

    Observed radar reflectivity on 27 Jul 1991

  • View in gallery
    Fig. 3.

    Accumulated rainfall obtained from portable automated mesonet (PAM) sites (+ sign) and National Climatic Data Center (NCDC) observing stations (diamond) on 27 Jul 1991 between 2100 and 2400 UTC. The analysis was produced using a Barnes interpolation

  • View in gallery
    Fig. 4.

    Calculated model radar reflectivity at a height of 3 km for simulated model runs

  • View in gallery
    Fig. 5.

    Histogram of radar observations and calculated radar results averaged from 1000 to 0100 UTC for 3-km inner domain. [“Bin” refers to method 2 (where reflectivity was calculated from its definition), while “bulk” refers to method 1, where the radar reflectivity was calculated from bulk hydrometeor values]

  • View in gallery
    Fig. 6.

    Average and maximum rainfall obtained from observations and model simulations from the locations of rainfall observations shown in Fig. 3 (within the inner domain of Fig. 1)

  • View in gallery
    Fig. 7.

    Accumulated rainfall between 2100 and 0000 UTC for the model simulations shown

  • View in gallery
    Fig. 8.

    Three-dimensional structure of total mass content in SBM FastM and in Reisner2 at 1900 UTC. Solid line denotes land–sea boundary

  • View in gallery
    Fig. 9.

    Three-dimensional structure of rainwater content in SBM FastM and in Reisner2 at 2300 UTC. Solid line denotes land–sea boundary

  • View in gallery
    Fig. 10.

    South view of total ice content from SBM FastM, Reisner2, and SBM FastC at 2300 UTC. Also shown is an infrared satellite picture at the same time (from NCDC)

  • View in gallery
    Fig. 11.

    Side view of total ice condensate looking down from the southern direction

  • View in gallery
    Fig. 12.

    The side view of total hydrometeor content in SBM FastM

  • View in gallery
    Fig. 13.

    Two-dimensional cross sections of the vertical velocity field from SBM FastM, SBM FastC, and Reisner2 schemes at different times for each, but during peak cloud development

  • View in gallery
    Fig. 14.

    Change in CCN concentration for the cloud shown in Fig. 13, but for the times 20.25 and 20.75 (2015 and 2045) UTC

  • View in gallery
    Fig. 15.

    Same as Fig. 6, but for simulations without breakup (SBM FastNoB) and without effects of turbulence on collision rate (SBM FastNoT). The results from SBM FastM are shown for comparison

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 226 112 18
PDF Downloads 145 82 12

Spectral (Bin) Microphysics Coupled with a Mesoscale Model (MM5). Part II: Simulation of a CaPE Rain Event with a Squall Line

Barry H. LynnThe Hebrew University of Jerusalem, Givat Ram, Jerusalem, Israel

Search for other papers by Barry H. Lynn in
Current site
Google Scholar
PubMed
Close
,
Alexander P. KhainThe Hebrew University of Jerusalem, Givat Ram, Jerusalem, Israel

Search for other papers by Alexander P. Khain in
Current site
Google Scholar
PubMed
Close
,
Jimy DudhiaNational Center for Atmospheric Research, Boulder, Colorado

Search for other papers by Jimy Dudhia in
Current site
Google Scholar
PubMed
Close
,
Daniel RosenfeldThe Hebrew University of Jerusalem, Givat Ram, Jerusalem, Israel

Search for other papers by Daniel Rosenfeld in
Current site
Google Scholar
PubMed
Close
,
Andrei PokrovskyThe Hebrew University of Jerusalem, Givat Ram, Jerusalem, Israel

Search for other papers by Andrei Pokrovsky in
Current site
Google Scholar
PubMed
Close
, and
Axel SeifertUniversity of Karlsruhe, Karlsruhe, Germany

Search for other papers by Axel Seifert in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

Spectral (bin) microphysics (SBM) has been implemented into the three-dimensional fifth-generation Pennsylvania State University–NCAR Mesoscale Model (MM5). The new model was used to simulate a squall line that developed over Florida on 27 July 1991. It is shown that SBM reproduces precipitation rate, rain amounts, and location, radar reflectivity, and cloud structure much better than bulk parameterizations currently implemented in MM5.

Sensitivity tests show the importance of (i) raindrop breakup, (ii) in-cloud turbulence, (iii) different aerosol concentrations, and (iv) inclusion of scavenging of aerosols. Breakup decreases average and maximum rainfall. In-cloud turbulence enhances particle drop collision rates and increases rain rates. A “continental” aerosol concentration produces a much larger maximum rainfall rate versus that obtained with “maritime” aerosol concentration. At the same time accumulated rain is larger with maritime aerosol concentration. The scavenging of aerosols by nucleating water droplets strongly affected the concentration of aerosols in the atmosphere.

The spectral (bin) microphysics mesoscale model can potentially be used for studies of specific phenomena such as severe storms, winter storms, tropical cyclones, etc. The more realistic reproduction of cloud structure than that obtained with bulk parameterization implies that the model will be more useful for remote sensing applications and in the development of advanced rain retrieval algorithms. The model can also simulate the effect of cloud seeding on rain production.

Corresponding author address: Prof. Alexander P. Khain, The Hebrew University of Jerusalem, Givat Ram, Jerusalem, Israel. Email: Khain@vms.huji.ac.il

Abstract

Spectral (bin) microphysics (SBM) has been implemented into the three-dimensional fifth-generation Pennsylvania State University–NCAR Mesoscale Model (MM5). The new model was used to simulate a squall line that developed over Florida on 27 July 1991. It is shown that SBM reproduces precipitation rate, rain amounts, and location, radar reflectivity, and cloud structure much better than bulk parameterizations currently implemented in MM5.

Sensitivity tests show the importance of (i) raindrop breakup, (ii) in-cloud turbulence, (iii) different aerosol concentrations, and (iv) inclusion of scavenging of aerosols. Breakup decreases average and maximum rainfall. In-cloud turbulence enhances particle drop collision rates and increases rain rates. A “continental” aerosol concentration produces a much larger maximum rainfall rate versus that obtained with “maritime” aerosol concentration. At the same time accumulated rain is larger with maritime aerosol concentration. The scavenging of aerosols by nucleating water droplets strongly affected the concentration of aerosols in the atmosphere.

The spectral (bin) microphysics mesoscale model can potentially be used for studies of specific phenomena such as severe storms, winter storms, tropical cyclones, etc. The more realistic reproduction of cloud structure than that obtained with bulk parameterization implies that the model will be more useful for remote sensing applications and in the development of advanced rain retrieval algorithms. The model can also simulate the effect of cloud seeding on rain production.

Corresponding author address: Prof. Alexander P. Khain, The Hebrew University of Jerusalem, Givat Ram, Jerusalem, Israel. Email: Khain@vms.huji.ac.il

1. Introduction

Lynn et. al. (2005, hereafter Part I) used spectral (bin) microphysics (SBM) to simulate cloud development in the three-dimensional fifth-generation Pennsylvania State University–National Center for Atmospheric Research (Penn State–NCAR) Mesoscale Model (MM5). The model includes water drops in warm clouds and drops, ice crystals, aggregates, graupel, and hail in mixed-phase clouds. Each type of hydrometeor is described using size (mass) distribution functions. Nucleation of droplets depends on both aerosol concentration and the supersaturation, and budget equations are used to calculate changes in aerosol concentration that occur because of advection or scavenging.

Part I showed simulated clouds that approached the west coast of Florida, prior to sea-breeze development on 27 July 1991. Clouds in “maritime” aerosol conditions precipitated sooner than clouds in “continental” aerosol conditions. As a result, the maritime clouds quickly spawned secondary convective clouds. But, these clouds were comparatively weaker than the secondary cloud that formed in the simulation with continental aerosols.

Hence, the secondary clouds in the continental air mass had stronger vertical velocity, higher cloud tops, larger maximum rain water content (RWC), and heavier rain rates than the clouds in the maritime air mass. Since the duration of continental-like clouds was longer than that of clouds that developed in clean air, the time increment between formation of the first and the secondary clouds was also longer. The duration of convective clouds was longer in the SBM simulations than in a simulation with a bulk parameterization, and the structure of the clouds was more similar to those shown in the literature in past studies of Florida sea-breeze development (e.g., Yuter and Houze 1995).

As described in Part I, the precision of bulk parameterizations in the reproduction of microphysical processes (e.g., raindrop formation, freezing, and secondary ice generation processes such as rime splitting and fragmentation, sedimentation, breakup, etc.) is limited by a number of factors, including a priori prescription of the size distributions of hydrometeors. In addition, bimodal size distributions of ice particles of the same type (e.g., Field 2000) are difficult to describe using bulk microphysics. Also, differences in assumed mean fall velocities in bulk parameterizations between hydrometeors of different types can lead to an unrealistic separation of these hydrometeors with height. Taking into account that cloud particles fall within a horizontal wind shear, the use of mean fall velocities also introduces errors in the horizontal distribution of cloud hydrometeors. Such errors can adversely impact the simulation of convective clouds and lightning (e.g., Hunter et al. 2001), as well as storm energy budgets (e.g., Lackmann et al. 2002). Perhaps for these reasons, recent studies have cast doubt on whether improved grid resolution will lead to improved precipitation characteristics using bulk parameterizations (Colle et al. 2000; Roebber and Eise 2001; Mass et al. 2002). Furthermore, all microphysical processes depend on changes in the shape of the cloud droplet size distribution that cannot easily be calculated using bulk parameterizations. We suggest that these limitations can lead to large errors in the prediction of precipitation in operational weather forecast models.

This study concerns the same case described in Part I pertaining to the evolution of a convective system over Florida on 27 July 1991, during the Convection and Precipitation Electrification Experiment (CaPE). While Part I concentrated on the simulation of single clouds in microphysically maritime and continental atmospheres, the present study represents the first attempt to use spectral microphysics in a three-dimensional nested grid mesoscale model to simulate a mesoscale rain event accompanied by squall-line development. The structure of the rest of the paper is as follows. In section 2, the design of numerical experiments is presented. Section 3 describes the convection on the experiment day. Section 4 is dedicated to the description of simulation results. A summary and conclusions can be found in section 5, while section 6 describes some current developments in the field of microphysics and an upcoming study.

2. Design of experiments

a. Model setup

A detailed description of model dynamics and microphysics was presented in Part I. Some additional details of the microphysics can be found in studies by Khain and Sednev (1996), Khain et al. (2000, 2001a,b), and Khain et al. (2004b). The main feature of this microphysical scheme is the solving of an equation system for size distribution functions of water droplets and six types of ice particles: columnar, platelike, and branch ice crystals, aggregates (snowflakes), graupel, and hail. The model microphysics is specially designed to simulate cloud–aerosol interaction. For this purpose, the model contains a size distribution function for aerosol particles serving as cloud condensation nuclei (CCN). Note that the version of the microphysics used here and in Part I is referred to as SBM Fast because it uses a set of three size distributions to describe the six original types of ice particles.

Microphysical processes include droplet and ice particle nucleation; diffusional growth/evaporation of droplets; deposition/sublimation of ice particles; secondary generation of ice; all possible collisions between drops and ice particles; freezing of drops, melting of ice, breakup of large drops, etc. The effect of the turbulence– inertia mechanism on collisions is also taken into account.

The design of the experiment is similar to that described in Part I wherein results of early formation of sea-breeze-induced clouds were simulated in experiments with a low concentration of maritime aerosols (experiment SBM FastM) and high concentration of continental aerosols (SBM FastC). To simulate the mesoscale event for a comparatively long period of time and over a significant area, the SBM was used on a 3-km grid mesh (see Fig. 1).

The SBM Fast microphysics covered the area shown in Fig. 1, which is 270 km in the north–south direction and 300 km in the east–west direction. The “Reisner2” bulk-microphysical parameterization (Reisner et al. 1998) ran concurrently on the outer border of the SBM domain. This outer border was approximately 120 km wide. This type of implementation was done to reduce computational time in areas outside the region of interest. Nonetheless, the simulation with SBM required 10 times more processing time than using solely Reisner2 throughout the domain. The mean hydrometeor masses obtained from Reisner2 were decomposed into bin spectra along the boundaries of the SBM micophysics zone as described in Part I.

The SBM results were compared to the Reisner2 scheme and two other bulk parameterizations: “GSFC” (Goddard Space Flight Center) (Tao et al. 2003) and “Schultz” (Schultz 1995). The GSFC scheme (Tao et al. 2003) is a descendent of the Lin et al. (1983) microphysics scheme and predicts bulk contents of cloud water, rain, snow, ice crystals, and graupel/hail. Similarly, the Schultz (1995) scheme predicts cloud, rainwater, ice crystals, snow, and graupel/hail/sleet. A summary of all schemes used is presented in Table 1.

The SBM results presented in the paper were obtained using a 32 processor, SGI 2000 computer. Each simulation took about 8 or 9 days. If recently developed supercomputers were to be used for simulation of MM5 with SBM, we would expect a significant decrease in the time to obtain meaningful results.

b. Dataset

The observational results used in this study include surface rainfall and radar reflectivity. The surface rainfall data are described in Halverson et al. (1996) and Lynn et al. (2001). The observed radar reflectivity was originally provided by Weather Services International and derived from 0.5° base reflectivity scans from the region's radar sites. The resulting product was binned into 5-dBZ intervals, although details of the technique to mosaic each radar site's data are unknown. Unfortunately, radar data were unavailable from 1815 to 1945 UTC.

c. Calculation of model radar reflectivity

To compare the model data with observations, Part I calculated radar reflectivity using a bulk-parameterization method from Tao and Simpson (1984) and McCumber et al. (1991). This is referred to as method 1. Note that this comparison concerns only radar reflectivity results obtained with the model's liquid water for which there are established relationships.

Method 2 can only be used with spectral (bin) microphysics. Here, the “bulk” radar reflectivity is calculated according to its definition (nonparameterized) and is the integral of radar reflectance from each bin (Khain and Sednev 1996). Yet, for purposes of comparison between SBM and results with bulk parameterization, most results are shown using method 1.

3. Design of case study

a. Description of simulated day

Halverson et al. (1996), Lynn et al. (2001), Baker et al. (2001), and Part I each discuss in various detail the convective development that occurred on 27 July 1991. Here, we give a very brief description and introduce an additional figure for comparison later with model observations.

The initial atmospheric conditions across the Florida peninsula on 27 July 1991 had a relatively small initial convective available potential energy (CAPE) of 740 J kg−1. Lynn et al. (1998) showed that an average sounding derived from observations in the CaPE network had a relatively low lifting condensation level pressure of 1010 mb (with a surface pressure of 1018 mb), a level of free convection of 839 mb, and a high equilibrium level of 190 mb. Upon heating and further moistening of the planetary boundary layer, the air mass more than doubled its CAPE (1664 J kg−1), making it quite conductive to the development of deep moist convection. The wind profile was characterized by weak (5 m s−1) low-level westerlies, light winds (<2 m s−1) in the middle troposphere, and weak (∼4 m s−1) easterlies higher up. As discussed by Halverson et al. (1996), a weak short-wave trough crossing the peninsula enhanced convective squall-line development. Initial cloud base was at about 500 m and the freezing level at about 5.1 km.

The prevailing westerly winds triggered summertime convection typical to this type of day (Blanchard and Lopez 1985). Figure 2 presents observed radar reflectivity and shows weak west coast convection followed by much more robust and organized convection later near the east coast. The west coast sea breeze penetrated farther inland than the east coast sea breeze as the prevailing west winds enhanced the movement of the west coast sea-breeze front while hindering the westward movement of the east coast sea breeze.

Satellite pictures (see Lynn et al. 2001) showed that clouds along the west coast sea-breeze front covered the western half of the Florida peninsula by 2000 UTC. At 2100 UTC (1600 LST), cold outflow from the west coast sea-breeze front collided with the east coast sea-breeze front. The focal point of this collision led to enhanced radar reflectivity echoes as seen in Fig. 2. A well-developed squall line resulted and subsequently moved toward the east coast. This squall line extended from northwest of Cape Canaveral south to near Lake Okeechobee. At 2300 UTC, a broad stratiform rain region had expanded to the rear of the dissipating squall line (shown below).

Figure 3 shows rainfall distribution from 2100 to 2400 UTC and reveals two large maxima in rainfall on the eastern side of the peninsula associated with the squall line mentioned above. There is also a large area of light stratiform precipitation surrounding, but mostly to the rear of, these rainfall maxima.

4. Results

a. Comparison of SBM and bulk-microphysics model results

1) Radar reflectivity

Radar reflectivity derived from each model experiment is shown in Fig. 4. At 2100 UTC, SBM FastM produced the radar reflectivity indicative of short convective lines near the squall line over east Florida. In contrast, the SBM FastC and bulk models produced numerous convective cells with less apparent organization. At 2200 UTC, the radar reflectivity in SBM FastM showed the most realistic development of the squall line, although it did not produce the straight-line structure seen in observations (Fig. 2). Both SBM FastM and SBM FastC simulated well the transition of convective cells to weaker reflective stratiform clouds at 2300 UTC. The Reisner2 and GSFC bulk models did not well simulate the squall line, but rather isolated convective cells and short lines. The Schultz scheme produced a much too intense squall line. Only GSFC of the three bulk parameterizations produced radar reflectivity associated with stratiform clouds.

Figure 5 shows a histogram of observed and simulated radar reflectivity. Excluding the times that data were missing, the observational data were averaged from 1000 to 0100 UTC. The simulated data were also averaged over the same time period. Using method 1 (calculations based on the integral values) to calculate the radar reflectivity, SBM FastM and Reisner2 appear to produce the best distribution of radar reflectivity (SBM FastC is not shown). The GSFC and Schultz simulations produced far too numerous grid elements in most reflectivity bins. For the 45–60-dBZ range, the radar reflectivity calculated in SBM FastM using method 1 exceeds the observed radar reflectivity, as well as the value of radar reflectivity calculated using method 2 (which is based on the utilization of size distributions). This reflects the fact that droplet size distribution calculated by the SBM deviates from that assumed in the bulk-parameterization formula. A certain underestimation of reflectivity by method 2 as compared to observations within the range 45–60 dBZ can be attributed to the effect of breakup, which, supposedly, is too strong in the scheme (Seifert et al. 2005).

Table 2 shows the absolute difference between observed and simulated radar reflectivity [using the time averages (shown in Fig. 5) obtained for each 5-dBZ range]. The SBM FastM had the smallest absolute difference. Among the bulk schemes, Reisner2 was the best. When method 2 (explicit bin calculation) was used to calculate the radar reflectivity for SBM FastM, the results for SBM FastM were even better.

2) Rainfall

Figure 6 shows average and maximum rain amounts. The average precipitation is the mean from all observing sites (mostly in east Florida). The maximum precipitation refers to the highest recorded rainfall from any of these data recording sites.

No model simulation exactly reproduced the time dependence of precipitation at the observing sites. In fact, all simulations produced too much rain too early (associated with the east coast sea-breeze front). Besides this, SBM FastM produced the most realistic time evolution of average and maximum rain. Note that the onset of the heavier precipitation in SBM FastC was delayed when compared to SBM FastM, but that SBM FastC simulated larger rain-rate amounts (typical of those occurring from clouds in continental-type air masses). The bulk models produced unrealistically large rain average and maximum amounts, although Reisner2 was again the best.

The average accumulated rain amount between 1600 UTC 27 July and 0100 UTC 28 July 1991 was 0.88 cm. Table 3 shows that SBM FastM and SBM FastC produced accumulated rain closer to observations than Reisner2, GSFC, and Schultz.

The spatial distribution of rainfall for each simulation from 2100 to 2400 UTC is shown in Fig. 7 (similar to Fig. 3). All simulations produced more rain at certain locations than found in observations. However, the location of heavy rain in the SBM FastM compared better with the location of heavy rain in the observations (to the northwest and south of Cape Canaveral). Similar results were obtained in SBM FastC, but it produced an enhanced rainfall maximum just west of Cape Canaveral that was not in the observations. All schemes produced patches (and even lines) of heavy rain not seen in observations; however, the SBM simulations clearly reveal evidence of stratiform rain whereas the bulk schemes do not.

For comparison, Fig. 7 also shows results from a simulation labeled “warm.” This bulk parameterization does not simulate any ice processes and produced fewer regions with heavy rainfall than did the parameterizations with ice processes. The results in the warm simulation suggest that adding ice physics to the bulk-parameterization schemes did not improve the surface rainfall predictability. This can be attributed to limitations of bulk-parameterization schemes in the representation of ice processes as well as ice–water interactions.

3) Cloud structure and vertical velocity

Simulated cloud structures were examined at each hour from 1700 to 2300 UTC. Figure 8 shows a three-dimensional picture of total cloud hydrometeor content (liquid plus ice) for SBM FastM (top) and Reisner2 (bottom) at 1900 UTC. The Reisner2 scheme produced columnar-shape clouds with large, circular anvils. SBM FastM had a number of convective clouds of different size, but without the large anvils.

Figure 9 shows rainwater content, respectively, for SBM FastM and Reisner2 at 2300 UTC. SBM FastM had rainwater structures of different sizes and shapes indicating both supercooled convective rain, as well as stratiform rain (i.e., resembling those observed in situ). In contrast, the scheme by Reisner2 produced columnar rain structures of similar height.

Figure 10 shows that SBM produced ice along convective clouds in the squall line and in a convective cloud line extending westward to the south and rear of the squall line. In Reisner2, ice was produced within the squall line and was transported by strong vertical velocities to the upper atmosphere. Ice then spread out over most of the simulation domain.

An examination of a satellite picture at 2300 UTC (Fig. 10) suggests an area of frozen hydrometeors along the squall line, with an extension westward along the southernmost part of the simulated domain. Comparing with the pictures above, we see that Reisner2 quite overproduces the area coverage of ice. In contrast, SBM FastM captures the basic structure of the ice coverage.

A comparison of total ice contents in SBM FastM and SBM FastC (Fig. 11) indicates that in the case of high aerosol concentration more upper-level ice is formed within the two areas or storm clusters: near the northern end of the squall line and on the western side of the southern cloud extension. In SBM FastC, the maximum vertical velocity was higher than in SBM FastM (see below) and a larger amount of ice crystals was transported to the upper troposphere. Figure 11 also shows that aerosols tend to concentrate convection, at the same time increasing its intensity. This result agrees well with conclusions reached by Khain et al. (2004a,b) in two-dimensional simulations with the SBM model. Figure 12 indicates that the main source of stratiform rain in SBM FastM was the melting of graupel and snowflakes. A similar result was obtained for SBM FastC (not shown).

Figure 13 shows two-dimensional cross sections of the vertical velocity field from SBM FastM, SBM FastC, and Reisner2 schemes at different times and locations for each, but during peak cloud development. Between the three simulations, the weakest maximum vertical velocity is found in SBM FastM, centered near 3 km in height. SBM FastC had a maximum located near 4 km in height. Reisner2 had a maximum in the vertical velocity located near 7 km in height. Yuter and Houze (1995) showed that the majority of the moisture flux was associated with vertical velocities closer in magnitude to SBM FastM.

b. Effect of moist convection on CCN concentration

The SBM model includes the nucleation scavenging of aerosols. Figure 14 shows the change in aerosol concentration in the SBM FastC simulation a half hour prior to and at the same time as in Fig. 13. One can see that most of the scavenging took place within about a half hour (or the time during which the convective cloud developed). The maximum change in aerosol concentration occurred near the top of the convective cloud at about 6.5 km in height. Scavenging has removed more than 60% of the available CCN during the development of the convective cloud. Thus, moist convection significantly decreased the CCN concentration and changed the size distribution. This scavenging process can affect the aerosol conditions in which subsequent clouds arise.

c. No breakup or turbulence

Figure 15 shows the time dependence of average rain and maximum rainfall from two simulations: the first did not include breakup of raindrops (SBM FastNoB), while the second did not include the effects of in-cloud turbulence on cloud particle collisions (SBM FastNoT). Both simulations were started with an initial maritime concentration of aerosols, as in SBM FastM. With no breakup of large drops, the average rainfall in SBM FastNoB remained at its maximum value for a longer period of time than in SBM FastM (or SBM FastC), while the maximum value of rainfall was also increased relative to SBM FastM. In total, the accumulated rainfall was more than in SBM FastM (1.67 versus 1.44 cm). The reason for this is that large drops collide more efficiently than smaller drops, and larger drops evaporate less than smaller drops, thus reaching the ground in larger total mass.

With no impact of in-cloud turbulence on drop collision rate, cloud development in SBM FastNoT took on characteristics more similar to continental clouds. This is because removing in-cloud turbulence increases the time required to convert droplets to raindrops. Yet, the total rainfall was less (1.22 cm), since eliminating in cloud turbulence reduced collisions that can further sustain precipitation.

5. Summary and conclusions

This paper described the implementation and testing of spectral (bin) microphysics (SBM) in the MM5. The microphysics package is based on solving a system of equations for size distribution functions for cloud condensational nuclei (CCN), water drops, and three types of ice crystals, as well as snowflakes (aggregates), graupel, and hail. The model was tested for a convective, mesoscale precipitating system with squall-line formation that developed over Florida on 27 July 1991. The model results suggests that the use of SBM Fast (a faster version of the original SBM physics) in a mesoscale model led to an important improvement in the simulation of cloud structure and cloud shape, average and maximum precipitation amounts, and accumulated precipitation. The spatial distribution of precipitation and simulated radar reflectivity were also closer to observations than the results obtained with various bulk-microphysical parameterization.

The bulk schemes overpredicted maximum vertical velocities, rainfall amounts, and coverage of ice cloud anvil. This suggests limitations in current bulk schemes in their ability to reproduce accurate cloud structure and precipitation. The bulk schemes also significantly underestimated the area covered by precipitation, including stratiform rain.

Even the use of a bin microphysical scheme led to an overprediction of accumulated rainfall amounts. We attribute this discrepancy to the utilization of a relatively low 3-km finite-difference grid. This decreased the vertical updrafts, lowered the level of raindrop formation, and decreased, therefore, the ice mass reaching the upper troposphere. Khain et al. (2004a) showed that a decrease in the grid resolution leads to formation of clouds with more “maritime” characteristics. Furthermore, simulations with coarser resolution produce more slowly developing squall lines than simulations with higher grid resolution and can overpredict the vertical mass transport (Weisman et. al. 1997). We would like to stress, however, that even with the crude model resolution used in the study, SBM indicated a significant improvement in the reproduction of rain rates and accumulated rain amount, as well as spatial structure and type of clouds and precipitation.

The use of the raindrop breakup scheme improved model simulation of rainfall, by limiting the size of large drops. The effect of turbulence on collisions of drops (and drops with graupel) also improved the timing and development of precipitation, leading to more “maritime”-type clouds. However, the sensitivity of the model results to the effect of in-cloud turbulence on collisions (and hence broadening of the droplet spectrum) could have been partly masked by the use of coarse resolution. This is because the coarse grid resolution also fostered formation of wider droplet spectra at lower heights, as was mentioned above. Thus, even greater sensitivity to in-cloud turbulence effects would likely be obtained if finer model resolution were used.

The CCN properties affect the initial drop size spectrum, cloud dynamics, the timing, and maximum precipitation rate. In the area of the squall line, the hourly precipitation rate in the simulation with continental aerosol was higher than in maritime aerosols. It was shown that the use of continental CCN concentration (SBM FastC experiment) led to an overall delay in growth of hourly rainfall over the simulated mesoscale region. At the same time, the enhanced aerosol concentration led to the formation of a stronger squall line with higher updraft maxima. This result supports the conclusion made by Khain et al. (2004b) that aerosols leading to a precipitation decrease from single clouds can foster formation of strong convective structures like squall lines. At the same time accumulated precipitation is larger in the maritime aerosol case.

Since both vertical velocities and supersaturation values are affected by model resolution, more detailed investigation of aerosol effects on cloud dynamics, microstructure, and precipitation, as well as the effects of turbulence, requires better grid resolution.

The aerosol concentration changed appreciably in the simulation with continental aerosols, in response to precipitation processes. SBM Fast microphysics can account for changes in cloud structure that occur during the course of storm development, associated with changes in aerosol concentration.

In the present study, an initially horizontally uniform distribution of aerosols was used. In case more detailed data are available, a nonhomogeneous aerosol distribution can be used, including possible sources of anthropogenic aerosols.

The model can be used for simulation of both aerosol effects on precipitation amount and spatial distribution, the effect of clouds on vertical and horizontal transport of aerosols, estimation of the rate of aerosol scavenging, and atmospheric cleaning by clouds. These results could be of significant importance for parameterization of corresponding processes in large-scale (i.e., regional or GCM) models.

The model includes a detailed description of sophisticated ice processes. It includes the effect of aerosols on warm rain and ice formation. Thus, the model might be used to better understand the microphysical cold processes that depend very much on drop size spectrum that can lead to extreme hail and lightning events and severe wintertime icing and snow, including changes in the storm's energy budget that can affect storm evolution. The model can be used to simulate the effects of cloud seeding on rain enhancement. The simulation of more realistic cloud structure and shape suggests that the model can be better used than bulk parameterization for remote sensing and development of advanced rain retrieval algorithms.

The ability to forecast, with lead times of 12–24 h, the location and timing of clouds, cloud clusters, mesoscale convective complexes (MCCs), squall lines, etc. would be an important step forward in providing accurate forecasts to both ground transportation, aviation interests, and the general public. With the advent of very fast processors and PC clusters, spectral (bin) microphysics has the potential to meet these needs.

6. Future work

The Reisner2 bulk parameterization recently was “upgraded” (Thompson et al. 2004) by incorporating the Berry and Reinhardt (1974) scheme instead of the Kessler (1969) autoconversion formula. This scheme includes a dependence on cloud spectra parameters. A simulation was made with the new Reisner2 scheme. Improved rainfall, but not cloud structure, was obtained. Future studies are planned at higher grid resolution with this version of the Reisner2 scheme as well as with another scheme developed by Seifert and Beheng (2004) and calibrated versus the HUCM spectral microphysics. In this work, we hope to compare modeled three-dimensional fields of cloud type/amount and wind fields with observations. This dataset (derived from radar observations) is currently under development.

Acknowledgments

This study was supported by the Binational U.S.–Israel Science Foundation (Grant 2000215), the Israel Ministry of Sciences (Grant WT 0403), and The Israel Water Company, as well as by the EU project SMOCC. The authors express their deep gratitude to the reviewers for their comments and suggestions that led to significant improvement in both parts of the papers. The authors also thank Robert Rillling at UCAR for providing WSI radar data and to Dennis Buechler and Bill Crosson for processing these data. Derek Posselt provided the figures of three-dimensional cloud structure.

REFERENCES

  • Baker, R. D., B. H. Lynn, A. Boone, W-K. Tao, and J. Simpson, 2001: The influence of soil moisture, coastline curvature, and land-breeze circulations on sea-breeze-initiated precipitation. J. Hydrometeor, 2 , 193211.

    • Search Google Scholar
    • Export Citation
  • Berry, E. X., and R. L. Reinhardt, 1974: An analysis of cloud drop growth by collection: Part I. Double distributions. J. Atmos. Sci, 31 , 18141824.

    • Search Google Scholar
    • Export Citation
  • Blanchard, D. O., and R. E. Lopez, 1985: Spatial patterns of convection in south Florida. Mon. Wea. Rev, 113 , 12821299.

  • Colle, B. A., C. F. Mass, and K. J. Westrick, 2000: MM5 precipitation verification over the Pacific Northwest during the 1997–99 cool seasons. Wea. Forecasting, 15 , 730744.

    • Search Google Scholar
    • Export Citation
  • Field, P. R., 2000: Bimodal ice spectra in frontal clouds. Quart. J. Roy. Meteor. Soc, 126 , 379392.

  • Grell, G., 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, 121 pp.

    • Search Google Scholar
    • Export Citation
  • Halverson, J., M. Garstang, J. Scala, and W-K. Tao, 1996: Water and energy budgets of a Florida mesoscale convective system: A combined observational and modeling study. Mon. Wea. Rev, 124 , 11611180.

    • Search Google Scholar
    • Export Citation
  • Hunter, S. M., S. J. Underwood, R. L. Holle, and T. L. Mote, 2001: Winter lightning and heavy frozen precipitation in the southeast United States. Wea. Forecasting, 16 , 478489.

    • Search Google Scholar
    • Export Citation
  • Kessler, E., 1969: On the Distribution and Continuity of Water Substance in Atmospheric Circulations. Meteor. Monogr., No. 10, Amer. Meteor. Soc., 84 pp.

    • Search Google Scholar
    • Export Citation
  • Khain, A. P., and I. Sednev, 1996: Simulation of precipitation formation in the eastern Mediterranean coastal zone using a spectral microphysics cloud ensemble model. Atmos. Res, 43 , 77110.

    • Search Google Scholar
    • Export Citation
  • Khain, A. P., M. Ovtchinnikov, M. Pinsky, A. Pokrovsky, and H. Krugliak, 2000: Notes on the state-of-the art numerical modeling of cloud microphysics. Atmos. Res, 55 , 159224.

    • Search Google Scholar
    • Export Citation
  • Khain, A. P., M. Pinsky, M. Shapiro, and A. Pokrovsky, 2001a: Collision rate of small graupel and water drops. J. Atmos. Sci, 58 , 25712595.

    • Search Google Scholar
    • Export Citation
  • Khain, A. P., D. Rosenfeld, and A. Pokrovsky, 2001b: Simulating deep convective clouds with sustained supercooled liquid water down to −37.5°C using a spectral microphysics model. Geophys. Res. Lett, 28 , 38873890.

    • Search Google Scholar
    • Export Citation
  • Khain, A. P., A. Pokrovsky, M. Pinsky, A. Seifert, and V. Philips, 2004a: Simulation of effects of atmospheric aerosols on deep turbulent convective clouds by using a spectral microphysics mixed-phase cumulus cloud model. Part I: Model description and possible applications. J. Atmos. Sci, 61 , 29632982.

    • Search Google Scholar
    • Export Citation
  • Khain, A. P., D. Rosenfeld, and A. Pokrovsky, 2004b: Aerosol effects on dynamics and microphysics of cumulus clouds. Proc. 14th Int. Conf. on Cloud and Precipitation, Vol. 1, Bologna, Italy, ICCP and WMO, 119–122.

    • Search Google Scholar
    • Export Citation
  • Lackmann, G. M., K. Keeter, L. G. Lee, and M. B. Ek, 2002: Model representation of freezing and melting precipitation: Implications for winter weather forecasting. Wea. Forecasting, 17 , 10161033.

    • Search Google Scholar
    • Export Citation
  • Lin, Y-L., R. D. 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
  • Lynn, B., W-K. Tao, and P. J. Wetzel, 1998: A study of landscape-generated deep moist convection. Mon. Wea. Rev, 126 , 928942.

  • Lynn, B., and Coauthors, 2001: Improved simulation of Florida summer convection using the PLACE land model and a 1.5-order turbulence parameterization coupled to the Penn State–NCAR mesoscale model. Mon. Wea. Rev, 129 , 14411461.

    • Search Google Scholar
    • Export Citation
  • Lynn, B., A. Khain, J. Dudhia, D. Rosenfeld, A. Pokrovsky, and A. Seifert, 2005: A spectral (bin) microphysics coupled with a mesoscale model (MM5). Part I: Model description and first results. Mon. Wea. Rev, 133 , 4458.

    • Search Google Scholar
    • Export Citation
  • Mass, C. F., D. Ovens, K. Westrick, and B. A. Colle, 2002: Does increasing horizontal resolution produce better forecasts? The results of two years of real-time numerical weather prediction over the Pacific Northwest. Bull. Amer. Meteor. Soc, 83 , 407430.

    • Search Google Scholar
    • Export Citation
  • McCumber, M., W-K. Tao, J. Simpson, R. Penc, and S-T. Soong, 1991: Comparison of ice-phase microphysical parameterization schemes using numerical simulations of tropical convection. J. Appl. Meteor, 30 , 9851004.

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

    • Search Google Scholar
    • Export Citation
  • Roebber, P. J., and J. Eise, 2001: The 21 June 1997 flood: Storm-scale simulations and implications for operational forecasting. Wea. Forecasting, 16 , 197218.

    • 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
  • Seifert, A., and K. Beheng, 2004: A two-moment cloud microphysics parameterization for mixed-phase clouds. Part I: Model description. Meteor. Atmos. Phys., in press.

    • Search Google Scholar
    • Export Citation
  • Seifert, A., A. Khain, U. Blahak, and K. Beheng, 2005: Possible effects of collisional breakup on mixed-phase deep convection simulated by a spectral (bin) cloud model. J. Atmos. Sci., in press.

    • Search Google Scholar
    • Export Citation
  • Tao, W-K., and J. Simpson, 1984: Cloud interactions and merging: Numerical simulations. J. Atmos. Sci, 41 , 29012917.

  • Tao, W-K., and Coauthors, 2003: Microphysics, radiation and surface processes in a non-hydrostatic model. Meteor. Atmos. Phys, 82 , 97137.

    • Search Google Scholar
    • Export Citation
  • 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
  • Weisman, M. L., W. C. Skamarock, and J. B. Klemp, 1997: The resolution dependence of explicitly modeled convective systems. Mon. Wea. Rev, 125 , 527548.

    • Search Google Scholar
    • Export Citation
  • Yuter, S. E., and R. A. Houze Jr., 1995: Three-dimensional kinematic and microphysical evolution of Florida cumulonimbus. Part I: Spatial distribution of updrafts, downdrafts, and precipitation. Mon. Wea. Rev, 123 , 19211940.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

Model geometry used in simulations. (center) The 3-km domain was used for the nested domain, with the spectral (bin) microphysics (right) nested inside at the same model grid resolution

Citation: Monthly Weather Review 133, 1; 10.1175/MWR-2841.1

Fig. 2.
Fig. 2.

Observed radar reflectivity on 27 Jul 1991

Citation: Monthly Weather Review 133, 1; 10.1175/MWR-2841.1

Fig. 3.
Fig. 3.

Accumulated rainfall obtained from portable automated mesonet (PAM) sites (+ sign) and National Climatic Data Center (NCDC) observing stations (diamond) on 27 Jul 1991 between 2100 and 2400 UTC. The analysis was produced using a Barnes interpolation

Citation: Monthly Weather Review 133, 1; 10.1175/MWR-2841.1

Fig. 4.
Fig. 4.

Calculated model radar reflectivity at a height of 3 km for simulated model runs

Citation: Monthly Weather Review 133, 1; 10.1175/MWR-2841.1

Fig. 5.
Fig. 5.

Histogram of radar observations and calculated radar results averaged from 1000 to 0100 UTC for 3-km inner domain. [“Bin” refers to method 2 (where reflectivity was calculated from its definition), while “bulk” refers to method 1, where the radar reflectivity was calculated from bulk hydrometeor values]

Citation: Monthly Weather Review 133, 1; 10.1175/MWR-2841.1

Fig. 6.
Fig. 6.

Average and maximum rainfall obtained from observations and model simulations from the locations of rainfall observations shown in Fig. 3 (within the inner domain of Fig. 1)

Citation: Monthly Weather Review 133, 1; 10.1175/MWR-2841.1

Fig. 7.
Fig. 7.

Accumulated rainfall between 2100 and 0000 UTC for the model simulations shown

Citation: Monthly Weather Review 133, 1; 10.1175/MWR-2841.1

Fig. 8.
Fig. 8.

Three-dimensional structure of total mass content in SBM FastM and in Reisner2 at 1900 UTC. Solid line denotes land–sea boundary

Citation: Monthly Weather Review 133, 1; 10.1175/MWR-2841.1

Fig. 9.
Fig. 9.

Three-dimensional structure of rainwater content in SBM FastM and in Reisner2 at 2300 UTC. Solid line denotes land–sea boundary

Citation: Monthly Weather Review 133, 1; 10.1175/MWR-2841.1

Fig. 10.
Fig. 10.

South view of total ice content from SBM FastM, Reisner2, and SBM FastC at 2300 UTC. Also shown is an infrared satellite picture at the same time (from NCDC)

Citation: Monthly Weather Review 133, 1; 10.1175/MWR-2841.1

Fig. 11.
Fig. 11.

Side view of total ice condensate looking down from the southern direction

Citation: Monthly Weather Review 133, 1; 10.1175/MWR-2841.1

Fig. 12.
Fig. 12.

The side view of total hydrometeor content in SBM FastM

Citation: Monthly Weather Review 133, 1; 10.1175/MWR-2841.1

Fig. 13.
Fig. 13.

Two-dimensional cross sections of the vertical velocity field from SBM FastM, SBM FastC, and Reisner2 schemes at different times for each, but during peak cloud development

Citation: Monthly Weather Review 133, 1; 10.1175/MWR-2841.1

Fig. 14.
Fig. 14.

Change in CCN concentration for the cloud shown in Fig. 13, but for the times 20.25 and 20.75 (2015 and 2045) UTC

Citation: Monthly Weather Review 133, 1; 10.1175/MWR-2841.1

Fig. 15.
Fig. 15.

Same as Fig. 6, but for simulations without breakup (SBM FastNoB) and without effects of turbulence on collision rate (SBM FastNoT). The results from SBM FastM are shown for comparison

Citation: Monthly Weather Review 133, 1; 10.1175/MWR-2841.1

Table 1.

Synopsis of microphysics options used for this paper. An “X” indicates that the microphysics option (first column) simulates the mass of this particular hydrometeor. Reisner2 calculates the concentration of cloud ice (NCI) and uses this information to calculate a more realistic ice mass field as compared to schemes without any information on number concentration. The column labeled GIC refers to a generic ice particle simulated with the different bulk options. For Schultz bulk parameterizations, the column labeled GRIC refers to a general ice hydrometeor consisting of rimed particles of graupel, hail, and sleet. The model labeled GSFC now has an option to simulate either graupel or hail, but this option was not available in the model version used in this study. SBM Fast calculates the number concentration of all cloud particles/hydrometeors

Table 1.
Table 2.

Simulated radar values and radar observations summed every 3 h to obtain the total absolute difference from 1000 UTC 27 Jul to 0100 UTC 28 Jul 1991 for MM5 with SBM FastM, GSFC, Reisner2, or Schultz [using method 1 or bulk (integral) values to calculate radar reflectivity]

Table 2.
Table 3.

Average accumulated rainfall from observations and from model simulations, during the hours 1000 to 0100 UTC. Each model grid point closest to an observing site was identified and used to calculate the average accumulated rainfall

Table 3.
Save