Laserfiche WebLink
<br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br /> <br />systems; and longer-range prediction from global <br />models (including general circulation models), <br />which focus primarily on accurate representations <br />of cloud-radiation interactions. Professor Orville <br />has been a leader in developing state-of-the-art <br />cloud microphysical models for use in many of <br />these types of investigations. Furthermore, as a <br />credit to his mentorship, many of his former <br />students have become leaders in subsequent <br />generations of atmospheric research. <br /> <br />4. BULK MICROPHYSICS <br /> <br />The degree of sophistication of microphysics <br />models is wide ranging, depending on the <br />availability of computational resources and the <br />relative importance of microphysics within the <br />overall modeling effort. Among the most popular <br />and least computationally expensive approaches <br />are bulk microphysical parameterizations. These <br />schemes calculate changes in the mixing ratios of <br />liquid and/or ice. The scheme of Kessler (1969) is <br />still being used in some models to represent <br />warm-rain processes (I.e., absent ice-phase <br />processes) because of its simplicity. Other <br />investigators have extended their microphysical <br />formulations to include ice-phase processes at <br />varying degrees of sophistication. In my opinion, <br />this is the area of research where Prof. Orville's <br />legacy will cast the longest shadows into the <br />foreseeable future. Uu and Orville (1969) added <br />the effects of rain to the cloud model of Orville <br />(1965, 1968). They also formulated an alternative <br />representation for the autoconversion of cloud <br />water to rain that is used in some models. Wisner <br />et a/. (1972) included the representation of hail in <br />a one-dimensional model, and provided excellent <br />documentation of the thermodynamic <br />considerations for modeling hail growth. Orville <br />and Kopp (1977) expanded the Wisner et al. <br />formulation to include cloud ice and incorporated <br />the package into a slab-symmetric two- <br />dimensional (2D) model. Hsie et al. (1980) <br />improved the microphysical interactions between <br />liquid and ice phases, and added a prognostic <br />variable for the seeding agent that included <br />parameterizing a variety . of ice nucleation <br />mechanisms. Un et a/. (1983; LFO) added snow <br />to the scheme, such that two forms of precipitation <br />ice are modeled. <br /> <br />The bulk microphysics developed at the South <br />Dakota School of Mines and Technology <br />(SDSMT) have served as the building blocks for <br />most of the microphysical schemes clirrently <br /> <br />available in the NCAR/Penn State Mesoscale <br />Model Version 5 (MM5) and its successor, the <br />Weather Research and Forecastin RF Model. <br />These sc emes a e composed of a system of <br />prognostic equations for temperature (or its <br />equivalent), water vapor mixing ratio, and mixing <br />ratios of one or more species of condensate in the <br />form of small (nonprecipitating) cloud droplets, <br />small (non precipitating) ice crystals, (precipitating) <br />raindrops, and larger (precipitating) ice particles. <br />The first two species are treated assuming a <br />monodisperse distribution (I.e., they have the same <br />size), which is approximately consistent with <br />observations of narrow droplet spectra in clouds. <br />The last two precipitating species are assumed to <br />have an exponential distribution based on <br />observed spectra (Marshall and Palmer, 1949). In <br />response to diabatic cooling and moistening from <br />ascent, cloud droplets (also referred to as "cloud <br />water") form by condensation. Small ice crystals <br />(also referred to as "cloud ice") are initiated by one <br />or more modes of ice nucleation, and subsequently <br />grow in size by vapor deposition. Small crystals <br />also form through various means of fracturing or <br />breakup of larger ice particles. Rain is initiated <br />either by the self-collection of small cloud droplets <br />into small drizzle-size droplets (80-100 J.l.m) or by <br />melting of ice falling below the DoC level. Rain can <br />also grow by continuous collection of smaller cloud <br />droplets as they fall through the air. Precipitating <br />ice particles form as a result of depositional growth <br />of small ice particles to larger sizes, self-collection <br />(aggregation) of small ice crystals, riming of cloud <br />water onto falling ice particles, and freezing of <br />raindrops carried above the DoC level in strong <br />updrafts. <br /> <br />If <br /> <br />Evolution of bulk schemes since LFO has been <br />primarily variations upon this basic treatment. LFO <br />represented large precipitation ice using two <br />categories: "snow", which is low-density, lightly <br />rimed aggregates with modest terminal fall speeds <br />of 1-2 m S"1, and "hail" (referred to as graupel in the <br />paper) in the form of large particles with a density <br />close to pure ice with fall speeds in excess of 5 m <br />S"1. An important aspect of this approach was a <br />recognition that the particle trajectories are almost <br />always very different between "snow" and "hail" for <br />all types of convective storms organized at all <br />scales of motion. In deep convection, snow is <br />typically carried aloft in strong updrafts into storm <br />anvils or is transported by storm-relative winds into <br />adjacent areas of stratiform precipitation. Hail falls <br />rapidly out of the storm, melts as it falls above OOC <br /> <br />5 <br />